Child Rearing in America: Challenges Facing Parents with Young Children

This page intentionally left blank

Child Rearing in America Challenges Facing Parents with Young Children

Child Rearing in America, a rich and well-researched volume, comes in the wake of intense national interest in young children. Leading scholars from diverse disciplines use relevant data from The Commonwealth Survey of Parents with Young Children to present a wealth of new information about the lives of families with very young children. They explore how parents spend their time with their children, the economic and social challenges they face, and the supports they receive to improve their children’s health and development. Such a broad portrait based on nationally representative data has not been attempted before. Drawing on their extensive expertise and research in the issues being addressed, the contributors examine and elaborate on the Survey findings. They synthesize the major themes emerging from the data and consider the family, community, and policy implications to frame and interpret the results. What emerges is a picture of the complex forces that influence families and child-rearing in the early years. Neal Halfon is Professor of Pediatrics, School of Medicine, and Professor of Community Health Sciences, School of Public Health, at the University of California, Los Angeles, where he serves as Director of the Institute for Children, Families, and Communities. Kathryn Taaffe McLearn is a Senior Fellow and Research Scientist at The National Center for Children in Poverty, Mailman School of Public Health, Columbia University. For the past seven years she has served as the assistant vice president at The Commonwealth Fund, where she oversaw and co-designed a national pediatric initiative, Healthy Steps for Young Children. Mark A. Schuster is Associate Professor of Pediatrics, School of Medicine, and Associate Professor of Health Services, School of Public Health, at the University of California, Los Angeles, and Senior Natural Scientist at RAND. He is Director of the UCLA/RAND Center for Adolescent Health Promotion, which conducts community-based research on adolescent health.

Child Rearing in America Challenges Facing Parents with Young Children

Edited by NEAL HALFON University of California, Los Angeles

KATHRYN TAAFFE McLEARN Columbia University

MARK A. SCHUSTER University of California, Los Angeles, and RAND

          The Pitt Building, Trumpington Street, Cambridge, United Kingdom    The Edinburgh Building, Cambridge CB2 2RU, UK 40 West 20th Street, New York, NY 10011-4211, USA 477 Williamstown Road, Port Melbourne, VIC 3207, Australia Ruiz de Alarcón 13, 28014 Madrid, Spain Dock House, The Waterfront, Cape Town 8001, South Africa http://www.cambridge.org © Neal Halfon, Kathryn Taaffe McLearn, Mark A. Schuster 2004 First published in printed format 2002 ISBN 0-511-03016-9 eBook (Adobe Reader) ISBN 0-521-81320-4 hardback ISBN 0-521-01264-3 paperback

To our families and parents

Contents

Tables Figures Contributors Acknowledgments 1 Introduction and Overview Neal Halfon, Kathryn Taaffe McLearn, Mark A. Schuster part i conditions of families with young children 2 Resources Devoted to Child Development by Families and Society M. Rebecca Kilburn, Barbara L. Wolfe 3 Preparing for Parenthood: Who’s Ready, Who’s Not? Constance T. Gager, Sara S. McLanahan, Dana A. Glei part ii child-rearing practices 4 Meeting the Challenges of New Parenthood: Responsibilities, Advice, and Perceptions Allison Sidle Fuligni, Jeanne Brooks-Gunn 5 Reading, Rhymes, and Routines: American Parents and Their Young Children Pia Rebello Britto, Allison Sidle Fuligni, Jeanne Brooks-Gunn 6 Child Discipline in the First Three Years of Life Lawrence S. Wissow 7 Breastfeeding in the United States Today: Are Families Prepared? Wendelin M. Slusser, Linda Lange

page ix xiii xv xix 1

21 50

83

117 146

178

vii

viii

Contents

8 Depressive Symptoms in Parents of Children Under Age 3: Sociodemographic Predictors, Current Correlates, and Associated Parenting Behaviors Karlen Lyons-Ruth, Rebecca Wolfe, Amy Lyubchik, Ronald Steingard part iii delivery of health services to mothers and children 9 Prenatal Care, Delivery, and Birth Outcomes Paul H. Wise 10 Access to Health Care for Young Children in the United States Paul W. Newacheck, Miles Hochstein, Kristen S. Marchi, Neal Halfon 11 Anticipatory Guidance: What Information Do Parents Receive? What Information Do They Want? Mark A. Schuster, Michael Regalado, Naihua Duan, David J. Klein part iv future directions and policy implications 12 New Models of Pediatric Care Barry Zuckerman, Steven Parker 13 Families with Children Under 3: What We Know and Implications for Results and Policy Neal Halfon, Kathryn Taaffe McLearn Index

217

263 293

320

347

367 413

Tables

2.1 Characteristics Influencing Both Financial and Time Resources page 27 2.2 Allocation of Purchased Resources 28 2.3 Child’s Health Insurance by Provider and Income Category 31 2.4 Allocation of Time Resources 32 2.5 Psychological Resources 35 2.6 Human Capital Resources 37 2.7 Parents’ Use of Government Resources for Child Development 38 2.8 Other Sources of Support for Families with Young Children 39 2.9 Incidence of Risk Factors in Surveyed Families 41 2.10 Number of Risk Factors in Survey Families 42 2.11 Incidence of Risk Factors Relative to Entire Sample 43 3.1 Means and Frequency Distributions for Readiness to Parent Outcome and Predictor Variables 58 3.2 Bivariate Relationships between Parents’ Characteristics and Readiness Indicators 63 3.3 Effects of Parents’ Characteristics on the Odds of Pregnancy Intention Status 66 3.4 Effects of Parents’ Characteristics on the Odds of Parent Education Class Attendance 68 3.5 Effects of Parents’ Characteristics on the Odds of Social Support 69 3.6 Effects of Parents’ Characteristics on the Odds of Coping with Parenthood 71 3.7 Effects of Parents’ Characteristics on Overall Readiness 73 3.8 Effects of Readiness Indicators on Parenting Behaviors 74 3.9 Predicted Likelihood of Being Ready for Parenthood, by Mother’s Marital Status and Education Level 76 ix

Tables

x

4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 4.10 4.11 4.12

4.13 4.14 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 5.10 5.11 5.12 5.13 5.14 6.1 6.2

Division of Child Care Responsibilities in Two-Parent Families (%) Logistic Regression Predicting Incidence of Mother as Primary Caregiver Satisfaction with Time Spent with Child in Two-Parent Families Linear Regressions Predicting Satisfaction with Time Spent with Child Satisfaction with Time Spent with Child in Single-Parent Families Parents Receiving Parenting Information from Hospital Logistic Regression Predicting Information Given at the Hospital Top Four Sources of Parenting Advice Logistic Regression Predicting Information-Seeking from Health Professionals Use of Media Sources for Parenting Advice Logistic Regression Predicting Parent Seeking Information from Media Sources Parents Attending Parenting Classes, by Education and Marital Status (Percentage of parents in each education category) Logistic Regression Predicting Attending Parenting Class Linear Regression Predicting Parental Frustration Levels Frequency of Shared Book Reading Frequency of Playing with Child Maintenance of Family Routines Comparison of Parents Who Read to Their Child Weekly and Daily Comparison of Parents Who Play with Their Child Daily and Weekly Comparison of Parents Who Hug or Cuddle Their Child Daily and Weekly Comparison of Parents Who Do and Do Not Follow Daily Routines Effects of Parental Depression on Daily Shared Book Reading Effects of Parental Depression on Daily Playing with Child Effects of Parental Depression on Following Daily Routines Effects of Information-Seeking on Shared Book Reading Effects of Information-Seeking on Daily Shared Book Reading Effects of Information-Seeking on Daily Playing with Child Daily Shared Book Reading: Results from Two Studies Reported Use of Disciplinary Practices by Child Age (%) Correlates of Spanking by Child Age

90 92 93 94 95 98 98 100 101 102 103

103 104 108 122 123 125 127 128 129 130 131 132 133 134 135 136 141 156 157

Tables 6.3 Logistic Regression Analysis of Factors Associated with Ever Spanking an Infant 6.4 Logistic Regression Analysis of Factors Associated with Ever Spanking a Child Age 12–36 Months 6.5 Correlation of Disciplinary Practices and Parent-Child Activities with Parent Reports of Ever Spanking Their Child 6.6 Characteristics of Parents and Children in Four Clusters, Defined by Discipline and Parent-Child Interactions 7.1 Association between Maternal Characteristics and Health Services Indicators and Infant Feeding Choice for Biological Mothers (%) 7.2 Unadjusted Odds of Breastfeeding Initiation Compared with Never Breastfeeding for Biological Mothers 7.3 Likelihood of Breastfeeding Initiation Compared with Never Breastfeeding, Controlling for Significant Demographic and Health Services Factors (n = 1178) 7.4 Unadjusted Odds of Breastfeeding Duration of Greater than 1 Month, compared with Breastfeeding Duration of Less than 1 Month, for Biological Mothers 7.5 Likelihood of Breastfeeding Duration Greater than 1 Month Compared with Breastfeeding Duration of Less than 1 Month, by Demographic and Health Services Factors (n = 661) 8.1 Potential Predictors and Correlates of Parental Depressive Symptoms: Bivariate Statistics 8.2 Percentage of Parents Endorsing Two or More Depressive Symptoms by Sociodemographic Factors, Prebirth Events, and Current Circumstances 8.3 Hierarchical Regression Model of Factors Associated with Mothers’ Depression: Sociodemographic Factors, Stressful Events, Health, Social Support, and Parenting Confidence 8.4 Hierarchical Regression Model of Factors Associated with Fathers’ Depression: Sociodemographic Factors, Stressful Events, Health, Social Support, and Parenting Confidence 8.5 Parent-Child Interactions Associated with the Level of the Parent’s Depressive Symptoms: Bivariate Odds Ratios 8.6 Predictors of Less Optimal Parent-Child Interactions: Multivariate Analyses 9.1 Prenatal Care Initiation by Health Insurance Coverage 9.2 Association of Selected Variables with Late Initiation of Prenatal Care: Unadjusted (Univariate) and Adjusted (Logistic Regression Model) Odds Ratios 9.3 Association of Selected Variables with Early Discharge, Readmission, and Not Breastfeeding: Adjusted (Logistic

xi

162 163

165 170

190 193

195

196

198 228

230

232

236 242 245 276

278

Tables

xii

10.1 10.2 10.3 10.4 10.5 10.6 10.7 10.8 10.9 11.1

11.2

11.3 11.4 11.5

Regression Model) Odds Ratios and 95% Confidence Intervals Usual Source of Care and Site of Care: U.S., 1993–94 Access Indicators for the Usual Source of Care: U.S., 1993–94 Missed Care: U.S., 1993–94 Delayed Care and Inability to Obtain Care Due to Cost: U.S., 1993–94 Average Annual Physician Contacts: U.S., 1993–94 Satisfaction with Usual Provider among Young Children with a Usual Source of Care: U.S., 1996 Need for and Receipt of Information on Early Childhood Development: U.S., 1996 Percentage of Families Reporting Ever Receiving Specific Services and How Useful They Considered Them: U.S., 1996 Difficulties Paying for Care: U.S., 1996 Percentage of Parents Who Had Not Discussed Each Topic, Who Could Use More Information, and Who Were in “Unaddressed Need” Groups Parents’ Mean Percentile Score for Number of Topics They Did Not Discuss and for Number of “Unaddressed Need” Groups They Are In Receipt of Services and Usefulness of Services Ratings of Clinicians among Parents Whose Child Usually Sees the Same Health Professional Percentage of Parents Who Gave at Least One Nonexcellent Rating and Percentage Willing to Pay an Extra $10 per Month for Anticipatory Guidance Discussions and Special Services

283 300 302 303 305 306 308 310 311 313

326

327 333 334

335

Figures

2.1 Incidence of Additional Risk Factors, Given One Risk Factor by Race/Ethnicity page 42 4.1 Which person are you most comfortable talking with about how to raise your children? 99 4.2 In general, how well do you feel you are coping with the demands of parenthood? 105 4.3 In a typical day, how many times would you say you feel frustrated or aggravated with your child’s behavior or that he/she gets on your nerves? 106 6.1 Child Discipline and Parent-Child Interactions in Clusters 168 8.1 Prevalence of Depression among Mothers and Fathers in Three Types of American Families 227 8.2 When the Child Is Not Firstborn: Prevalence of Depression among Mothers by Family Profile 234 8.3 Percentage of Parents Who Feel Aggravated with Their Young Children Two or More Times a Day by Parental Depressive Symptoms 244 8.4 Percentage of Parents Who Are Low on Two or More Types of Positive Interaction by Parental Depressive Symptoms 247 9.1 Month of Prenatal Care Initiation by Household Income 275 9.2 Maternal Hospital Stay for Vaginal and Cesarean Section Deliveries 282 9.3 Parents Reporting That Their Newborn Spent Too Little Time in the Hospital 285 13.1 Influence of Risk Reduction and Health Promotion Strategies on Health Development 375

xiii

xiv

13.2 Strategies to Improve School Readiness Trajectories 13.3 Readiness to Learn Trajectory: Service Sectors and Programs That Influence School Readiness 13.4 Service Organization for Early Child Development and Parenting

Figures 402 403 405

Contributors

Pia Rebello Britto, PhD, Research Scientist, Center for Children and Families, Teachers College, Columbia University. Jeanne Brooks-Gunn, PhD, Virginia and Leonard Marx Professor of Child Development and Education; Co-Director, Center for Children and Families, Teachers College, Columbia University; Director, Columbia University Institute of Child and Family Policy. Naihua Duan, PhD, Professor in Residence, Department of Psychiatry and Biobehavioral Sciences, School of Medicine, Department of Biostatistics, School of Public Health, University of California, Los Angeles. Allison Sidle Fuligni, PhD, Research Scientist, Center for Children and Families, Teachers College, Columbia University. Constance T. Gager, PhD, Visiting Assistant Professor, Department of Sociology, University of Pennsylvania. Dana A. Glei, PhD, Independent Research Consultant, Santa Rosa, California. Neal Halfon, MD, MPH, Professor of Pediatrics and Community Health Sciences, UCLA Schools of Medicine and Public Health; Director, UCLA Center for Healthier Children, Families and Communities. Miles Hochstein, PhD, Assistant Director, National Center for Infancy and Early Childhood Health Policy of the Center for Healthier Children Families and Communities, UCLA, Los Angeles, California. M. Rebecca Kilburn, PhD, Director, RAND Child Policy Project; Senior Economist, RAND, Santa Monica, California. David J. Klein, MS, Senior Quantitative Analyst, RAND, Santa Monica, California. xv

xvi

Contributors

Linda O. Lange, DrPH, RN, Research Coordinator, Child and Family Health Program; Associate Director, UCLA Breastfeeding Resource Program, UCLA Center for Healthier Children, Families and Communities. Karlen Lyons-Ruth, PhD, Associate Professor of Psychiatry, Cambridge Hospital, Harvard Medical School. Amy Lyubchik, PhD, Postdoctoral Fellow in Psychiatry, Harvard Medical School. Kristen Skenfield Marchi, MPH, Principal Research Analyst, Department of Family and Community Medicine, University of California, San Francisco. Sara S. McLanahan, PhD, Professor, Sociology and Public Affairs, Princeton University, Director, Center for Research on Child Wellbeing. Kathryn Taaffe McLearn, PhD, Senior Fellow and Research Scientist, National Center for Children in Poverty, Mailman School of Public Health, Columbia University. Paul W. Newacheck, DrPH, Professor of Health Policy and Pediatrics, University of California, San Francisco. Steven Parker, MD, Director, Division of Developmental and Behavioral Pediatrics, Boston Medical Center; Associate Professor of Pediatrics, Boston University School of Medicine. Michael Regalado, MD, Associate Professor of Pediatrics and Community Health Sciences, UCLA Schools of Medicine and Public Health. Mark A. Schuster, MD, PhD, Associate Professor of Pediatrics and Health Services, UCLA Schools of Medicine and Public Health; Senior Natural Scientist, RAND; Director, UCLA/RAND Center for Adolescent Health Promotion. Wendelin M. Slusser, MD, MS, Assistant Clinical Professor of Pediatrics, UCLA School of Medicine; Director, Breastfeeding Resource Program, UCLA Center for Healthier Children, Families and Communities. Ronald J. Steingard, MD, Professor of Psychiatry and Pediatrics; Vice Chair, Child and Adolescent Psychiatry, University of Massachusetts Medical School, University of Massachusetts Medical Center. Paul H. Wise, MD, Professor of Pediatrics, Boston University School of Medicine and Boston Medical Center; Lecturer, Harvard Medical School; Associate, Department of Medicine, Children’s Hospital, Boston, MA. Lawrence S. Wissow, MD, MPH, Associate Professor of Health Policy and Management, Johns Hopkins School of Public Health.

Contributors

xvii

Barbara L. Wolfe, PhD, Professor, Department of Economics and Population Health Sciences, University of Wisconsin, Madison. Rebecca J. Wolfe, Doctoral Candidate, Department of Psychology, Harvard University. Barry Zuckerman, MD, The Joel and Barbara Alpert Professor of Pediatrics, Professor of Public Health, Chairman of Pediatrics, Boston University School of Medicine; Chief of Pediatrics, Boston Medical Center.

Acknowledgments

We would like to acknowledge The Commonwealth Fund, which provided the financial support that made this book possible. We express our gratitude to Anne MacKinnon for her expert and gifted editing of the manuscript, to Phinney Leah Ahn for her tireless assistance in helping prepare the manuscript, to Naihua Duan, PhD, for his assistance with statistical analyses, and to David Klein, MA, for his management of the dataset. We would like to thank the contributors for the rich scientific perspectives and insights they brought to each of their chapters. We also acknowledge the support of our editor, Julia Hough.

xix

1 Introduction and Overview Neal Halfon, Kathryn Taaffe McLearn, and Mark A. Schuster

recent perspectives on early childhood The well-being of young children has become a popular topic as expansions of knowledge in the neuro- and behavioral sciences have documented the importance of the first three years in children’s long-term learning, behavior, and health. Recent scholarly books and national research conferences have examined these issues from a variety of perspectives. In 1997, the National Academy of Sciences convened a three-year, multidisciplinary commission to examine and report on the science of early childhood. In 1999, RAND issued Investing in Our Children, which documented the benefits and savings associated with targeted early intervention programs. In 1996, 1997, and 2000, Time and Newsweek devoted entire special issues to the development of the young child. In 1997, the White House hosted conferences on childcare and on early child development and learning. Federal agencies have signaled their interest by expanding data collection on the early life predictors of educational success and supporting expansions of the Head Start program for children from birth to age 3. Numerous states have initiated aggressive early childhood agendas, focusing on childcare, health care, universal preschool, home visiting, and family support (Cauthen et al. 2000). All these efforts are evidence of a recognition of the importance of early child experience and a growing commitment to public policy that fosters the healthy development of our youngest children. The burgeoning focus on early childhood issues in the United States must also be understood in a broader context of concern expressed by the American public about children and their families. Recent polling data from Public Agenda suggest that 82 percent of Americans believe it is harder to be a child today than in past years. By a margin of almost two to one, Americans believe that most parents face times when they really need help raising their children (Public Agenda 1997). The current domestic 1

2

Halfon, McLearn, Schuster

policy focus on improving educational opportunities and outcomes, and a constellation of related political factors, all suggest that the coming decade presents a window of opportunity to address the concerns and needs of families with young children. This strong interest in the lives and potential life trajectories of young children is also being driven by concerns that many young children are seriously and predictably at risk for poor long-term developmental and educational outcomes. For nearly 20 years, poverty rates for children younger than 6 years of age have exceeded 20 percent nationally; in several major urban areas, poverty rates continue to exceed 30 percent. These high rates of early childhood poverty have persisted despite sustained economic growth for most of the 1990s. The deleterious effects of growing up in poverty on child health and development are increasingly well recognized (Duncan and Brooks-Gunn 1997). And there is evidence to suggest that poverty may be especially damaging to the child in the early years (Shonkoff and Phillips 2000). Of related concern is the growing recognition that disparities in health and social outcomes of children, youth, and adults have their origins in disparities that begin early in life (Keating and Hertzman 1999). Disparities in health and emotional states that develop early in life not only persist but are often compounded over time. As one economist writing about policies to foster the growth of human capital observed, “Early learning begets later learning and early success begets later success just as early failure begets later failure” (Heckman 1999). A host of short- and long-term research studies in various fields is providing empirical evidence that disparities in human development manifest early in life and persist across decades. For example, the British 1946 National Birth Cohort Study, which followed more than 5,000 children over half a century, clearly documents that events in early childhood are independent predictors of cardiovascular, respiratory, and neurological health in mid-adulthood (Wadsworth and Kuh 1997; Smith 1999). Children in the cohort who experienced frequent lower respiratory infection in the first two years of life were much more likely to have chronic obstructive pulmonary disease in middle age. Other research studies on emotional and cognitive development have also established links between early life experiences and long-term emotional and intellectual functioning. As the scientific evidence has grown, theoretical constructs to explain the role of family and environmental contexts on child health and development have also evolved significantly. Subtle and important differences in approaches persist, yet a remarkable consensus has emerged. Researchers in a range of disciplines generally accept that family contexts, relationships, and activities play an important role in determining child development and health outcomes (Boyce et al. 1998).

Introduction and Overview

3

Changing Views of Childhood and Family Our understanding of childhood and the social roles of children has changed dramatically over the past century. Children are shaped by the culture in which they grow, and American children do not all begin with the same chances for success. What parents bring to the job of nurturing their children’s development, especially in the early years, is recognized as a critical influence, but so too are political forces, practical economics, and implicit ideological commitments to children and their families. As such, childhood has become a focus of study in many different academic disciplines, each providing a unique perspective on the roles, needs, and prospects of children and their optimal development. r Economists have focused on the cost of raising children and on household inputs needed to produce the circumstances necessary for children’s attainment (Becker 1973). Some studies have examined economic incentives that influence family choices of child care arrangements, health care, and other factors that directly impact the ability to provide for children (Leibowitz 1974; Haveman and Wolfe 1995). Recent research has also focused on the effects of certain kinds of jobs and work relationships on how parents feel, and therefore on how they are able to support their children’s development when they return home from the workplace. r Sociologists have focused on family and social structures and their influence on children’s development and life course transitions. Studies such as Glen Elder’s classic Children of the Great Depression demonstrate that historical forces shape the social trajectories of families, determine the availability of educational and employment opportunities, and thus influence behavior and development (Elder 1974). More recently, Robert Sampson studied adolescents coming of age in Chicago’s inner city in order to dissect the webs of social processes and relationships that influence the life courses of adolescents (Sampson 1997). r Psychologists have examined emotional and cognitive development and the determinants of the onset of psychopathology (Sroufe 1997; Rutter 1996). A range of longitudinal studies display the impact of early life experiences and specific traumatic events, such as childhood sexual abuse, on the developing self. r Pediatricians and child health specialists have focused on the determinants of health status and risks to child health (Hoekleman and Pless 1988), and especially on changes in those areas over the past century. Epidemiological studies had documented dramatic reductions in infant mortality due to improved living conditions and medical care, decreases in childhood infection as a result of immunizations and antibiotics, and the persistent threat of injury as a cause of morbidity and mortality.

Halfon, McLearn, Schuster

4

Changes in the epidemiology of childhood health and disease have also given rise to what some pediatric researchers have termed the “new morbidities” – conditions such as child abuse, learning disorders, and psychopathology that may be heavily influenced by social factors (Haggerty 1975). While each of these research traditions approaches the subject of children and childhood differently, using different conceptual models and analytical tools, each has also necessarily focused on and highlighted the changing nature of the family. While families still play an essential role in the lives of all children, families are changing in size, structure, earning potential, and expectations. Children born in the year 2000 are much more likely to experience divorce, to live in a single-parent household, and to feel certain stresses and demands than were children born 50 years earlier. Today’s children are also more likely to be raised by parents who are better educated, who are more isolated from extended family relationships, and in two-parent households, who are both employed to support the family. What we expect parents to know and do to promote and support their child’s development has also changed. Social norms for parents and parenting expectations have shifted dramatically over the century, and these changes seem to have accelerated over the past decade. Norms regarding acceptable childcare, discipline, and education have been redefined, even during the past two decades. Parents have access to growing volumes of information about child development, discipline, and parenting techniques from a range of professional and eclectic purveyors. Talk shows, specialty magazines, and now the internet have the capacity to provide non-stop information on what to do and how to encourage a child’s future development. Yet little is known about how good this advice is, whether parents who need it are gaining access to it, or whether they are able to act on it in ways that actually promote the development of their children. Is this information helping parents be better parents? We simply do not know. The New Focus on Early Childhood Just as childhood and children have become the focus of study and debate, early childhood has grown to become a field in its own right, with its own areas of research focus and public policy concern. r Economists have highlighted the wisdom of investing in the youngest children, when the potential for return is greatest (Heckman 1999). Recent reports from RAND and the National Bureau of Economic Research have reviewed the economic assumptions that underlay the provision of intervention services early in childhood and calculated the costs and benefits of programs specifically targeted at improving the development of children from birth to age 3 (Karoly 1997; Heckman 1999).

Introduction and Overview

5

r Developmental psychologists have suggested that young children 0–3 learn more about the world, other people, and language than what was previously understood (Gopnik 1999). They have outlined mechanisms that “program” various behavioral response strategies, as well as the role of critical relationships in providing the emotional scaffolding that supports learning and cognitive development (Dawson 1994; Sroufe 1997). r In neurobiology, researchers have highlighted the exceptional flexibility and activity of the brain of the young child, and how profoundly its neuronal structure and function can be influenced by experience. The early neurobiology of cognitive function is coming into greater focus, as are the neurobiology of early emotional states and the influence of selective experiences in the “wiring” of the brain (Schore 1994; Fox 1995). r Epidemiologists and child health researchers have highlighted the origins of adult conditions and disease states that begin early in life. This work suggests that metabolic pathways and response patterns set forth during the first years of life may influence the onset of coronary artery disease, diabetes, and hypertension five or six decades later (Barker 1998; Kuh and Ben-Shlomo 1997). One study, for example, traced the possible programming effect of breast milk on cholesterol metabolism, while another explored the possible influence of early nutrition and growth on non-insulin dependent diabetes (Lucas 1998). r Sociologists and developmental psychologists have juxtaposed the fact that most American parents are now employed outside the home against the fact of young children’s dependence on others for caregiving and nurturing. Even the youngest children are now likely to be cared for by non-family members and childcare centers. r Last, a growing number of studies have demonstrated that life course trajectories can be altered by interventions that change the dynamic relationships within a family early in life (Olds 1997; Ramey 1992; Campbell and Ramey 1995). Although enriched early intervention programs do not seem to alter long-term cognitive outcomes measured by IQ, they substantially alter noncognitive skills and social attachments of participating children and families. Public Policy and Leadership in Early Childhood At the beginning of the 1990s, Beyond Rhetoric (National Commission on Children 1991) and the Carnegie Corporation’s Starting Points (1994) set a new and broader public policy context for considering the needs of families with young children. And in late 2000, the Institute of Medicine and National Research Council released a seminal report, From Neurons to Neighborhoods (Shonkoff and Phillips 2000), which

6

Halfon, McLearn, Schuster

provided a guide to what science-based policies and programs could mean for our nation’s youngest children. The 1997 White House conferences examined the public policy implications of new findings from brain research literature and the role of childcare as more women are employed as full-time workers. Several federal agencies have focused new attention on the development of young children, collecting new data, launching new programs, and expanding oversight and support for states and local communities that are trying to make a difference in the lives of young children. Over the past 15 years, federal legislation has formed the basis for a more defined set of social policies on young children. Each legislative effort not only demonstrated a concern with the development of young children but recognized that the most important single instrumental method for supporting the development of young children is to build family capacity. r Passed in 1986 and fully implemented in the early 1990s, the Individuals with Disabilities Education Act (IDEA) Part C legislation has funded states to identify and address the needs of young children with developmental disabilities and those who are at risk for having developmental disabilities through a comprehensive family-focused approach. r Head Start has grown and expanded during the 1990s, and the federal government created Early Head Start for children 0–3 in 1996. The Early Head Start model also employs early intervention to build family capacity, drawing on home visiting techniques that have been developed and tested over the past two decades. r The 1992 Family and Medical Leave Act allows mothers and fathers to take an unpaid leave of up to 12 weeks in a 12-month period without penalty in the workplace after the birth, adoption, or foster placement of a child and when they need to care for a sick child. At the state level, there has also been widespread policy activity focused on young children. A few examples illustrate the range of these new state initiatives. r North Carolina’s Smart Start Program is an early childhood initiative to promote school readiness in children. r Vermont has created Success by Six, a statewide health and developmental improvement campaign that has had a significant impact on several child outcomes. r California’s Proposition 10, the Children and Families First Act, passed in 1998, is a major initiative that will provide about $700 million annually to improve services focused on children 0–5. Proposition 10 also builds a new quasi-governmental infrastructure in each county to serve as a community outcomes trust for young children.

Introduction and Overview

7

the purpose of this book This book is about families with very young children. It offers a broad and in-depth look at families with infants and toddlers: how they prepared for parenthood, how they spend their time together, the nature and patterns of their routines and practices, their relationships with extended family members and their other supports, and the stresses and strains they experience. Its perspective is multidiscplinary, exploring the range of social, economic, family, and individual factors that interact to alter and shape the development of young children. The contributors include scholars and practitioners from different disciplines, including economics, sociology, developmental psychology, psychiatry, pediatrics, and health policy. The impetus for preparing the book was a unique survey conducted in 1995–96 by The Commonwealth Fund. The Commonwealth Survey of Parents with Young Children provides a representative snapshot of the conditions of families with young children – and the pressures and concerns faced by parents in shaping the home environments and lifestyles of those families. A national sample of more than 2,000 mothers and fathers answered questions about a broad spectrum of topics, including preparation for parenthood, birth of a child, family economic structure, participation in the workforce, child-rearing practices and discipline, child and family health, and access to and satisfaction with child health and developmental services. Their responses were the starting point for many of the analyses presented in this book. The survey did not to focus on child-care issues since so many other surveys, national studies, and books have recently been written about the subject. In fact, the decision was made to focus on more neglected topics relevant to the parents of young children – what they do and believe and the resources they need, especially from the health care system. The contributors were encouraged to complement their analysis of the survey data with other national datasets, a wide range of published studies on families with young children, and relevant theoretical frameworks. Each chapter builds on the survey data to provide a well-rounded analysis of an important area of family life and to examine policies and programs that address the needs of young children and their families. The book is also unique in its focus on the role of the health-care system and health-care providers in supporting the child-bearing and childrearing needs of families. While other recent studies have examined the influence of child-care environments, there has been much less attention paid to health-care providers–representatives of one of the main societal institutions with continuous and ongoing contact with families and young children. Their interactions are sufficiently intense during the first three years of life to provide a unique access point for transferring information and initiating interventions that can benefit children’s health and well-being.

8

Halfon, McLearn, Schuster

The richness of the survey dataset allowed the contributors to examine in detail the patterns of child rearing, the routines and rhythms of family life, and certain behaviors that are potentially sentinel indicators of how mothers and fathers interact with their young children. We have paid particular attention to family routines, especially those focused on reading and early literacy, breastfeeding, and discipline. We believe that the focus on these behaviors and the examination of similarities and differences across geographical regions, racial and ethnic groups, and different family structures tells a great deal about the similarities and differences that exist for children living in the United States today. Survey Methods The Commonwealth Fund developed the Survey of Parents with Young Children to provide an overview of the health and social conditions of families with young children in the United States. It is the first national survey of a representative sample of parents with children ages 0–3 to focus on factors that have been shown by recent research to be important in determining child health and development outcomes, and it also has the advantage of including a large number of fathers in the sample. Questions cover a period beginning before conception and include the early childhood years. Parents were asked about planning to have a child, strategies for child rearing, the challenges of taking care of children while meeting other responsibilities, social supports, and interactions with the medical system. It does not include some important topics, such as childhood immunizations and child care, that have already been covered in detail by other national surveys. A total of 2,017 parents were surveyed during 25–minute telephone interviews carried out between July 1995 and January 1996. Stratified random-digit dialing was used to obtain a nationally representative sample of parents with children 0–3 years of age in the United States. The sample included 1,320 mothers and 697 fathers. African American and Hispanic households were oversampled to facilitate subgroup analyses, and results were weighted accordingly. With oversampling, the sample included 392 non-Hispanic African American parents and 419 Hispanic parents, along with 1,109 non-Hispanic white parents, 84 “other” parents, and 13 parents who did not report race or ethnicity. Respondents were screened by telephone for eligibility, and those who were eligible were asked to participate in an interview. Eligible respondents included biological parents, adoptive parents, stepparents, or other guardians of a child younger than 3 years old living in their households. (Overall, fewer than 1 percent were non-parents.) When two parents were present at the time of the interview, one was randomly selected to participate in the interview. Parents who did not live with their children were

Introduction and Overview

9

excluded. Many of the interview questions were specific to a child younger than 3. If the respondent had more than one child younger than 3 years old, either the oldest or youngest age-eligible child was randomly selected as the index child about whom specific interview questions were asked. This sampling procedure excluded middle children when there were more than two age-eligible children in the same family. Because only 1 percent of families in the study had more than two age-eligible children, this exclusion should not result in any serious bias in the sample. Among 66,633 telephone numbers sampled, eligibility status was determined for 82 percent (54,462) through screening; other means for determining eligibility included the following: telephone number not in service (12,132) and number was for a business, telephone booth, fax, or modem line (7,808). Among 34,522 who completed screening, 8.6 percent (2,959) were found to be eligible to participate in the study. Of those who met the eligibility criteria, 68 percent (2,017) completed the phone interview. The results are weighted to represent the overall distribution of parents with children younger than 3 years old in the United States. For all statistical inference testing, standard errors were adjusted for the design effect (due to unequal sampling probabilities and stratification) using the sandwich variance estimator (StataCorp 1997). Organization of the Book The editors recruited contributors who are experts in their fields and represent a variety of disciplines. After completing their first drafts, the lead contributors gathered at The Commonwealth Fund for a two-day meeting of cross-disciplinary discussion of the findings. In addition, a statistician was available for consultation and review of statistical methods for the analyses. The editors guided the development of the chapters and reviewed analyses and content. The book is organized into four parts. Part I describes the contemporary conditions of families in the United States with children birth to age 3. Part II examines contemporary child-rearing practices. Part III focuses on the health-care system and parents’ relationships with healthcare providers. Part IV looks toward the policy implications of the survey findings. Part I includes two chapters that chronicle the conditions that affect a family’s ability to care for their young children. In the opening chapter of the section (Chapter 2), economists Kilburn and Wolfe present models of child development, highlighting the important role of resource allocation. The authors suggest that four types of resources – financial, time, psychological, and human capital – are inputs to child development. The chapter documents the allocation of resources in these categories by parents, the

10

Halfon, McLearn, Schuster

government, and others from the prenatal period through the first three years of a child’s life; it then develops an index of risk factors for young children. A unique aspect of the chapter and the Survey of Parents with Young Children is the unparalleled information on two important resources for child development: human capital and psychological resources. The authors note that these two resources have received little attention, despite their increasing prominence in the discourse on child-development modeling. They conclude that although parents provide the majority of inputs, a large number of families struggle to provide inputs for their young children. Government inputs are relatively common and mainly serve as a safety net, conditional on a family’s being in need or at risk. In Chapter 3, sociologists Gager, McLanahan, and Glei discuss the psychosocial and familial characteristics associated with being ready to parent among mothers and fathers who have had a child within the last three years. Based on numerous studies that document that intended pregnancies are associated with better parenting and child outcomes, this chapter focuses on two questions: What proportion of parents is ready and what risk factors are associated with preparation for parenthood? The authors examine four dimensions of readiness: intention to become pregnant, participation in childbirth and child rearing classes, availability of social support from family and friends, and psychological resources or preparedness. The authors conclude that the vast majority of mothers and fathers appear to have been prepared for their new responsibilities as parents. At the same time, they caution that some parents seem to have been unprepared for their new responsibilities, and that the likelihood of being unprepared is higher among racial and ethnic minorities and parents with low educational attainment. Some findings were inconsistent with previous work, and the authors suggest possible implications for social policy and programs. For example, these data indicate that closeness to kin appears to be a function of family structure rather than race or ethnicity, and that parents living in nontraditional families are more likely to live near a grandparent. Somewhat surprisingly, first-time parents appear to be more prepared than other parents on almost all indicators of readiness to parent. Part II contains five chapters, which together provide a rich and contemporary picture of child rearing in U.S. families. Each focuses on a set of parenting behaviors that are believed to influence the trajectory of a child’s health and development. These chapters also present a more in-depth analysis of the data on psychological, time, and human capital resources presented in the opening chapter by Kilburn and Wolfe. Developmental psychologists Fuligni and Brooks-Gunn present data in Chapter 4 about how today’s parents of infants and toddlers divide the child-rearing responsibilities, where parents turn for advice and support, and how well they feel they are managing their parenting responsibilities. The findings provide some important information about the effects of

Introduction and Overview

11

employment patterns, and especially the large numbers of mothers with young children in the workforce, on child rearing patterns. Fuligni and Brooks-Gunn find higher rates of shared caregiving than other recent studies of two-parent families, yet mothers are still the primary caregivers of young children, often working the “second shift.” The authors report that parents are coping well with the demands of parenthood in general, but are pressed for time and would like more time with their young children. The data suggest that certain groups of parents – those with low incomes or education or who are unemployed – tend not to avail themselves of formal supports and child-rearing information that can make parenting more satisfying and positive. In Chapter 5, Britto, Fuligni, and Brooks-Gunn examine the frequency with which parents of young children engage in three representative parenting practices: shared book reading, daily routines (bed, nap, and meal time), and nurturing activities such as hugging and cuddling. Summarizing a substantive body of research, the authors suggest that these three parenting activities provide unique opportunities for stimulating cognitive, social, physical, and emotional development in young children. Several common threads emerge from the data that have practice and policy implications. First, the age of the child is an important determinant of parental activities, with parents of infants less likely to read to them and parents of toddlers less likely to play with them, although research indicates that both activities are important for both age groups. Second, only about half of parents with children 1 year or older maintain daily routines. This chapter also adds to existing literature suggesting that young, low income, less-educated parents are less likely to engage in child-rearing activities associated with fostering children’s healthy cognitive, social, and emotional development. The authors discuss the relevance of their findings for welfare reform and initiatives for at-risk parents. Discussing discipline, pediatrician and child psychiatrist Wissow opens Chapter 6 with an extensive review of the history of physical punishment in the United States and the current debate surrounding this particular type of child discipline. Wissow defines both discipline and punishment and summarizes the findings from large national surveys about discipline and child-rearing practices. He notes evidence that the use of physical punishment is declining, although it is practiced in about half of American families, and that parents use it despite having relatively little faith in its effectiveness. Wissow goes on to present findings from the Commonwealth Survey, the only representative study with data about discipline solely from parents of infants and toddlers. The data suggest that counseling about disciplinary practices and alternatives to spanking needs to take place early: More than 40 percent of parents reported that they could use more information on how to discipline their young children. By the time children are 18 months old, more than a third have been spanked.

12

Halfon, McLearn, Schuster

Wissow also identifies the contexts in which spanking takes place, and explains that spanking is most common among a group of parents who tend to be single, have low incomes, and experience more depressive symptoms. These parents were also less likely than others to use nonphysical punishments, such as time out or taking things away. Also, parents of firstborn children were less likely than parents with more children to say they could use more information about discipline – a finding that suggests that parents with more children may be more receptive to parenting advice or that having one child is less stressful. In Chapter 7, Slusser and Lange provide the reader with a unique portrait of the social, demographic, and family factors that influence the decision to breastfeed in the United States today, as well as the role of support and encouragement by professionals. As the authors explain in their chapter, “Breastfeeding in the United States Today: Are Parents Prepared,” breastfeeding is one of the most important child-rearing and healthpromoting behaviors in infancy, yet it is frequently ignored in policy discussions and books on child health and development or family policy. Slusser and Lange present a comprehensive review of the physical and psychosocial benefits of breastfeeding for both child and mother, the economic savings for families, and social barriers to breastfeeding. Their analysis confirms previous findings that education remains the strongest demographic predictor for breastfeeding. Mothers with the lowest breastfeeding rates include single, low-income women, African Americans, those who receive government assistance, and those with lower levels of education. According to the survey, almost three-quarters of mothers report being encouraged to breastfeed in the hospital and that professional support was associated with the initiation of breastfeeding but not its duration. The authors conclude with a discussion of the policy, programmatic, and legislative efforts to overcome barriers to breastfeeding. Part II closes with a discussion by Lyons-Ruth, Wolfe, Lyubchik, and Steingard of the prevalence and impact of parental depressive symptoms on child rearing. The authors begin Chapter 8 with an extensive literature review documenting the increased incidence of maternal depression among women of childbearing age and its association with impairments in adaptive parenting behaviors and developmental outcomes for children. The Commonwealth data confirm widespread developmental findings that negative correlates of parental depression are evident during a child’s first three years of life. Unique in the inclusion of both mothers and fathers, the data indicate that parents of both genders who are depressed are less likely to play, read, maintain routines, and hug or cuddle their infant or toddler, and are more likely to be frustrated, aggravated, and use negative discipline such as yelling, spanking, and hitting. The authors suggest that even as parental depression constitutes a risk factor for young children, young children constitute a risk factor for adult depression.

Introduction and Overview

13

Among fathers who assume primary child-rearing responsibility, the rate of depressive symptoms is equal to that of mothers who are their children’s primary caretakers. Part III focuses on the delivery of health services to young children and their families in three chapters on diverse topics: prenatal care and perinatal services, young children’s access to health care, and the parent-physician relationship. In Chapter 9, on prenatal care and perinatal services, Wise describes recent trends in the use of prenatal care and the implications of a changing medical marketplace – most notably, early discharge – on the care of mothers and newborns. The author concludes that although a vast majority of women in the United States start prenatal care early, powerful social disparities still exist in the use of prenatal care. A mother’s income, race, age, educational status, and intention to become pregnant are all associated with prenatal care utilization and attendance at childbirth classes. The survey findings also document that lack of insurance significantly elevates the chance that a woman will begin prenatal care late or not at all, and not attend childbirth classes. Wise judiciously suggests that these factors are likely to be highly interactive, “reflecting the underlying clustered expression of social status in American daily life.” The author reports that one-day hospital stays have become common throughout the United States, with no associated wide-scale use of at-home follow-up for low-risk mothers. The survey shows that despite their concerns about short stays, mothers report high levels of satisfaction with their hospital care. In Chapter 10, Newacheck, Hochstein, Marchi, and Halfon provide a comprehensive profile of young children’s access to health care in the United States using two complementary data sources: the Commonwealth Fund Survey of Parents with Young Children and the 1993–94 National Health Interview Survey and its supplements on access to care and health insurance coverage. The authors show that partly as a result of Medicaid expansions, the vast majority of young children have a usual source of health care, yet there is still a group of infants and toddlers – many of them poor, minority, or uninsured – who are more likely than other young children to go without a usual source of pediatric care. These children are significantly less likely to receive their care in a physician’s office, private clinic, or HMO, and more likely to use community health centers and hospital outpatient clinics. The authors also document the fact that poor and uninsured children face the greatest obstacles in obtaining needed services such as dental care, prescriptions, or mental health services. Using both datasets, the authors report that poor, nonwhite, and uninsured parents are more likely to report being dissatisfied with the health care their child receives and less likely to get information on child growth and development. Nearly one in ten young children had no coverage for preventive services such as well child care and immunizations.

14

Halfon, McLearn, Schuster

Part III closes in Chapter 11 with a discussion by Schuster, Regalado, Duan, and Klein of the parent-physician relationship, the evolution of the role of the physician as educator and counselor, and the expectations of the physician in this role for parents with young children. The authors suggest that the pediatric clinician’s role has been transformed over time by multiple factors, including changes in traditional pediatric morbidity and mortality, the growth of pediatric subspecialties, increased knowledge about normal child development, and parents’ expectation that the health system will address psychosocial concerns in addition to traditional child health issues. They also note that the current literature gives little indication of how adequately pediatric clinicians are meeting parents’ needs for child-rearing information. Using the Commonwealth Fund Survey of Parents with Young Children, the authors find that most parents are not discussing child-rearing information with their child’s doctor, despite the fact that they say they want more information on learning, discipline, sleep, toilet training, and other topics. Schuster, Regalado, Duan, and Klein conclude that parent satisfaction with care is positively related to the number of topics discussed and the breadth of services provided. Part IV presents a discussion of future directions and policy implications. In Chapter 12, pediatricians Zuckerman and Parker describe new models of pediatric care, and list sociocultural forces that influence the need for new models: rapid social change, growing income inequity, new knowledge about early development, and the link between parental health and child health. The chapter draws on the data highlighted in the rest of the book as well as research literature on early intervention. The authors suggest that pediatric practice needs to strengthen its emphasis on child development, using five related strategies: (1) expanding child development services, (2) creating a two generation approach to child health, (3) strengthening links to other services, (4) providing parental support, and (5) serving as an advocate for parents. Building on their clinical experience and work with families, Zuckerman and Parker suggest a clinical model, Healthy Steps, as one approach to enhance pediatric care. The authors conclude with a discussion of the potential benefits of developmentally oriented primary care interventions in an era of cost containment. In the concluding chapter (Chapter 13), editors Halfon and McLearn consider how the research findings and interpretations presented by the authors could inform social and health policy for families with very young children. The discussion begins by considering the family, community, and policy context of early childhood, then goes on to summarize major findings and suggest policy implications. In looking across the chapters, what becomes clear is that income, education, employment, and other family relationships are important predictors

Introduction and Overview

15

of child-rearing practices and behaviors. The encouraging news is that the survey data indicate that information and guidance can make a difference in parents’ child-rearing practices and their ability to cope with the demands of parenthood. The discouraging news is that the United States has not yet been able to translate what has emerged from research on child development and family functioning into universal policy responses. The health-care system potentially plays a major role in the lives of young children, yet it misses important opportunities to help many parents do their best in rearing young children. Using new information and examples of innovation in communities across the United States, local communities can pursue integrated strategies to promote the healthy development of their youngest residents. Moreover, state and federal policy makers must consider a more integrated policy response based on the services and supports that many families with young children need.

references Barker, D. (1998). Mothers, babies, and health in later life. Edinburgh, Scotland: Churchill Livingstone. Becker, G., & Lewis, H.G. (1973). On the interactions between the quantity and quality of children. Journal of Political Economy, 81(2), S279–288. Boyce, W.T., Frank, E., Jensen, P.S., Kessler, R.C., Nelson, C.A., & Steinberg, L. (1998). Social context in developmental psychopathology: Recommendations for future research from the MacArthur Network on psychopathology and development. Development and Psychopathology, 10, 143–164. Brooks-Gunn, J., & Duncan, G.J. (1997). The effects of poverty on children. The Future of Children, 7(2), 55–71. Campbell, F.A., & Ramey, C.T. (1995). Cognitive and school outcomes for highrisk African American students at middle adolescence: positive effects of early intervention. American Education Research Journal, 32(4), 743–772. The Carnegie Corporation of New York. (1994). Starting Points: Meeting the needs of our youngest children. New York: Carnegie Corporation. Cauthen, N., Knitzer, J., & Ripple, C. (2000). Map and Track: State Initiatives for young Children and Families. New York, NY: National Center for Children in Poverty. Dawson, G., Hessel, D., & Frey, K. (1994). Social influences on early developing biological and behavioral systems related to risk for affective disorder. Development and Psychopathology, 6, 759–779. Duncan, G.J., & Brooks-Gunn, J. (1997). Consequences of Growing up Poor. New York: Russell Sage Foundation. Elder, G. (1974). Children of the Great Depression. Chicago: University of Chicago Press. Farkas, S., & Johnson, J. (1997). Kids These Days: What Americans Really Think about the Next Generation. New York: The Public Agenda, 17. Fox, N.A., Rubin, K.H., Calkins, S.D., Marshall, T.R., Coplan, R.J., Porges, S.W., Long, J.M., & Stewart, S. (1995). Frontal activation asymmetry and social competence at four years of age. Child Development, 66, 1770–1784. Gopnik, M. (1999). Familial language impairment: More English evidence. Folia Phoniatrica Logopedica, 51(1–2), 5–19.

16

Halfon, McLearn, Schuster

Halfon, N., Inkelas, M., & Hochstein, M. (2000). The health development organization: An organizational approach to achieving child health development. The Milbank Quarterly, 78(3), 447–497. Haggerty, R.J., Roghmann, K.J., & Pless, I.B. (1975). Child health and the community. New York: John Wiley & Sons. Haveman, R., & Wolfe, B. (1995). The determinants of children’s attainments: A review of methods and findings. Journal of Economic Literature, 33, 1829–1878. Heckman, J.J. (1999). Policies to foster human capital. NBER Working Paper Series, Working Paper 7288. Cambridge: National Bureau of Economic Research. Hoekleman, R.A., & Pless, I.B. (1988). Decline in mortality among young Americans during the 20th century: Prospects for reaching national mortality reduction goals for 1990. Pediatrics, 82(4), 582–595. Karoly, L., Greenwood, P.W., Everingham, S.S., Hoube, J., Kilburn M.R., Rydell, C.P., Sanders, M., & Chiesa, J. (1997). Investing in our children. Santa Monica: RAND Corp. Keating, D.P., & Hertzman, C. (1999). Developmental health and the wealth of nations: Social, biological and educational dynamics. New York: The Guilford Press. Kuh, D., & Ben-Shlomo, Y. (1997). A life course approach to chronic disease epidemiology. New York: Oxford University Press. Leibowitz, A. (1974). Home investment in children. Journal of Political Economy, 82(2), S111–131. Lerner, R.M. (1998). Theories of human development: Contemporary perspectives. In W. Damon & R.M. Lerner (eds.), Handbook of Child Psychology. New York: John Wiley. Lucas, A. (1998). Programming by early nutrition: An experimental approach. Journal of Nutrition, 128, 401S–406S. National Commission on Children. (1991). Beyond rhetoric: A new American agenda for children and families. Washington DC: U.S. Government Printing Office. NRC National Survey of Health and Development. Paediatric Perinatal Epidemiology, 111(1), 2–20. Olds, D.L., Eckenrode, J., Henderson, C.R., Kitzman, H., Powers, J., Cole, R., Sidora, K., Morris, P., Pettitt, L.M., & Luckey, D. (1997). Long-term effects of home visitation of maternal life course and child abuse and neglect. Fifteen-year follow-up of a randomized trial. JAMA, 278(8), 637–643. Penn, A.A., & Shatz, C.J. (1999). Brain waves and brain wiring: The role of endogenous and sensory-driven neural activity in development. Pediatric Research, 45(4), 447–458. Ramey, C.T., Bryant, D.M., Wasik, B.G., Sparling, J.J., Fendt, K.H., & LaVange, L.M. (1992). Infant health and development program for low birth weight, premature infants: Program elements, family participation, and child intelligence. Pediatrics, 3, 454–465. Rutter, M. (1996). Developmental psychopathology: Concepts and prospects. In M. Lenzenweger & J. Havgaard (eds.), Frontiers of developmental psychopathology. New York: Oxford University Press. Sampson, R.J., Raudenbush, S.W., & Earls, F. (1997). Neighborhoods and violent crime: A multilevel study of collective efficacy. Science 277, 918–924.

Introduction and Overview

17

Schore, A.N. (1994). Affect regulation and the origin of the self: The neurobiology of emotional development. Hillsdale, NJ: Lawrence Erlbaum, Assoc. Shonkoff, J.P., & Phillips, D. (2000). From Neurons to Neighborhoods: The Science of Early Childhood Development. Washington, DC: National Academy Press. Smith, J. (1999). Healthy bodies and thick wallets. The dual relation between health and economic status. Journal of Economic Perspectives, 13(2), 145–166. Sroufe, L.A. (1997). Psychopathology as an outcome of development. Development and Psychopathology, 9(2), 251–268. StataCorp. Stata Statistical Software: Release 5.0. College Station, TX: Stata Corporation. 1997. Wadsworth, M.E., & Kuh, D. (1997). Childhood influences on adult health: A review of recent work from the British 1946 national birth cohort study, the NRC National Survey of Health and Development. Paediatric Perinatal Epidemiology, 111(1), 2–20.

part

I

CONDITIONS OF FAMILIES WITH YOUNG CHILDREN

2 Resources Devoted to Child Development by Families and Society M. Rebecca Kilburn and Barbara L. Wolfe

Both the public sector and families invest heavily in children’s development. One estimate of the average total cost of these investments per child is about $16,030 annually (1992 figures, converted to 1999 dollars), or about 15 percent of GDP (Haveman and Wolfe 1995). Of the total average spending on a child from birth to age 18, about one-third represents public investment and the other two-thirds, private expenditure. This ratio of public to private investment in children is not constant throughout childhood or across the income distribution. Over two-thirds of government spending on children supports education, which typically does not begin until age 5. Before children enter elementary school, the family is the primary source of inputs into child development. In recent years, the child development literature has elevated the early years to a new place of prominence in the human life course. Since inputs into early child development are primarily the purview of parents, realizing the potential of early development requires an understanding of parenting behaviors and the contextual factors that influence parents’ allocation of resources. The Commonwealth Survey of Parents with Young Children focuses on families with children in their early years, from birth to age 3. This chapter documents the allocation of resources to child development made by parents, government, and others during this critical period of development. In doing so, we recognize that families’ investments in children include not only purchased goods and services but also considerable time. In their estimates of spending on children, Haveman and Wolfe (1995) include direct out-of-pocket costs, such as food, housing, and health care, along with indirect costs, or the value of time parents spend caring for children. They estimate that about 80 percent of parents’ expenditures on children are in the form of direct costs, with the remaining 20 percent due to indirect time costs. They admit, however, that their estimate of indirect time costs is necessarily a “lower bound” because of the 21

22

Kilburn, Wolfe

challenges inherent in accounting for all the time parents spend caring for children.

child development models and the importance of resources Many different approaches are used to model child development and the role of families and society in that process. These approaches reflect a range of aspects of young children’s growth and development, including cognitive development, physical growth, emotional maturation, social functioning, and more. They also span academic disciplines, encompassing the social sciences, medical science, demography, and others. Prominent child development models include the economic or rational choice framework (see Haveman and Wolfe 1995; Becker and Tomes 1986), the ecological context model (Bronfenbrenner 1979; Garbarino 1990), social capital theory (Coleman 1988), and other perspectives (see discussion in Brooks-Gunn et al. 1995). Rather than selecting one model to describe child development, we choose in this discussion to emphasize similarities among the various approaches. Since the intersections of the models often guide data collection, this discussion should help place the Commonwealth Survey in the context of the child-development literature and highlight the unique features of the data. One similarity that connects many child-development models is the recognition that children do not start life with equal chances of success. That is, most approaches to child development identify factors that may place some children at higher risk of poor outcomes. These factors, often called “risk factors,” may include disadvantaged socioeconomic status, physical disabilities, parents’ psychological problems, or other characteristics that have been linked to developmental shortcomings (see Karoly et al. 1998; Meisels and Wasik 1990). A second similarity is that most models characterize child development as a process. That is, child development happens over time and has numerous phases, and subsequent development builds on foundations established at earlier ages (see discussion in Brooks-Gunn et al. 1995; Werner 1990). As a result, research on children’s development often focuses on a particular age or period of development, such as prenatal development, adolescence, or preschool years. As part of the recognition of child development as a multistage process, some models consider outcomes of earlier stages to be inputs at later stages (Brooks-Gunn et al. 1995). For example, the social skills and pre-reading skills a child develops in preschool can be considered inputs into development in the early elementary school years. Third, the majority of models view a child’s environment as playing a key role in development. Many view child development as taking place in a

Child Development Resources

23

multilayered context that involves not only families but also schools, communities, and varying levels of government (Blank 1997). The bioecological systems perspective characterizes the relationship between individuals and their environment – nature and nurture – as a bidirectional interaction that produces an individual’s actual traits from his or her potential traits (Sameroff et al. 1987; Bronfenbrenner and Ceci 1994). According to this model, without appropriate environmental stimulation, some traits may not develop fully even if an individual inherits the genotypic predisposition to exhibit the trait. Finally, all models of child development stress the importance of resources, which provide inputs into the child-development process. In the early years that are the focus of the Commonwealth Survey, these resources are provided predominantly by parents. Although government allocates some resources to early childhood programs for all families (such as educational campaigns regarding immunizations and subsidized vaccines), it targets more resources to families at higher risk of being unable to provide adequate resources for child development due to specific parental circumstances such as poverty or chemical dependence. For these families, the government might provide special assistance, such as children’s health insurance or free vaccinations. In addition to parents and the public sector, other individuals, including extended family members, religious organizations, or neighbors, may also provide inputs into child development. This analysis of the resources allocated to support child development in the early years examines the four types recognized by Brooks-Gunn et al. (1995): financial, time, psychological, and human capital. Financial resources, or the economic means to purchase goods and services that promote child development, are referred to in this chapter as “purchased resources.” The relationship between purchased resources and children’s outcomes has been explored extensively through data collection and analysis (see, for example, Mayer 1997b; Duncan and Brooks-Gunn 1997). Parents’ investment of time in children’s development has received less research attention, although its importance has captured the interest of the public: In a recent opinion poll, the small amount of time parents spent with their children was cited as the number one problem for U.S. families (Benton Foundation 1998). “Time investments” are not restricted to time spent in enriching activities, such as reading. Rather, these inputs are a measure of all time spent producing child-related goods and services in the home rather than purchasing them in the market. Child care, for example, can be purchased in the market or provided by a parent. In standard models of economic choice, the decision to produce a good using purchased resources or time resources relies on the quantity of both available, the relative price of the good in terms of income or time, and the parents’ efficiency in converting time to income (see Becker and Michael 1976, for an exposition of this model). For example, children’s meals can be produced using

24

Kilburn, Wolfe

varying amounts of parents’ time and purchased goods or by purchasing meals from a restaurant. As discussed in more detail later, because families’ purchased resources are strongly related to income earned during time parents spend working, purchased resources and time resources are inextricably linked. Although adequatepurchased resources and time are critical to a family’s ability to provide inputs into child development, the availability of those resources does not guarantee that parents will allocate them in a way that optimizes child development. In addition to making choices about how much purchased and time resources they devote to child development, families must also decide how to alleviate allocate resources to various alternative uses within the realm of child development (see, for example, Leibowitz 1974 and Mulligan 1997). These allocation decisions are related to the two remaining types of resources: psychological and human capital resources. While analyses of purchased resources and time resources focus on the quantity of inputs families provide for child development, analyses of psychological and human capital resources stress the quality of the resources. “Psychological resources” are emotional, cognitive, and behavioral factors such as parental knowledge, depression, or substance abuse that can play a role in child development. A substantial body of research points as well to the importance for child development of another type of psychological resource: parenting style (Maccoby and Martin 1983; Beckwith 1976; Barnard and Kelly 1990). The term “parenting style” refers to a collection of behaviors that includes, for example, degree of willingness to allow a toddler to explore and tendency to display warmth toward the child. The hazards of parents’ drug and alcohol use for child development have been recognized for some time (Sher 1991; Bauman and Levine 1986). The role of parents’ depression, and in particular the depression of low-income mothers, in children’s development has received more attention in recent years (Beckwith 1990). The last type of resource that contributes to child development is human capital. In this context, “human capital” denotes the skills and knowledge parents bring to their decision making, interaction, and other behaviors that affect child development. To deliver an appropriate diet to a child, for example, a family needs knowledge about nutrition, as well as adequate financial and time resources to purchase and prepare food. Parents’ human capital may be acquired through experience, but it can also be increased through information from health care providers, reading, or other avenues.

the allocation of resources to support child development The common features of child-development models provide a context for describing the results of the Commonwealth Survey of Parents with Young Children. In this section, we generate a portrait of parents’ and societal

Child Development Resources

25

support for child development as portrayed by these data. The survey interviewed 2,017 parents of children less than age 3, focusing on inputs to child development in the prenatal period, birth, and infancy. While some surveys target families that meet some risk factor criteria, the Commonwealth study interviewed a random sample of parents with small children. Eligible parents included biological parents, adoptive parents, step parents, or guardians of a child under age 3 living in their household. At each sampled household, either the child’s mother or father was randomly selected to complete the interview. Of the 2,017 families interviewed, 1,320 respondents were mothers or female guardians and the remaining 697 were fathers or male guardians. Blacks and Hispanics were oversampled to generate large enough samples for comparative analysis. To account for the fact that blacks and Hispanics were not sampled in proportion to their true representation in the population, all statistics in this chapter are weighted to reflect the national population of parents in the United States with children less than 3 years old. A comparison of the demographic characteristics of the Commonwealth Survey data to a nationally representative sample, the 1994 Current Population Survey, indicates a very close match on every item. The survey asked all respondents to provide information about experiences during pregnancy, as well as immediately after the child’s birth. The children in the survey are about equally distributed across each of the first three years of age. Nearly a fifth (18 percent) of the children are between 0 and 5 months of age, and another 14 percent are 6 months to 1 year old. About a third (36 percent) are 1 year old, and nearly another third (32 percent) are 2 years old. Less than half the children (41 percent) are the firstborn child of the mother. The sample contains almost equal numbers of boys (51 percent) and girls (49 percent). The majority of the children live in families with two or fewer children (71 percent). Another 18 percent of families have three children, and 11 percent have four or more children. We examine information collected by the survey regarding the resources devoted to young children’s development. We begin by discussing the resources parents allocate to support child development, distinguishing between families’ purchased, time, psychological, and human capital resources. Next, we describe the families’ reliance on government resources for child-development inputs. Finally, we enumerate other sources of support for child development that do not come from families or government. Parental Inputs: Purchased Resources and Time Resources The standard economic model of family decision making (Becker 1991) posits that a parent has a total amount of time, say T, that can be divided between working, TW , and not working, T N , where T = TW + T N . The more time a parent spends working – the higher is TW – the more financial resources will be available to a family to purchase direct-cost items such

26

Kilburn, Wolfe

as food, housing, and health care (Haveman and Wolfe 1995), yet the less time the parent will have to engage in activities that contribute to child development. Haveman and Wolfe categorize these as indirect costs of investing in children. In the parlance of economic models of household decision making, the time not spent working is available to devote to “home production.” This relationship between purchased resources and time resources formalizes the reason that two-parent households are better able to invest in children: With two parents in the household, 2T rather than T amount of time is available to allocate between working and not working. The net result is that most two-parent families are able to choose a combination of purchased and time resources that exceeds the choices available to most one-parent families. An unresolved question in the literature is the relative merit of allocating more time to home production or to procuring financial resources through employment (Mayer 1997a; Smith et al. 1997). For single mothers, welfare reform policy has chosen employment: Temporary Assistance for Needy Families, the program that has replaced the prior federal welfare system, encourages mothers to work rather than be full-time caregivers for their children. The Survey of Parents with Young Children reports on several characteristics that influence the total time available to a family to devote to working or home production, including number of parents in the household and their employment status. Although current divorce rates are high by historical standards, and nonmarital births continue to rise (Zill and Nord 1994), the overwhelming majority of families in the sample (86 percent) contain two adults (Table 2.1). Nearly every two-parent household has at least one parent devoting part of his or her time to paid employment (97 percent) that produces financial resources available to the family. In single-parent families, less than half (43 percent) of parents are in the labor force. The difference is even more striking when rates of full-time participation in the labor force are considered: 95 percent of two-parent families have a parent who works full time, compared with 30 percent of one-parent families. The difference between the two groups of families is less dramatic when comparing the availability of parents for home production. Despite their high employment rates, two-parent families do not seem to be sacrificing time for home production. Among two-parent families, about two-thirds (66 percent) have a parent available for home production at least part time, a share very similar to that found among single-parent families (70 percent). Surprisingly, only 40 percent of two-parent families have a parent who is not employed and is therefore available to care full-time for the child, compared with 57 percent of single-parent families. Parental employment varies a great deal with the age of the youngest child, with parents being more likely to work as children get older. Although the survey sample includes only families with children 3 years

Child Development Resources

27

table 2.1. Characteristics Influencing Both Financial and Time Resources Variable

Category

(%)

Marital status of parents

Married/living as married No longer married Never married Full-time Part-time Not employed Full-time/full-time Full-time/part-time Full-time/not emp Not emp/not emp Other combinations 1 2 3 4 or more

86 4 10 30 12 57 34 20 40 3 2 35 36 18 11

Single-parent households Employment status of parent Two-parent households Employment status of parents

Number of children under 18 in household

old or younger, the data exhibit this pattern. For example, among single parents, nearly two-thirds (65 percent) of parents with children less than 18 months old do not work; after a child’s second birthday, however, less than half (48 percent) of single parents are not working. Similarly, in twoparent families, at least one parent is at home in 44 percent of households where the child is less than 6 months old; for families with a child above age 2, the share is 34 percent. To summarize, children in single-parent and two-parent families are about equally likely to have a parent who engages in home production at least part time. Children in two-parent families are more likely to have a parent who spends time working, making purchased resources available from earnings. Another critical facet of family resources for child development is the number of individuals in the family. The fewer children in a family, the further the available money and time will go. Economic models of investment in children formalize this notion by specifying a multiplicative relationship between the number of children and the amount invested in each child (Becker 1991). The number of children is referred to as child “quantity,” or n, and the amount invested in each child as child “quality,” or q. The amount the family allocates to child investment, I, is then related to the quantity and quality of children as I = nq. This simple relationship implies that for any level of investment in children, as the number of children (n) rises, the amount invested in each child (q) must fall.1 Approximately one-third (35 percent) of parents in the survey have only one child under age 18 in the household, and another third (36 percent) 1 This exposition assumes a unitary price of a unit of quality in Becker’s model.

Kilburn, Wolfe

28 table 2.2. Allocation of Purchased Resources Variable Trimester of pregnancy in which mother first visited health professional Total annual family income

Category

First trimester Second trimester Third trimester/never <$10,000 $10,000–$19,999 $20,000–$29,999 $30,000–$39,999 $40,000–$59,999 $60,000+ Parents who have had “some trouble” Prenatal care or “lot of trouble” paying for each Medical expenses for birth of following types of expenses Child’s health and medical expenses Child care Supplies: formula, food, diapers, etc. Type of health insurance that covers Through employer child Government Some other source Uninsured

(%) 94 5 1 11 13 16 14 23 23 14 17 18 17 24 62 25 7 6

have two children (Table 2.1). Of the remaining third, 18 percent have three children, and 11 percent have four or more. The median number of children in U.S. families is now one child, after falling for decades (U.S. Bureau of the Census 1997). The survey does not include childless families, and is more likely to capture families with a large number of children; as a result, the median number of children in the survey families is two. Purchased Resources. Beginning with inputs during the prenatal period, the overwhelming majority of survey families (94 percent) received care from a health professional beginning in the first trimester of pregnancy (Table 2.2). Prenatal care is believed to improve birth outcomes by identifying high-risk pregnancies that may require special care and by helping mothers modify unhealthy behaviors (Frick and Lantz 1996, Lantz and Partin 1997).2 The rate of receiving first trimester prenatal care is unusually high among survey families: Among all registered births in the United States in 1996, about 82 percent of mothers obtained prenatal care during the first trimester, while 4 percent received care in the third trimester or no care at all (U.S. Bureau of the Census 1999). This difference could reflect the fact that the national statistics are drawn from health care providers’ 2 Lantz and Partin (1997) note that due to difficulties inherent in studying the efficacy of prenatal care, self-selection issues cloud the interpretation of results from many studies.

Child Development Resources

29

records, while the survey data is based on parents’ own reports, which may overstate actual use of prenatal care. Another finding regarding purchased resources is the incidence of parents who report “some trouble” or “a lot of trouble” paying for basic expenses associated with early child rearing. Approximately one in seven parents had trouble paying for expenses associated with pregnancy, and about one in six had problems paying medical expenses associated with the child’s birth. Similar numbers of parents report trouble paying for health expenses and child care as the child got older. Nearly a quarter of parents (24 percent) have had trouble paying for basic child-rearing supplies, such as formula, food, diapers, and other items. These problems are closely linked to families’ eligibility for public assistance programs. In results that presage findings for children’s health insurance, we find that families with income below $20,000 per year are less likely to have trouble paying for prenatal care and birth expenses than families earning between $20,000 and $40,000 per year. Nearly three-quarters of families earning less than $20,000 said they do not have to pay for these services, which probably indicates their participation in public programs. More families in the lower income group indicated that they have had trouble paying for supplies such as diapers and food: 41 percent of families earning under $20,000 per year had “some trouble” or “a lot of trouble” paying for supplies, compared with 23 percent of parents in the rest of the sample. The high percentage of families with higher incomes who report having had trouble paying for various basic child-rearing expenses, however, suggests that parents are regularly making tough spending choices. Family income is the most prevalent measure of families’ financial resources. Studies have repeatedly established an association between poverty and negative outcomes for children (see Duncan and Brooks-Gunn 1997). There is no consensus, however, as to whether the relationship is causal or whether poverty is simply associated with other factors that affect children’s well being (Mayer 1997a, 1997b; Brooks-Gunn et al. 1997). Throughout the 1980s and 1990s, children have continued to be more likely to live in poverty than other groups in the population (Hernandez 1993; Brooks-Gunn et al. 1997). During the years when children are young and parents elect to trade time working for time to devote to child rearing, the relative deprivation of many families is startling. In 1995, about one in five children under the age of 18 lived in poverty, compared with about one in seven people in the entire population (U.S. Bureau of the Census 1997). The rates tend to be even higher for children under five years of age (Hernandez 1993). The child poverty rate is routinely higher in the United States than in other Western industrialized countries (Smeeding and Rainwater 1995). The distribution of family income as reported for survey respondents (Table 2.2) is similar to the distribution of nationally representative data for similar families. In 1995, the poverty threshold was $10,504 for a family

30

Kilburn, Wolfe

with one adult and one child and $12,267 for a family with two adults and one child (U.S. Bureau of the Census 1996). In the Commonwealth Survey, which asked parents to place their own income within a set of income categories, 11 percent of families say they earn less than $10,000 per year, and another 13 percent earn between $10,000 and $19,999. Many of these families have only one child and one parent, yet nearly three-quarters of survey families with annual income below $10,000 have two or more children (72 percent). As in other data, the incidence of low income is strongly related to there being only one parent, usually the mother (Hernandez 1997). Families headed by a single parent are more likely to have a low income: Nearly a third of single-parent families (32 percent) have an annual income of less than $10,000, compared with less than 6 percent of two-parent families. Not surprisingly, family income is also linked to parents’ employment status: families with at least one parent working full time have much higher incomes on average. Among one-parent households, about a quarter (24 percent) of those with a parent who works full time earn less than $10,000 per year, while only 5 percent of two-parent households with at least one full-time worker earn so little. In contrast, nearly half (46 percent) of one-parent families where the parent works part time or is not employed earn less than $10,000 annually, and nearly the same percentage (44 percent) of two-parent families in which no adult is working full time have earnings in this range. Motivated by the recognition that official poverty standards are set extremely low (Brooks-Gunn et al. 1997; Citro and Michael 1995) and renewed emphasis on employment for recipients of public assistance, analysts have increasingly examined the situation of the “working poor.” In general, studies find that parental employment does not guarantee that a family will not be poor, although parental employment reduces the likelihood that the family will enter poverty and is associated with other positive family characteristics (Zill and Nord 1994). For example, Wertheimer (1999) uses the following definition of “working poor” – families with children that have incomes below the federal poverty standard and in which two parents work a total of at least 35 hours per week or a single parent works 20 hours per week. He found that over half (52 percent) of children in twoparent poor families had working parents according to this definition, and 30 percent of children in single-parent poor families had a working parent. There is one notable exception to the generalization that having an employed parent generates benefits for children. This exception is children’s health insurance coverage. Families in the lowest income group (less than $10,000 annually) do not have the lowest rates for children’s health insurance (Table 2.3). Among survey parents in the lowest income category, 10 percent have uninsured children, while 77 percent receive insurance for the child from the government. This reflects the use of Medicaid, the

Child Development Resources

31

table 2.3. Child’s Health Insurance by Provider and Income Category Insurance provider (%) Income category

Parent’s employer

Government

Other

Uninsured

< $10,000 $10,000–$19,999 $20,000–$29,999 $30,000–$39,000 $40,000–$59,999 $60,000+

6.5 25.6 53.3 75.3 84.3 84.6

77.1 48.1 27.6 11.6 5.8 8.8

6.0 8.9 7.8 5.1 7.5 6.4

10.0 16.9 11.0 6.8 1.6 0.2

joint federal-state program that provides coverage for children living in families with incomes below the poverty line. The income category with the highest rate of uninsured children is the second lowest category ($10,000–$19,999), with 17 percent of parents having uninsured children. This finding is consistent with government reports and the results of other studies. For example, Dubay and Kenney (1996) find that children in families with incomes between 100 and 133 percent of the poverty level are more likely to be uninsured than children living in poverty. In the third lowest category ($20,000–$29,999), the rate of uninsured children is 11 percent, close to that of the lowest income group.3 These patterns suggest that the government has been successful at delivering health insurance to children in the neediest families. In families earning $30,000 per year or more, fewer than 7 percent in each category have uninsured children. In the three higher income groups, employer-provided insurance covers children more than three-quarters of the time. It is the parents in between whose children have neither employer-provided insurance nor government insurance, and their children are more likely than others to be uninsured or covered by some other type of insurance. Another pattern worth noting is the low rate of children’s health insurance among Hispanics. While 6 percent of all surveyed parents say their child is uninsured, the rate for Hispanic parents is 14 percent, more than twice as high. White parents’ children are uninsured at the same rate as the entire sample, while black parents’ children are uninsured at half the rate (3 percent). Time Resources. Another set of inputs into child development is generated through home production – that is, using parents’ time resources. 3 The percentage of families in the $10,000–$19,999 income category with uninsured children is significantly different from the percentage of families in the lowest income category with uninsured children at a level of p = 0.05. The percentage of families in the $10,000–$19,999 income category with uninsured children is also significantly different from the percentage of families in the next highest income category at a level of p = 0.08.

Kilburn, Wolfe

32 table 2.4. Allocation of Time Resources Variable

Category

(%)

Length of time child was breastfed

More than 1 month Less than 1 month Not at all Read or look at book with child Play with child Sing/play music Hug child Right amount of time Lot more time Little more time Little less time Don’t know

53 13 33 39

Percent of parents who do each of following with child at least once a day or more often

Percent of parents who feel they spend about right amount of time with child, or would like to spend more or less time with child

84 59 89 37 38 19 4 2

The first of these inputs is breastfeeding. Compared with formula feeding, breastfeeding has been found to reduce feeding problems, constipation, and infections among babies. More recently, breastfeeding has also been associated with better cognitive outcomes for young children (Anderson et al. 1999). Research has demonstrated that maternal employment and breastfeeding are not always compatible, another example of trade-offs between financial and time resources. Results in Roe et al. (1999) indicate that employment characteristics such as maternal leave policies influence the duration of breastfeeding. In the Commonwealth Survey sample, about half of children were breastfed for more than one month (53 percent), 13 percent were breastfed for less than one month, and a third were not breastfed. Breastfeeding among survey families does not appear to be strongly related to the labor force participation of the parents. Another survey question inquired about the frequency with which parents engage in certain activities that may promote young children’s development: reading to the child, playing with the child, singing or playing music with the child, and hugging the child (Table 2.4). Although these activities may have specific developmental benefits for children (for example, reading books develops pre-reading skills) (Phillips and Love 1997), many types of parent-child interaction contribute to children’s emerging social and emotional competencies, as well as cognitive and linguistic skills (Beckwith 1990). Most parents reported playing with the child (84 percent) and hugging the child (89 percent) at least once a day, or more often. Over a third said they read or look at a book with the child (39 percent), and over half indicated that they sing or play music for the child (59 percent). The frequency

Child Development Resources

33

of some of these activities varies with a child’s age. For example, less than a quarter of parents with children under 1 year old say they read to their child, although the share is closer to half among parents of 2-year-olds. The survey also asked parents whether they feel they spend about the right amount of time with the child. More than a third of parents (37 percent) say they spend about the right amount of time with the child, and only 4 percent report that they would like to spend a little less time with the child. The majority of parents indicate that they would like to spend more time with the child: 38 percent want to spend a lot more time, and 19 percent a little more time. Responses to this question are related to parents’ labor force status. Among single parents, full-time workers are nearly twice as likely to say they would like to spend more time with the child (59 percent). Parents in two-parent households are also more likely to respond this way when both parents are in the workforce full time. To summarize some of the key patterns of purchased and time resources, the data show that in two-parent families, one parent usually spends time engaging in home production and the other parent spends time in the labor force procuring financial resources. In one-parent families with less parental time to “spend,” nearly all single parents spend a significant amount of time in home production, which implies that these families are much less likely to have a parent spending time in the labor force. This pattern of time use in single-parent families may reflect the nature of public assistance available to parents. Government resources to support child development are typically in the form of financial assistance, such as the Temporary Assistance to Needy Family (TANF) program or health insurance for children. Government support is much less likely to take the form of time assistance, although child care is increasingly a component of benefits provided by government. Hence, single parents may choose to allocate their time to home production rather than working because there is an alternative source of financial resources but not of time resources. The data also show that families with the lowest incomes are not the most likely to be without health insurance for their children or to have trouble affording prenatal or birth care, probably because they are eligible for public programs that provide for those needs. In contrast, basic supplies such as diapers, formula, and food are less often provided by government for needy families. The lowest income groups report having the most trouble paying for those items. Subsequent to the collection of the Commonwealth Survey data, welfare reform substituted TANF for Aid to Families with Dependent Children (AFDC). Welfare reform increased incentives to work, established time limits on the receipt of cash assistance, and set aside additional funds to subsidize child care. Given TANF’s emphasis on maternal employment, it is likely that low-income families are facing greater pressures to work, and so spend less time in home production. In addition, cash assistance

34

Kilburn, Wolfe

is now more closely tied to work effort through the Earned Income Tax Credit (EITC). Parental Inputs: Psychological and Human Capital Resources The most unique aspect of the Commonwealth Survey data is the unsurpassed information they provide on parents’ psychological and human capital resources. These resources have received less attention in other research and data collection on child-development despite their increasing prominence in the child-development modeling discourse. Characteristics related to psychological resources – parents’ beliefs, communication style, methods of discipline, warmth toward the child, and other parenting behaviors – have been demonstrated to be related to children’s well-being (Beckwith 1990; Holmbeck et al. 1995; Bornstein 1995; Maccoby and Martin 1983). Human capital resources include the skills, knowledge, and information parents bring to parenting, such as knowing how to treat childhood illnesses, familiarity with children’s nutritional requirements, and understanding whom to consult when a question arises. This category may also include parents’ overall education level, which has repeatedly been shown to be related to children’s outcomes and parents’ use of public services (Haveman and Wolfe 1995; Strauss and Thomas 1995; Sastry 1998). Psychological Resources. The Commonwealth Survey produced information about parents’ psychological resources in a series of questions that asked parents to assess their parenting competency. Despite the formidable challenges of child rearing, a majority of parents (67 percent) say they felt “very confident” when they first brought home their newborns, and an additional 25 percent felt “somewhat confident” (Table 2.5). Forty-four percent of parents of first-born children say they were “very confident” with their newborns, while 83 percent of other parents were “very confident.” Very few parents in the sample felt “only a little confident” or “not confident at all.” Even among first time parents, only 14 percent report little or no confidence. Other survey findings also reflect parents’ generally positive feelings about parenting. First, over half of parents (56 percent) responded that they are coping “very well” with the demands of parenthood (56 percent), while almost the entire remaining sample (43 percent) report that they are “about average.” In other words, almost no parents in the survey feel that they are coping less well than the average parent. Second, 68 percent of parents report that things in life overall have gotten better since they brought the child home. An additional 28 percent indicate that things have stayed about the same, and only 2 percent say things have gotten worse.

Child Development Resources

35

table 2.5. Psychological Resources Survey question

Response

(%)

How confident did you feel about caring for child when you first brought it home?

Very confident Somewhat confident Only a little confident Not confident at all Don’t know Very well About average Not too well Not well at all Don’t know Gotten better Gotten worse About same Some ways better/ some ways worsea Don’t know Depressed Enjoyed life Had crying spells Felt sad Felt people dislike me Excellent Good Only fair Poor Never Once Two to three times Four or more times Don’t know

67 25 6 1 1 56 43 0 0 1 68 2 28 1

In general, how well do you feel you are coping with demands of parenthood?

Since you brought child home, have things in your life overall gotten better, worse, or stayed same?

Have you felt or behaved in following ways some of time or most of time recently?

In general, is your mental health excellent, good, fair or poor?

In typical day, how many times do you feel frustrated or aggravated with child?

a Respondents

1 18 95 6 21 10 65 32 3 0 31 42 22 4 1

volunteered this response.

Third, when asked about their own feelings and behavior, nearly all parents (95 percent) say they have “enjoyed life” some or most of the time recently. Even so, a non-trivial number also report recent feelings that are not positive some or most of the time: 18 percent have felt depressed, 6 percent have had crying spells, 21 percent have felt sad, and 10 percent have felt that people dislike them. Finally, about two-thirds of parents report that their mental health is excellent, with most of the remainder saying that their mental health is good. Only 3 percent of parents characterize their mental health as “fair.” The survey also asked how many times a day on average the parent became frustrated with the child. Nearly a third of parents responded

36

Kilburn, Wolfe

“never” (31 percent), with the most common response being “once” (42 percent), and 22 percent saying “two to three times.” Only 4 percent of parents reported feeling frustrated with the child four or more times a day. In general, while some parents report feeling depressed or frustrated, nearly all have positive feelings about parenting. Human Capital Resources. Another path-breaking line of inquiry in the survey covers the human capital resources that parents bring to their childrearing responsibilities. The survey includes questions about parents’ education, a traditional measure of human capital resources. In addition, it includes several questions on parents’ sources of information caring for a young child. Several models postulate reasons for the relationship between parents’ education and children’s outcomes. According to one model, parent’s education is a proxy for the quality of interaction between the parent and child – for example, more-educated parents read more to their children than less educated parents (Timmer et al. 1985). A second model stresses values instilled by more-educated parents, who are thought to encourage their children to pursue more education, and also foster characteristics, such as independence and self-direction, that are rewarded in school and the labor market (Hernandez 1993). A third model stresses that educated parents are likely to earn higher income, which permits them to invest more in their children (Mayer 1997b). Over half the parents surveyed have education beyond high school (Table 2.6). About a quarter (27 percent) of all parents have been to trade school or some college, and another quarter (26 percent) have finished college or engaged in some postgraduate studies. A third of parents say they have a high school degree or GED, and 14 percent have not graduated from high school. This distribution of educational attainment is remarkably close to the national distribution of education for all adults ages 25–34 in 1996 (U.S. Bureau of the Census 1997). The Commonwealth Survey also provides information about how parents gain knowledge about child rearing. The data indicate that parents get information from a number of different sources, including health professionals, classes, written materials, and other media, such as videos or television. Only 4 percent of parents had not received information from at least one of the following sources: hospital or birthing center, childbirth classes, class or discussion about parenting, or books, magazines, newspapers, television, or videos. Over half (57 percent) had received information from at least three of those four sources. Despite the fact that the majority of parents appear to have received information from a number of sources, large numbers of parents also report that they could use more information on particular topics (Table 2.6). Most prevalent are “how to encourage the child to learn” (54 percent), although

Child Development Resources

37

table 2.6. Human Capital Resources Survey question

Response

(%)

What is parent’s highest level of education?

Did not graduate from high school High school or GED Trade school or some college College or postgrad Yes No Yes No Yes No Yes/sometimes No

14

How to discipline child How to encourage child to learn How to deal with child’s sleeping patterns How to care for newborn What to do when child cries How and when to toilet train child

42 54

Did you receive information in hospital/ birthing center about caring for newborn? Have you ever attended any childbirth classes? Have you ever attended class or discussion about parenting? Do you ever use books, magazines, newspapers, television or videos to get information about raising children? Could you use more information on following topics?

33 27 26 88 12 71 29 35 65 74 26

30 19 23 41

“how to discipline the child” (42 percent) and “how and when to toilet train the child” (41 percent) also elicit interest. In sum, most families appear comfortable with the human capital resources available to them for child rearing. Government Input Although parents provide the majority of input into child development before children turn 3, the public sector does offer some resources – primarily financial – to families with young children. Public spending on older children is typically universal, largely because public education accounts for the bulk of this spending. In contrast, public programs for younger children tend to be social safety net programs targeted toward the most needy families. These programs include food programs, housing assistance, social services, health insurance and care, legal assistance, and income maintenance programs (see Haveman and Wolfe 1995). Child care

Kilburn, Wolfe

38

table 2.7. Parents’ Use of Government Resources for Child Development Survey question

Response

(%)

At any time since child’s birth, have you received assistance from any government assistance program? What type of health insurance covers your child?

Yes No

47 53

Government Private employer Some other source Uninsured

25 62 7 6

subsidies are a growing component of the social safety net (Greenberg 1999). Although government inputs into early childhood development are largely conditional on a family being needy or at-risk, nearly half (47 percent) of families surveyed said they had received government assistance since the child’s birth (Table 2.7). On the face of it, this incidence seems high, given that the poverty rate for children in the United States is about one in five and nearly half of families in the survey had incomes greater than $40,000 per year. Yet several factors may explain this seeming incongruence. Parents who participated in the survey had children as old as 35 months. Since the question asks whether the family received assistance from the government at any time since the child’s birth, responses reflect the number of families receiving government assistance over a period of years, not the number receiving assistance at the time of the survey. In fact, parents with older children are more likely to respond affirmatively to this question. For similar reasons, the current poverty rate is likely to understate the number of families that ever receive public assistance over the first three years of a child’s life. Moreover, the poverty rate for families with very young children is higher than the rate for all families with children under 18 (Zill and Nord 1994). Finally, the families’ responses to the income question reflect contemporaneous income levels, not income levels in the past, when they might have received public support. The government also provides resources for early childhood development through child health insurance. At the time of the survey, a quarter of parents had children who were covered by government-provided health insurance, and one-sixth were themselves covered under government health insurance programs. A comparison of responses regarding assistance from any government program and assistance in the form of health insurance indicates that the majority of government support is received in the form of health care coverage. Of children whose families had received government assistance at some time, 60 percent were currently covered by government insurance or their parents were covered. Hence the

Child Development Resources

39

table 2.8. Other Sources of Support for Families with Young Children Survey question

Response

(%)

Do any of child’s grandparents live within an hour’s drive of your home? How many relatives or friends do you have living nearby whom you could count on for help in emergency?

Yes No Many Few None

78 22 70 27 3

other 40 percent either were not receiving government health insurance at the time of the survey or had received assistance other than health insurance through programs such as Food Stamps or WIC (the Women, Infants, and Children program). Other Sources of Support Several studies emphasizing the value of social capital have identified links between extended family and community inputs and children’s outcomes (see, for example, Runyan et al. 1998). The Commonwealth Survey examines some of those inputs through questions regarding the contributions of kin, friends, neighbors, and physicians, in addition to government resources, in early child development. In general, the survey data indicate that families have a number of sources outside family and government to which they can turn for support (Table 2.8). Over three-quarters of families (78 percent) live within a one-hour drive of at least one of the child’s grandparents. A similar share of families (70 percent) have many relatives and friends nearby on whom they can count in an emergency.

risk factors and the allocation of resources for child development Most models of child development assume that children do not enter various stages of development with the same chances of success (Brooks-Gunn et al. 1997). Children may start with different outcomes from earlier developmental processes, receive different inputs from family or nonparental sources, or have unequal access to resources of various types. These differences, often called risk factors, are predictors of which children are at risk for poor outcomes. The association between risk factors and poor outcomes is not necessarily causal, but children who experience multiple risk factors have a much higher likelihood of realizing worse outcomes than other children (Sameroff et al. 1987). Evidence suggests that multiple risk factors operate in a multiplicative rather than additive fashion (Werner and Smith 1982; Garmezy and

40

Kilburn, Wolfe

Tellegen 1984). For example, in the classic study by Rutter et al. (1975), children exhibiting one of four risk factors experienced outcomes no worse than children with no risk factors, while children exhibiting any two risk factors were four times as likely to have a psychological disorder. Thus, although the types of risk factors measured by the survey have been found to raise the likelihood that a child will experience poor outcomes, a relatively small number of children identified as being at risk will actually experience poor outcomes (Kirby et al. 1993; Finkelstein and Ramey 1980; Rutter 1985). This observation has led to a growing emphasis on the study of children’s resilience – that is, positive factors that protect children from developing poor outcomes despite exposure to risk – rather than their vulnerability (Werner 1990; Garmezy and Tellegen 1984). Research on the combinations of risk factors that best identify children who will have developmental difficulties is an important area for additional research. At least three approaches are used for identifying children at risk (Meisels and Wasik 1990). The first approach relies on medical diagnosis to identify children likely to experience developmental delays because of medical conditions such as Down syndrome or spina bifida. A second approach focuses on biological factors – such as low birthweight or premature birth – that are likely to have developmental implications. The final approach uses environmental factors, including parental resources and other sources of investment in child development, to identify at-risk children. The Commonwealth Survey data contain some information that can be used to assess children’s risk for developmental delays on the basis of medical diagnosis or biological risk. Although the survey does not ask parents about specific medical diagnoses, one question asks for an assessment of the child’s overall health status. Only 2 percent of parents characterize their children as having “only fair” health, and less than 1 percent of parents say the child’s health is “poor.” The rest describe their children’s health as excellent or good. Similarly, one question in the survey provides evidence on biological risk. When asked whether the child was born prematurely or weighed less than 5 12 pounds at birth, only 10 percent of parents respond affirmatively. In contrast, the data include numerous variables that can be used to assess risk from an environmental perspective (Table 2.9). By examining some of those variables and relating them to the two available measures of medical diagnosis and biological risk, we are able to generate a summary measure of the allocation of some of the most important resources for child development. The characteristics included in the risk analysis have been documented in the literature as having strong associations with poor child outcomes (see, for example, Kirby et al. 1993; Karoly et al. 1998; Cramer 1995; Mayer 1997a). There is, however, no set of risk factors agreed on to be the most important, and other characteristics and groupings could have been devised.

Child Development Resources

41

table 2.9. Incidence of Risk Factors in Surveyed Families Variable

Category

(%)

Parent’s race/ethnicity

Latino Black <25 Did not graduate high school <$10,000 $10,000–$19,000 Single-parent household Did not receive care during first trimester Uninsured Parent rates own mental health fair or poor

13 11 25 14 11 13 14 6

Age of mother at child’s birth Parent’s education Family income Marital status Early prenatal care Child’s health insurance coverage Parental mental health

6 3

Our analysis of risk factors considers five demographic characteristics that have been shown in the literature to be related to child development outcomes: race or ethnicity (Cramer 1995), age of mother at child’s birth (Kirby et al. 1993; Maynard 1997), parent’s education (Hernandez 1993), family income (Mayer 1997a; Duncan and Brooks-Gunn 1997), and marital status or number of parents in the household. In the survey data, 11 percent of the families are black, and 13 percent are Latino, proportions very similar to those within the U.S. population (U.S. Bureau of the Census 1997). A quarter of mothers were less than 25 years old at the time of the child’s birth, with 8 percent being younger than 21. Only about 14 percent of parents have not graduated from high school, and another third have only a high school degree or GED. Almost a quarter of survey families (24 percent) report incomes below $20,000 per year, or the lowest two income categories in the data. One in seven families report being headed by a parent who was previously married, has never been married, or is not living in a marriage-like relationship. We also examine three risk factors related to family health and health care: early prenatal care, health insurance status, and self-reported parental mental health. Only 6 percent of mothers in the sample say they did not visit a health professional during the first trimester of pregnancy. Six percent of children lack health insurance, a variable found by some studies to be linked to well-care visits and treatment for acute health conditions (Cunningham and Hahn 1994; Wolfe 1999) and to health outcomes (Currie and Gruber 1996a, 1996b), although other evidence is somewhat mixed (see Devaney et al. 1997). Only 3 percent of respondents indicate that their mental health is either fair or poor, a factor that may be related to a parent’s interactions with the child, ability to hold steady employment, or capacity to procure appropriate inputs for the child’s development.

Kilburn, Wolfe

42

table 2.10. Number of Risk Factors in Survey Families Number of risk factors

Percentage of families

0 1 2 3 4 5 6

48.6 21.6 11.1 10.0 6.2 2.2 0.3

It is not surprising that such a large number of families in the sample display multiple risk factors, because a family with at least one risk factor is more likely to develop additional risk factors than a family with none (see Meisels and Wasik 1990; Sameroff et al. 1987). The survey data show that although no more than a quarter of families display any particular risk factor, just over half the families in the sample exhibit at least one risk factor, and more than 30 percent have more than one (Table 2.10). About 22 percent of families have a single risk factor, 11 percent have two, and nearly 19 percent have three or more. It is not surprising that such a large number of families in the sample display multiple risk factors, since having at least one risk factor raises the likelihood of displaying another. That is, a family with one risk factor is more likely than a family with no risk factor to have another risk factor. This phenomenon is illustrated in Figure 2.1, which shows that among families who already have one risk factor besides their ethnicity, blacks and Latinos are more likely than whites to exhibit additional risk factors. Thirty-seven percent of whites, 66 percent of blacks, and 72 percent of Latinos have one or more additional risk factors. Table 2.11 provides 70 60 Percent

50 White

40

Black Hispanic

30 20 10 0 0

1

2

3

4

5

Number of Additional Risk Factors

figure 2.1 Incidence of Additional Risk Factors, Given One Risk Factor by Race/ Ethnicity.

Black Latino Young mother No care in first trimester Child uninsured Has not graduated high school Low income Single parent

Risk factor

1.59 1.29 — — — — — —

Young mother 1.28 2.09 1.70 — — — — —

No health care in first trimester

table 2.11. Incidence of Risk Factors Relative to Entire Sample

0.42 2.17 1.29 3.39 — — — —

Child uninsured 1.30 3.25 2.07 2.83 2.60 — — —

Has not graduated high school 1.80 1.98 1.26 2.21 2.19 2.44 — —

Low income

3.24 1.15 2.41 1.89 1.03 2.28 2.52 —

Single parent

1.86 2.65 1.09 1.85 1.98 3.00 1.93 2.14

“Fair” mental health

44

Kilburn, Wolfe

another way of seeing this. In nearly every case, families with a risk factor shown in the far left column are much more likely than the entire sample to exhibit the risk factors listed along the top of the table. For example, families in which the mother is young are more than twice as likely as the general population to have a parent who did not graduate from high school. Despite the variety of life chances into which children are born, the Commonwealth Survey data show that the majority of children are in excellent or good health (76 percent and 22 percent, respectively). Nevertheless, the data confirm previous research that finds a relationship between the rare cases of negative health outcomes and certain risk factors. In these data, 58 percent of families with premature or low birthweight babies exhibit at least one risk factor, compared with 49 percent of other families.4 The incidence of risk factors also differs according to whether the parent rated the child’s health as “excellent” or “good.” Among families rating their child’s health as good, 59 percent exhibited at least one risk factor, compared with 49 percent of families who rated their child’s health as excellent.5

conclusion Resources are allocated to child development by parents, the government, and others, and play a prominent role in most models of child development across the spectrum of academic disciplines. The Survey of Parents with Young Children includes data on young children and their families that reveal some of the fundamental patterns of resource allocation to child development. The remaining chapters in this book examine the role of resources in analyzing various aspects of child development. In the United States, parents provide the majority of inputs into child development, including direct resources and time. This chapter has examined four types of resources that parents devote to young children’s development: purchased, time, psychological, and human capital resources. We argue that purchased and time resources are partial substitutes for one another, due to parents’ labor force behaviors that limit both the amount of purchased resources the family can provide and the amount of time inputs the family can make to child rearing. Whether the family includes one or two resident parents is also critical to purchased and time resource allocation because the number of resident parents largely determines the total amount of time parents can divide between working and child rearing. The data show that about a quarter of families have income considered meager by federal poverty standards. Many of these are families with a single resident parent. 4 These two percentages are significantly different at the level p = 0.06. 5 These percentages are significantly different at the level p = 0.003.

Child Development Resources

45

According to the survey data, government’s role in child development is primarily to provide financial inputs to families with limited incomes. For example, government generally provides health insurance for children in families with incomes below the federal poverty line, and nearly half (47 percent) of families in the survey reported having used the “safety net” at some point. Since government programs primarily target the most disadvantaged families – that is, families in the lowest income quartile – we find that the families in the second-lowest income quartile may have the most trouble providing basic inputs into early child development. These parents, not those with the lowest incomes, are most likely to report that their children lack health insurance or that they have had problems paying for basic child-rearing expenses such as health services, formula, food, and diapers. Few government programs make up shortfalls in families’ time inputs. This may contribute to the choices single parents make between working and devoting time to child rearing. The survey data show that in twoparent families, there is usually one parent who uses time to engage in home production and one who spends time in the labor force procuring financial resources. One-parent families have less parental time to “spend.” The data suggest that nearly all single parents spend a significant amount of time in home production, yet one-parent families are much less likely to have a parent spending time in the labor force in addition to spending time in home production. The fact that single parents allocate their scarce time to home production rather than working is not surprising, given that government provides an alternative source of financial resources but not time resources. To assess the relative incidence of resource levels that might compromise child development, we examined the frequency with which survey families exhibited eight environmental risk factors that the literature has shown to be related to poor child outcomes. These risk factors span the four types of parental resources. More than half the families in the sample displayed at least one risk factor, with about 30 percent displaying two or more risk factors. Furthermore, the risk factors in these data were associated with poor child outcomes, such as premature birth, low birthweight, or poor child health. Nevertheless, while the incidence of risk factors among the families was high, nearly all children in these data are healthy: 97 percent of families report that their children are in good or excellent health. This is consistent with other research that has found that although risk factors raise the likelihood that a family experiences a negative outcome, there still may be a relatively low frequency of negative outcomes, even among families at higher risk. Overall, and despite the relative financial deprivation of many families and the large fraction that exhibit environmental risk factors, the picture that emerges regarding psychological and human capital resources

46

Kilburn, Wolfe

for young children’s development is generally quite positive. Only 4 percent of parents in the Commonwealth Survey sample had not received information about child development from the sources discussed in the survey, and the overwhelming majority were confident about caring for a baby when they first brought the child home, including first-time parents. Most parents feel that their lives improved overall after bringing their child home from the hospital, and no parents believe their abilities to cope with the demands of parenting are below average. Although parents in the survey face many challenges, most express satisfaction with the level of psychological and human capital resources they devote to child rearing.

references Anderson, J.W., Johnstone, B.M., & Remley, D.T. (1999). Breast-feeding and cognitive development: A meta-analysis. American of Journal of Clinical Nutrition, 70(4), 525–35. Barnard, K.E., & Kelly, J.F. (1990). Assessment of parent-child interaction. In Meisels, S.J., & Shonkoff, J.P. (eds.), Handbook of Early Childhood Intervention. Cambridge, UK: Cambridge University Press. Bauman, P.S. & Levine, S.A. (1986). The development of children of drug addicts. International Journal of the Addictions, 21(8), 849–63. Becker, G.S. (1976). The economic approach to human behavior. Chicago, IL: University of Chicago Press. Becker, G.S. (1991). A treatise on the family, enlarged edition. Cambridge, MA: Harvard University Press. Becker, G.S. & Michael, R.T. (1976). On the new theory of consumer behavior. In Becker, G.S. (ed.), The economic approach to human behavior. Chicago, IL: University of Chicago Press. Becker, G.S., & Tomes, N. (1986). Human capital and the rise and fall of families. Journal of Labor Economics, 4(2), S1–S139. Beckwith, L. (1976). Caregiver-infant interaction and the development of the high risk infant. In Tjossem, T.D. (ed.), Intervention strategies for high risk infants and young children, (pp. 119–140). Baltimore, MD: University Park Press. Beckwith, L. (1990). Adaptive and maladaptive parenting–Implications for intervention. In Meisels, S.J., & Shonkoff, J.P. (eds.), Handbook of early childhood intervention. Cambridge, UK: Cambridge University Press. Benton Foundation. (1998). Effective language for discussing early childhood education and policy. New York, NY: Benton Foundation. Blank, R.M. (1997). It takes a nation: A new agenda for fighting poverty. New York, NY: Russell Sage Foundation. Bornstein, M. (1995). Handbook of parenting. Hillsdale, NJ: Erlbaum Associates. Bronfenbrenner, U. (1979). The ecology of human development: Experiments by nature and design. Cambridge, MA: Harvard University Press. Bronfenbrenner, U., & Ceci, S.J. (1994). Nature-nurture reconceptualized in developmental perspective: A bioecological model. Psychological Review, 101(4), 568–86.

Child Development Resources

47

Brooks-Gunn, J., Brown, B., Duncan, G.J., & Moore, K.A. (1995). Child development in the context of family and community resources: An agenda for national data collection. In National Research Council and Institute of Medicine, Integrating federal statistics on children, report of a workshop. Washington, DC: National Academy Press. Brooks-Gunn, J., Duncan, G.J., & Maritato, N. (1997). Poor families, poor outcomes: The well-being of children and youth. In Duncan, G.J., & Brooks-Gunn, J. (eds.), Consequences of growing up poor. New York, NY: Russell Sage Foundation. Citro, C.F., & Michael, R.T. (1995). Measuring poverty: A new approach. Washington, DC: National Academy Press. Coleman, J.S. (1988). Social capital in the creation of human capital. American Journal of Sociology, 94, 95–120. Cramer, J.C. (1995). Racial and ethnic differences in birthweight: The role of income and financial assistance. Demography, 32(2), 231–47. Cunningham, P.J., & Hahn, B.A. (1994). The changing American family: Implications for children’s health insurance coverage and the use of ambulatory care services. The Future of Children, 4(3), 24–42. Currie, J., & Gruber, J. (1996a). Health insurance, eligibility, utilization of medical care, and child health. Quarterly Journal of Economics, 111(2), 431–66. Currie, J., & Gruber, J. (1996b). Saving babies: The efficacy and cost of recent changes in the medicaid eligibility of pregnant women. Journal of Political Economy, 104(6), 1263–96. Devaney, B.L., Ellwood, M.R., & Love, J.M. (1997). Programs that mitigate the effects of poverty on children. The Future of Children, 7(2), 88–112. Dubay, L.C., & Kenney, G.M. (1996). The effects of medicaid expansions on insurance coverage. The Future of Children, (6)1, 152–161. Duncan, G.J., & Brooks-Gunn, J. (eds.) (1997). Consequences of Growing Up Poor. New York, NY: Russell Sage Foundation. Finkelstein, N.W., & Ramey, C.T. (1980). Information from birth certificates as a risk index or educational handicap. American Journal of Mental Deficit, 84, 546–52. Frick, K.D., & Lantz, P.M. (1996). Selection bias in prenatal care utilization: An interdisciplinary framework and review of the literature. Medical Care Research and Review, 53(4), 371–96. Garbarino, J. (1990). The human ecology of early risk. In Meisels, S.J., & Shonkoff, J.P. (eds.), Handbook of early childhood intervention. Cambridge UK: Cambridge University Press. Garmezy, N., & Tellegen, A. (1984). Studies of stress resistant children: Methods, variables, and preliminary findings. In Morrison, F.J., Lord, C., & Keating, D.P. (eds.), Applied developmental psychology, vol. 1. Orlando, Florida: Academic Press. Greenberg, M. (1999). Child Care Policy Two Years Later. Center for Law and Social Policy. http://www.clasp.org/pubs/childcare/childcarepolicyarticlemhg.html. Haveman, R., & Wolfe, B. (1995). The determinants of children’s attainments: A review of methods and findings. Journal of Economic Literature, 33(4), 1829–78. Hernandez, D.J. (1993). America’s children: Resources from family, government and the economy. New York, NY: The Russell Sage Foundation. Hernandez, D.J. (1997). Poverty trends. In Duncan, G.J., & Brooks-Gunn, J. (eds.), Consequences of growing up poor. New York, NY: Russell Sage Foundation.

48

Kilburn, Wolfe

Holmbeck, G.N., Paikoff, R.L., & Brooks-Gunn, J. (1995). Parenting of adolescents. In Bornstein, M. (ed.), Handbook of parenting. Hillsdale, NJ: Erlbaum. Karoly, L.A., Greenwood, P.W., Everingham, S.S., Hoube, J., Kilburn, M.R., Rydell, C.P., Sanders, M., & Chiesa, J. (1998). Investing in our children: What we know and don’t know about the costs and benefits of early childhood interventions. Santa Monica, CA: RAND, MR-898-TCWF. Kirby, R.S., Swanson, M.E., Kelleher, K.J., Bradley, R., & Casey, P. (1993). Identifying at-risk children for early intervention services: Lessons from the infant health and development program. The Journal of Pediatrics, 122 (5, Part 1), 680–86. Lantz, P., & Partin, M. (1997). Population indicators of prenatal and infant health. In Hauser, R.M., Brown, B.L.V., & Prosser, W.R. (eds.), Indicators of children’s wellbeing. New York, NY: Russell Sage Foundation. Leibowitz, A. (1974). Home investments in children. Journal of Political Economy, 82, S111–31. Maccoby, E.E., & Martin, J.A. (1983). Socialization in the context of the family: Parent-child interaction. In Hetherington, E.M. (ed.), Handbook of child psychology, vol. 4: Socialization, personality, and social development. New York, NY: Wiley. Mayer, S.E. (1997a). Indicators of children’s economic well-being and parental employment. In Hauser, R.M., Brown, B.L.V., & Prosser, W.R. (eds.), Indicators of children’s well-being. New York, NY: Russell Sage Foundation. Mayer, S.E. (1997b). What money can’t buy: Family income and children’s life chances. Cambridge, MA: Harvard University Press. Maynard, R. (1997). Kids having kids. New York, NY: Urban Institute. Meisels, S.J., & Wasik, B.A. (1990). Who should be served? Identifying children in need of early intervention. In Meisels, S.J., & Shonkoff, J.P. (eds.), Handbook of early childhood intervention. Cambridge UK: Cambridge University Press. Mulligan, C.B. (1997). Parental priorities and economic inequality. Chicago, IL: University of Chicago Press. Phillips, D.A., & Love, J.M. (1997). Indicators for school readiness, schooling, and child care in early to middle childhood. In Hauser, R.M., Brown, B.L.V., & Prosser, W.R. (eds.), Indicators of children’s well-being. New York, NY: Russell Sage Foundation. Roe, B., Whittington, L.A., Fine, S.B., & Teisl, M.F. (1999). Is there competition between breast–feeding and maternal employment? Demography, 36(2), 157–71. Runyan, D.K., Hunter, W.M., Socolar, R.R.S., Amaya-Jackson, L., English, D., Landsverk, J., Dubowitz, H., Browne, D.H., Bangdiwala, S.I., & Mathew, R.M. (1998). Children who prosper in unfavorable environments: The relationship to social capital. Pediatrics, 101(1), 12–18. Russell, K.S., Swanson, M.E., Kelleher, K.J., Bradley, R.H., & Casey, P.H. (1992). Identifying at-risk children for early intervention services: Lessons from the infant health and development program. Little Rock, AR: Center for Research on Teaching and Learning, University of Arkansas. Rutter, M. (1985). Psychopathology and development: Links between childhood and adult life. In Rutter, M., & Hersov, L. (eds.), Child and adolescent psychiatry: Modern approaches, Second Edition. Oxford UK: Oxford University Press.

Child Development Resources

49

Rutter, M., Yule, B., Quinton, D., Rowland, O., Yule, W., & Berger, M. (1975). Attainment and adjustment in two geographical areas: III. Some factors accounting for area differences. British Journal of Psychiatry, 126, 520–33. Sameroff, A.J., Siefer, R., Barocas, R., Zax, M., & Greenspan, S. (1987). Intelligence quotient scores of 4-year old children: Social-emotional risk factors. Pediatrics, 79, 343–50. Sastry, N.M. (1998). Community characteristics, individual and household attributes, and child survival in Brazil. Demography, 33(2), 21–230. Sher, K.J. (1991). Children of alcoholics: A critical appraisal of theory and research. Chicago, IL: University of Chicago Press. Smeeding, T., & Rainwater, L. (1995). Cross-national trends in income poverty and dependence: The evidence for young adults in the eighties. In McFate, K. (ed.), Poverty, inequality and the future of social policy. New York, NY: Russell Sage Foundation. Smith, J.R., Brooks-Gunn, J. & Jackson, A. (1997). Parental employment and children. In Hauser, R.M., Brown, B.L.V., & Prosser, W.R. (eds.), Indicators of children’s well-being. New York, NY: Russell Sage Foundation. Strauss, J., & Thomas, D. (1995). Human resources: Empirical modeling of household and family decisions. In Behrman, J., & Srinivasan, T.N. (eds.), Handbook of development economics, Vol. IIIA. Amsterdam, Netherlands: Elsevier. Timmer, S., Eccles, J., & O’Brien, K. (1985). How children use time. In Juster, F.T., & Stafford, F. (eds.), Time, goods and well-being. Ann Arbor, MI: Institute for Social Research. U.S. Bureau of the Census. (1996). Poverty in the United States: 1995, Current Population Reports, P60-194. Washington, DC. U.S. Bureau of the Census. (1997). Statistical Abstract of the United States: 1997, 117th Edition. Washington, DC. U.S. Bureau of the Census. (1999). Statistical Abstract of the United States: 1999, 119th Edition. Washington, DC. Werner, E.E. (1990). Protective factors and individual resilience. In Meisels, S.J., & Shonkoff, J.P. (eds.), Handbook of early childhood intervention. Cambridge UK: Cambridge University Press. Werner, E.E., & Smith, R.S. (1982). Vulnerable but invincible: A longitudinal study of resilient children and youth. New York, NY: McGraw Hill. Wertheimer, R. (1999). Working poor families with children. Washington, DC: Child Trends. Wolfe, B. (1999). Poverty, children’s health, and health care utilization. Federal Reserve Bank of New York Economic Policy Review, September, 1–13. Zill, N., & Nord, C.W. (1994). Running in place: How American families are faring in a changing economy and an individualistic society. Washington, DC: Child Trends, Inc.

3 Preparing for Parenthood: Who’s Ready, Who’s Not? Constance T. Gager, Sara S. McLanahan, and Dana A. Glei

The birth of a child, especially the first child, represents a major life transition for most parents. Parents’ ability to cope successfully with their new responsibilities has lasting consequences not only for the health and development of their new baby but also for their own well-being and relationship with one another. Despite the predictability of these new challenges, adults enter parenthood with different levels of skill and knowledge and with different social and psychological resources. To explore levels of “readiness” that new parents bring to their role, we examined data from the 1996 Commonwealth Fund Survey of Families with Young Children in four areas: whether or not the pregnancy was intended, parents’ receipt of formal instruction about childbirth or parenting, parents’ access to social support from family or friends, and the extent to which parents are prepared psychologically for the challenges of parenting. Our analysis addresses two basic questions: What proportion of new parents are ready for their new role in each of these domains and What risk factors are associated with being unprepared for parenthood? Although other researchers have examined these indicators separately, this study is the first to consider them concurrently using a nationally representative data set. Overall, the results are encouraging. The vast majority of new parents appear to be prepared for their responsibilities. Over 95 percent say their child was wanted, over 75 percent attended a parenting class, and 70 percent have friends or relatives to call on in an emergency. The only area in which new parents appear vulnerable is psychological skills. Only 40 percent of new parents report that they feel confident about their ability to parent and believe that they are coping well with the demands of parenthood. With respect to risk factors, our findings support the conventional wisdom in some areas, contradict existing knowledge in other areas, and provide new information in still other domains. Parents who are married 50

Preparing for Parenthood

51

and highly educated, for example, are more likely than other parents to plan their pregnancy and to attend childbirth classes, as we expected. Less predictably, African American and Hispanic parents are less likely than other parents to report having relatives or friends to count on in case of an emergency, a finding that is contrary to some (but not all) previous research. Finally, parents who were abused as children are less ready to parent than those who were not abused – a result that is not surprising but has never before been documented using a national representative sample of parents.

the significance of “readiness to parent” Current literature offers a range of perspectives on parental “readiness” and the potential importance of the four indicators covered by the survey. Pregnancy Intention Numerous studies support the hypothesis that intended pregnancies are associated with better parenting and better child outcomes (see Brown and Eisenberg 1995 for an extensive review). Parents who plan their pregnancies are more likely to be prepared for their new responsibilities and to have the social, economic, and psychological resources to care adequately for their child. They are less likely to be in the midst of other major life changes, such as a residential move, job change, or divorce, factors that may undermine their ability to parent. Previous researchers have distinguished among three types of pregnancies: those that are wanted and occur at the desired time, those that are wanted but occur too soon, and those that are unwanted at any time. Unintended births include pregnancies that fall into the second and third categories (Forrest 1994; Morgan 1996; Brown and Eisenberg 1995). Women with unintended pregnancies are less likely than women who planned their pregnancies to receive adequate prenatal care. A woman with an unwanted pregnancy is from 1.8 to 2.9 times more likely to delay seeking prenatal care until after the first trimester, while a woman with a mistimed pregnancy is from 1.4 to 2.6 times more likely to delay seeking care. Women with unintended pregnancies are also more likely to engage in “risky” behaviors, including smoking and drinking, which are often linked to low birthweight and other adverse birth outcomes. Research suggests that eliminating unwanted births would result in a 7 percent reduction in low birthweight births among African Americans and a 4 percent reduction among whites (Kendrick et al. 1990). Eliminating mistimed births, however, would not significantly reduce low birthweight births. The effects of pregnancy intentions extend into infancy and childhood. Baydar and Grady (1993) found that children whose conceptions

52

Gager, McLanahan, Glei

were either mistimed or unwanted had significantly fewer resources and exhibited more developmental delays than children of intended pregnancies. Even after controlling for socioeconomic differences among the families, younger children whose births were unintended had lower levels of positive affect and older children had lower verbal skills than children whose conceptions were planned. Baydar and Grady hypothesize that adverse child outcomes may be due to parents’ being less available to their child. Finally, some research suggests an association between unintended pregnancy and child abuse and neglect, although the evidence on this issue is mixed. Zuravin (1991) suggests that under conditions of severe financial and emotional deprivation, unplanned childbearing increases the risk of child abuse (but not neglect), in part by increasing family size. Sorting out the causal relationship between pregnancy intention and parenting behavior is not a simple task. Many variables that predict pregnancy intention also predict parenting ability and child well-being through other routes. Compared with parents whose pregnancies are unintended, parents who plan their pregnancies are more likely to be better educated, older, married, and more economically secure, all of which are associated with more effective parenting and more successful child development (Brown and Eisenberg 1995). Although most of the empirical research described here is based on correlational evidence, there is some quasi-experimental evidence that supports the claim that pregnancy intentions per se affect parenting and child outcomes. Researchers in Sweden, Finland, and Czechoslovakia followed women who were denied abortions, and compared their birth outcomes with those of women who did not seek abortions.1 In these studies, seeking an abortion was taken to indicate an unwanted pregnancy, and denial of an abortion was treated as a “natural experiment.” The researchers reported adverse outcomes for children whose mothers were denied abortions (Forssman and Thuwe 1981, 1966; Blomberg 1980; Myhrman 1988; Matejcek et al. 1978). The Czech study is especially important because researchers were able to compare siblings, thereby controlling for unobserved common family background factors. These comparisons indicated that girls whose pregnancies were unintended experienced more emotional problems than did their sisters whose pregnancies were wanted. Participation in Parent Education Classes The second indicator of “readiness to parent” is participation in childbirth and/or parenting classes. Although birthing and parenting classes 1 Women were denied abortions because their pregnancies had advanced beyond 12 weeks, because of a chronic disease, or because they had terminated a pregnancy in the previous 6 months.

Preparing for Parenthood

53

are believed to improve child outcomes, parenting education services – and especially childbirth classes – have received little attention. Childbirth classes provide parents with information about the birth process, appropriate nutrition, and techniques to help them prepare for the birth of the child and cope with the delivery. Parenting education often includes specific information on how parents can foster a child’s intellectual, social, and physical development, as well as strategies for solving problems that go along with having a new baby. These classes aim to improve parent-child interactions and child development, which in turn are believed to improve child outcomes. Like planning for pregnancy, attending childbirth or parenting classes may be a proxy for some other characteristic of the parent, such as motivation to be a “good” parent, associated with both class attendance and parenting quality. Thus, benefits associated with parenting education may be attributable to some characteristic of the parent rather than to the classes themselves. Fortunately, we have some experimental evidence on the effects of parenting education to increase our confidence in this indicator. Analyses of the long-term effects of programs such as classes for parents or home visits by trained professionals indicate that, in some instances, parenting education has increased the likelihood that parents will provide a stimulating home environment by spending time talking and reading to their children (Gomby et al. 1993). Other research suggests that the benefits of parenting education have been modest or mixed (see St. Pierre et al. 1993). In their review of home visiting programs, Olds and Kitzman (1993) argue that the most effective programs are those that are comprehensive in their goals, intensive (that is, providing many visits by well-trained staff), and targeted toward families that are initially at high risk. Recent evaluations of two interventions designed to increase the parenting skills of teenage mothers also provide some insight into the effects of parenting education programs on parenting and child well-being. Mothers in the New Chance and Teenage Parent Demonstration programs were randomly assigned to control and experimental groups, and the latter were offered a variety of educational, healthcare, and childcare services. Many of these services included parenting workshops or classes as part of multifaceted intervention strategies. Evaluations of both programs indicated that the interventions had no lasting effect on the mothers’ parenting skills (for a review, see Reichman and McLanahan 1997). Finally, there is some evidence that participation in parenting classes may increase parents’ support networks, which in turn may improve parenting practices (Webster-Stratton 1997). These findings have led researchers to advocate parenting classes to improve child outcomes and also to strengthen parents’ social support networks. A specific goal of home visiting programs is to provide emotional, social, and practical support to parents in need (Wasik et al. 1997).

54

Gager, McLanahan, Glei

Access to Social Support Social support, defined as emotional or instrumental support elicited from others, and predicated on the presence of stable human relationships, can also be an important indicator of “readiness to parent.” Social support may derive from friends, a spouse, other family members, or individuals previously unknown to a parent, such as child care providers or counselors (Wilson et al. 1997; Belsky 1984). The principal argument is that social relationships and support improve parenting both directly, by providing information and household assistance, and indirectly, by buffering parents from the harmful effects of other stress. Stressful life events may include unemployment, divorce, and ongoing conditions such as financial hardship and single parenthood (House et al. 1988; McLoyd et al. 1994; Goldberg and Easterbrooks 1988; Umberson and Gove 1986; Cohen and Willis 1985). A large body of empirical research documents a positive relationship between social support and health or well-being (see reviews by Adler and Mathews 1994; House et al. 1988; Cohen and Willis 1985). Some literature focuses specifically on the effect of social support on parenting. Wakschlag et al. (in press) found that young mothers who received emotional support from their own mothers were more likely to be emotionally available to their children than were young mothers who did not receive support. Similarly, Crockenberg (1987) found that among economically deprived teen mothers, proximity to family members who were willing to help with household and childcare tasks was associated with increased maternal sensibility and accessibility. In other words, mothers who had family members nearby to help with chores and offer support and advice exhibited lower levels of stress and higher levels of positive parent-child interaction. Finally, researchers found that maintenance of social networks and reliance on social support were important coping mechanisms for parents in families under stress (McCubbin et al. 1983). Social support is also associated with reduced incidence of child abuse and neglect. Newburger and colleagues (1977) found that lack of social support significantly increased the likelihood of child abuse. Similarly, in a sample of poor African American mothers, McLoyd and colleagues (1994) reported that increased availability of social support reduced the incidence of harsh punishment. Focusing specifically on child neglect, researchers have shown that neglectful parents are more likely than others to be socially isolated (Polansky et al. 1985a, 1985b). Not all research points to a positive relationship between kin involvement and parenting skills. Whereas Stevens (1984) found that maternal grandmothers’ knowledge about infant development was a strong predictor of their adolescent daughters’ knowledge, Chase-Lansdale and colleagues (1994) found only a few significant correlations between mothers’

Preparing for Parenthood

55

and daughters’ parenting practices. They did, however, find a positive correlation between negative parenting styles; further, contrary to much previous research, they found that living with a grandmother has a negative effect on the quality of mothers’ parenting.2 These inconsistent results suggest that benefits accruing from a grandmother’s help depend on the grandmother herself, her relationship with the mother, and the financial status of the family. Previous research shows that harsh parenting practices persist across generations (Patterson 1986; Simmons et al. 1991), suggesting that negative or punitive parenting is learned and transmitted from one generation to the next (Chase-Landsdale et al. 1994). Thus, in instances where the grandmother herself used harsh disciplinary practices, her influence is likely to be more negative than positive. Psychological Resources Parents’ psychological status is the final indicator of “readiness to parent.” Just as social support can be an important resource for new parents, selfesteem and a sense of efficacy can play an essential role in easing the transition to parenthood and improving parenting skills among new parents. Emotionally healthy parents are more likely to provide a nurturing home environment, which fosters positive child outcomes. They are also more likely to be able to deal with the daily stresses that come with parenting and to establish a stable environment and regular daily routines for the child. Despite a paucity of research on changes in self-definition that accompany the transition to parenthood, we might assume that maternal selfconfidence is an important aspect in the process. Deutsch and colleagues (1988) found that levels of maternal self-confidence increase during the transition to parenthood, and argue that such changes may have an important influence on subsequent postpartum adaptation. A large body of research documents the link between parents’ emotional state and parenting behavior (for a review, see McLoyd 1990). For example, Conger and colleagues (1984) found that mothers who reported high emotional distress were less likely to exhibit positive parenting practices, such as giving hugs, praise, and supportive statements. Other researchers have found that mothers who are experiencing emotional distress and depression are more likely to abuse their children, use coercive discipline, and exhibit diminished maternal sensibility and satisfaction with parenting (Daniel et al. 1983; Crnic et al. 1983; Patterson 1988). 2 The negative association may be due to selection – that is, grandmothers may co-reside with daughters whom they perceive to be poorly equipped to raise their children.

56

Gager, McLanahan, Glei

All the evidence regarding social support and psychological resources is based on survey data. We have been unable to locate experiments designed specifically to increase mothers’ access to social support, self-confidence, or self-esteem, although some interventions for teen mothers and lowincome parents have included components relevant to those outcomes. Since positive relationships with friends and relatives and having a sense of efficacy and self worth are likely to be to correlated with other factors affecting parenting and child outcomes, we cannot say for certain that social support and psychological resources themselves have a causal effect. Nevertheless, the notion that they do has considerable intuitive appeal. Related Factors Some indicators of “readiness to parent” are correlated with parents’ economic and educational status. Race is also associated with many of these indicators, reflecting the fact that African Americans and Hispanics are disproportionately represented among lower socioeconomic groups (McLoyd 1990; Taylor and Roberts 1995). Seventy-five percent of women classified as living below the poverty line have reported having an unintended pregnancy, compared with only 45 percent of women living above the poverty line (Brown and Eisenberg 1995; Forrest and Singh 1990). Low income may also be associated with participation in childbirth or parenting classes, since those services are not usually provided free of charge. Low-income parents, and African Americans in particular, have been shown to have complex kin networks and to pool resources in response to harsh economic conditions (Stack 1974; Wilson 1987). Thus, we might expect parents in those groups to have better social support systems than higher income parents and whites. Previous research also suggests that poverty and economic insecurity increase depression and psychological distress (McLoyd 1990; Conger et al. 1984). Therefore, we would expect parents from low socioeconomic groups to be less prepared psychologically than others to cope with their new responsibilities. Family structure is also a risk factor for readiness to parent. Unmarried mothers are much more likely to report an unintended pregnancy than married mothers (Brown and Eisenberg 1995), and they are less likely to attend childbirth and parenting classes. If parents are married (or cohabiting), their principal source of support is usually their spouse (Belsky 1984). In contrast, single mothers are more likely to rely on their own parents or members of their extended family for support (Coletta 1979). Finally, we would expect unmarried mothers to have fewer psychological resources than married or cohabiting mothers, in part because of their precarious economic situation and in part because they lack the support of another parent.

Preparing for Parenthood

57

Parents’ experiences growing up within their own families should also affect their readiness to parent. A new line of research suggests that childhood sexual abuse may increase the likelihood of an unintended pregnancy (Boyer and Fine 1992; Butler and Burton 1990; Gershenson et al. 1989). Musick (1993) argues that the causal mechanisms operating here are low self-esteem and feelings of powerlessness and hopelessness (Briere 1992; Oates et al. 1985). Such feelings may result in early initiation of sexual activity, and are likely to impede the careful use of contraception once sexual activity begins. Individuals who were abused as children or who had negative role models may also be reluctant to rely on their own parents for social support. Thus, a mother’s relationship with her own mother is likely to affect her psychological resources. As noted already, some researchers have found that grandmothers’ knowledge of infant development is predictive of daughters’ knowledge (Stevens 1984). Having a positive role model is therefore likely to increase a mother’s selfconfidence, especially during the transition to parenthood (Ruble et al. 1990). Finally, mother’s age and experience are likely to be risk factors in her readiness to parent. Teenage mothers have a much higher incidence of unintended pregnancy than other mothers, while women near the end of their childbearing years also have a higher incidence of unwanted pregnancies. Thus, the relationship between age and pregnancy intentions is not linear (Brown and Eisenberg 1995). In addition, very young mothers may be less informed about parenting classes or may lack the resources to participate in such classes. They may also have less confidence in their ability to raise their new babies. In contrast, young mothers are more likely than other mothers to live near their own parents, and they may have more social support from extended kin. Parents of firstborn children may have some of the same risk factors as young parents because they lack experience: They may be more likely to have an unintended pregnancy, less likely to attend parenting classes, and less confident about their parenting ability. At the same time, because of their inexperience, they may receive more support from their parents and relatives.

survey data and analyses Indicators and Variables Parents’ answers to questions in the Commonwealth Survey offer insights into each of the four major “readiness to parent,” or outcome, indicators: pregnancy intention, parent education classes, access to social support, and psychological resources (Table 3.1). The survey also provides information about independent, or predictor, variables.

table 3.1. Means and Frequency Distributions for Readiness to Parent Outcome and Predictor Variables Predictor variables Educational attainment Less than high school High school graduate Some college College graduate African American Hispanic Family structure Two biological married parents Two biological cohabiting parents Never-married single parent Unstable family structurea Raise child differentlyb Respondent was abused while growing upc Age of respondent at child’s birth Child is firstborn Female respondent Outcome variables Pregnancy intention Wanted On-time Mistimed Unwanted Parent education class attendance Childbirth class Parenting class Either childbirth or parenting class Social support Grandparent nearby Relatives/friends to help Either grandparent or relative/friends Psychological resources Confident at birth and now Coping very well now Both confident and coping well Total readiness scored All four domains Three domains Two domains One domain Zero domains

Percentage (or Mean) 13.7 33.1 27.2 26.0 10.1 12.9 80.5 5.0 10.1 4.4 54.2 26.8 29.4 years 41.5 56.6

95.7 71.2 24.5 4.2 71.0 34.4 76.2 77.8 70.1 87.3 65.1 56.1 40.0 28.7 40.2 22.2 8.3 0.7

a Includes parents who are divorced, separated, remarried, or never married but cohabitating

with someone other than the child’s father. on the question: “In general, do you want to raise your child in the same way that you were raised, or do you want to raise him/her differently?” c Self report of verbal/emotional, physical, or sexual abuse while growing up. d The four domains are (1) an intended (wanted and on-time) pregnancy, (2) attending a childbirth or parenting class, (3) living near a grandparent or having many friends or relatives to help out, and (4) feeling very confident at birth and now, and coping well. b Based

58

Preparing for Parenthood

59

Pregnancy intentions are measured by a question that asks parents, “When (you/mother) became pregnant, did you want (her) to become pregnant at that time, would you have preferred (her) to become pregnant at a later time, or did you never want (her) to become pregnant?” Parents who said that they wanted the pregnancy to occur at that time or a later time were coded as having a wanted pregnancy. Parents who said that they wanted the pregnancy to occur at that time were coded as having an on-time pregnancy. Attendance at parent education classes is measured by two dichotomous variables indicating whether or not the respondent answered yes to two questions, “Have you ever, personally, attended childbirth classes” and “Have you ever, personally, attended parenting classes?” Access to social support is measured by two dichotomous variables. The first variable is based on the question, “Do any of the child’s grandparents live within one hour’s driving distance of your home?” This variable, which taps the proximity of social support networks, was coded 1 if respondents answered yes. The second variable is based on the question, “If you had some kind of problem or emergency and needed help, how many relatives or friends do you have living nearby who you could count on for help – would you say many, a few, or none?” This variable was coded 1 if the respondents reported “many.” Finally, psychological resources are measured by three indicators. Parents were queried, “In general, how well do you feel you are coping with the demands of parenting: very well, about average, not too well, or not well at all?” This variable was coded 1 if parents reported they were coping “very well.” To measure parenting confidence, we created a dichotomous variable by combining two questions on confidence, “How confident did you feel about caring for your child at birth” and “How confident do you feel about caring for your child now?” Parents who reported being “very confident” at birth and now were coded as 1. Parents’ educational attainment is measured by an ordinal variable with four categories: less than a high school degree, high school graduate or GED, attended some college, and college graduate. Less than 14 percent of respondents had not completed high school, 33 percent had graduated from high school, and more than 26 percent had completed college. Race and ethnicity are indicated by dummy variables for non-Hispanic African American and Hispanic; the reference group is non-Hispanic whites and others. Ten percent of the respondents are African American, 13 percent are of Hispanic origin, and the remaining 77 percent are white or other race. Family structure is measured by a set of dummy variables indicating whether the child is living with two married biological parents, two cohabiting biological parents, a never-married parent living alone, or a parent who is divorced, separated, remarried, or never-married but cohabiting

60

Gager, McLanahan, Glei

with a partner who is not the child’s father. The indicator of family structure reflects parents’ relationship at the time of the interview, not at the time of the child’s birth. The majority of children are living with both biological parents, either married (81 percent) or cohabiting (5 percent), 10 percent are living with a never-married parent, and 4 percent are living in an unstable family. Parents’ childhood experience is measured by two variables. The first is based on the question, “In general, do you want to raise your child in the same way that you were raised, or do you want to raise him/her differently?” An answer indicating a wish to raise the child differently may capture the parent’s lack of positive parenting role models, although it may also merely reflect changes in parenting style from one generation to the next. More than half of parents in the survey (54 percent) said they intend to raise their child differently. The second variable is based on the question, “When you were growing up, do you feel you were ever: (1) verbally or emotionally abused, (2) physically abused, or (3) sexually abused, or not?” The abuse variable was coded 1 if the respondent answered yes to any type of abuse. Twenty-seven percent of respondents reported some type of abuse. Age and parenting experience are measured by two variables: age of respondent at the birth of the child and whether the child is the first born. The age of the parent is measured as a continuous variable, with a mean of 29.4 years old. Nearly 42 percent of parents reported having only one child. Finally, we included a measure of the respondent’s sex. Nearly 57 percent of the respondents are women. Analytic Techniques We began our analysis by examining the bivariate relationships between the “parenting readiness” indicators and parents’ background characteristics. Next, we estimated multivariate logistic regression models, presenting three sets of estimates for each domain. For pregnancy intention, we estimated the effects of background variables on whether the birth was wanted, whether it occurred on time (conditional on being wanted), and whether the birth was intended (both wanted and on time). For parenting education, we estimated models for attending childbirth class, attending parenting class, and the combined measure – whether the parent attended either a childbirth class or a parenting class. For social support, we estimated models for whether a grandparent lives nearby, whether respondent has many relatives or friends to help in an emergency, and the combined measure – whether the parent has either a grandparent or many friends or relatives to call for help. Finally, for psychological resources, we estimated a model for whether respondents are coping very well with the demands of parenting, whether they feel very confident about their ability to care

Preparing for Parenthood

61

for their child (both at birth and at the time of the survey), and whether they feel confident and are coping well. To illustrate how “parenting readiness” differs by family type, we used the coefficients from the logistic regression models to calculate predicted probabilities for six subgroups of parents, differentiated by marital status and education level. We focused on two marital status groups (nevermarried parents and continually married parents) and three levels of education (high school graduate or GED, some college, and college graduate or more). Predicted probabilities were calculated by changing the values of the family structure and education variables, while leaving all other variables in the equation at their observed values. Next, we examined the relationship between parents’ background characteristics and a summary index of “parenting readiness.” First, we created an index, ranging from 0 to 4, that counts the number of domains in which parents reported being ready. For example, parents who reported (1) an intended pregnancy, (2) attending a birthing or parenting class, (3) living near a grandparent or having many friends or relatives to help out, and (4) feeling very confident and coping well with their new responsibilities got a score of 4. Second, we created a dichotomous variable that took the value of 1 if a parent scored 2 or higher on the readiness index and 0 if he or she scored 1 or below. The dichotomous indicator allowed us to distinguish between parents who are “not ready” and parents who pass a minimum threshold. Finally, to demonstrate the validity of our indicators, we estimated several models that treat “good parenting behaviors” as a function of the readiness indicators. The parenting behaviors include breastfeeding the child for more than one month, reading to the child frequently, refraining from yelling at the child, refraining from spanking the child, and establishing a regular bedtime routine for the child.

results The majority of new parents (96 percent) reported that their pregnancy was wanted (Table 3.1). Seventy-one percent said that the pregnancy was wanted and occurred at the right time, while 25 percent said that the pregnancy was wanted but they would have preferred a later time. More than three-quarters of new parents reported attending either childbirth or parenting classes, with 71 percent attending childbirth classes and 34 percent attending parenting classes. Nearly four out of five new parents said they had a grandparent living nearby, and 70 percent reported having many relatives or friends to help in an emergency. Altogether, more than 87 percent had either a grandparent nearby or many people to call on in an emergency. Nearly two-thirds of parents reported feeling very confident about their parenting skills both at birth and at the time of the survey, and 56 percent

62

Gager, McLanahan, Glei

reported that they were coping very well with the demands of parenting. Only 40 percent of all new parents scored high on both measures of psychological adjustment. Nearly 29 percent of new parents reported being prepared in all four “readiness to parent” domains, another 40 percent reported being ready in three out of four domains, and 22 percent said they were ready in two domains. Eight percent reported being ready in one domain, and less than 1 percent of parents in the survey were not ready in any domain. As presented in Table 3.2, bivariate analyses of parental characteristics and the “readiness to parent” indicators show that college-educated parents, white parents, parents of firstborn children, and parents in intact families were more likely than other parents to say that the pregnancy was wanted. Reports of childhood abuse and wanting to raise a child differently were not related to whether or not the child was wanted. College graduates and parents in intact families were much more likely to report an on-time pregnancy than parents who did not complete high school or were never married. Parents who were abused as children and said they wanted to raise their child differently were less likely to report on-time pregnancy. Not surprisingly, education is strongly associated with taking a childbirth or parenting class. Parents with a college degree were more than twice as likely to attend a birthing class as parents who had not completed high school. White parents were the most likely and Hispanic parents the least likely to attend classes. Married parents were more likely to attend classes than unmarried parents. For social support, African American, never-married mothers and parents who are high school graduates and first-time parents were the most likely to report having a grandparent nearby to help. Having many friends or relatives to help was not strongly related to parents’ demographic characteristics, although parents who reported childhood abuse or who wanted to raise their child differently were less likely to report having social support than other parents. Finally, parents with the least education were the most likely to report feeling very confident, with the level of confidence declining steadily as education level increased. Confidence was highest among Hispanics, divorced mothers, parents with other children, and parents who were not abused as children. In contrast, parents with the most education were most likely to report that they were coping very well. First-time parents, African American parents, and cohabiting parents were also more likely to report that they were coping well. Finally, parents who reported being abused and parents who wanted to raise their child differently were less likely than other parents to be coping well. Readiness scores across the four domains show that overall readiness is strongly related to family structure and education. While only 6.7 percent

49.2 74.4

71.9 77.9

Sex of parent respondent Female 94.2 Male 97.9

72.1 85.9

96.1 99.0

84.8 94.2

69.9

94.1

77.1

65.1

92.8

97.4

Race Non-Hispanic white African American Hispanic

Education Less than high school High school graduate Some college College graduate

70.3 72.0

54.1 52.7

76.2

80.9 86.5

64.8

34.1

36.3 32.0

34.3 30.7

34.9

37.7 44.5

27.9

24.4

79.1 75.9

82.6 60.1

80.2

82.8 67.5

86.2

67.2

Childbirth Parenting Grandparent Wanted On-time class class nearby (%) (%) (%) (%) (%)

68.3 72.4

63.4 64.7

71.8

71.2 63.5

75.3

67.6

63.7 66.9

69.9 72.9

63.3

64.7 57.5

67.5

73.9

54.2 58.4

65.2 54.1

55.3

54.0 62.3

51.9

57.8

(continued)

20.8 20.5

13.1 16.9

22.3

23.7 23.3

19.7

11.3

Many Confident at Coping Ready in all to help birth and well 4 domains (%) now (%) (%) (%)

table 3.2. Bivariate Relationships between Parents’ Characteristics and Readiness Indicators

63.6

29.8

59.9

93.8

88.1

85.2

95.3 96.3

94.9 96.0

Raise differently Yes No

Abused as child Yes No

67.7 76.9

70.9 78.4

79.4

74.5 74.3

97.2

97.8 94.3

Family structure Two biological married parents Two biological cohabiting parents Never-married single parent Divorced mother

Child is firstborn Yes No

70.3 71.3

69.3 73.1

49.5

45.4

56.5

75.3

69.7 71.9

40.3 32.4

35.1 33.7

35.5

33.2

30.9

34.7

32.3 35.9

79.0 77.4

77.9 77.9

84.9

87.1

73.1

76.4

81.7 75.0

Childbirth Parenting Grandparent Wanted On-time class class nearby (%) (%) (%) (%) (%)

table 3.2. (Continued)

62.5 73.1

65.4 75.7

71.8

75.3

74.1

69.2

75.0 66.6

60.8 66.5

66.9 63.2

69.4

62.6

61.6

65.3

42.2 81.5

49.9 58.1

53.0 59.6

53.9

57.6

61.1

55.5

59.7 53.4

15.3 22.7

18.4 23.4

10.9

6.7

14.8

22.7

16.6 23.6

Many Confident at Coping Ready in all to help birth and well 4 domains (%) now (%) (%) (%)

Preparing for Parenthood

65

of never-married single mothers were ready in all four domains, nearly 23 percent of parents in intact marriages had a score of 4. Similarly, whereas 11 percent of parents without a high school degree were ready for parenting in all four domains, 23 percent of parents with at least some college fell into this category. Finally, parents’ own childhood experiences were related to “parenting readiness.” The numbers reported in Tables 3.3 through 3.6 are odds ratios (eβ ) and control for other variables in the model. Pregnancy Intention As shown in Table 3.3, parents who completed college were more than four times as likely as parents who did not finish high school to report that their child was wanted. Level of education was not significantly related, however, to likelihood of having an on-time pregnancy, although it might be expected that education would have a positive and significant effect on both measures of intention. African Americans were much less likely than whites to report that their pregnancy was wanted, on-time, and intended. Compared with whites, the odds of an African American parent reporting a wanted pregnancy were nearly 73 percent lower, an on-time pregnancy nearly 46 percent lower, and an intended pregnancy 55 percent lower. These differences are net of other variables, such as education, marital status, and age of mother, although propensity to rationalize an unintended pregnancy may be relevant, since these data are based on retrospective reports. Never-married parents and cohabiting parents were less likely than married parents to report that their pregnancy was on-time or intended, as were parents in unstable relationships. Although these results suggest that marital status and family structure have significant effects on birth intentions, family structure itself may be determined by pregnancy intentions. For example, a “shotgun” marriage in response to an unintended pregnancy may result in a divorce. Since family structure is measured at the time of the interview as opposed to the birth (or conception) of the child, the sequencing of these variables – family structure and intentions – is somewhat ambiguous. The next variables provide information about respondents’ experiences with their own parents while growing up: whether they would raise their own child differently, and whether they were abused in some way. Consistent with expectations, growing up with an abusive parent reduced the odds of having an on-time or intended pregnancy. These results lend credibility to the argument that abuse during childhood increases the risk of having an unplanned pregnancy (Briere 1992; Oates et al. 1985). Older parents were more likely than younger parents to report that their pregnancy occurred at the right time and that it was intended. Parents of

0.17

0.10

0.13

0.43

0.41

0.29

0.99 0.75

0.95 2.25 0.46

Never-married single parent Two biological cohabiting parents Unstable family structure

Raise child differently Respondent was abused

Age of respondent at birth Child is firstborn Female respondent




3





1.01 5.03 0.98

1.74 1.43

0.64

1.66

1.08

0.60 1.29

1.97 3.24 17.72

1.07 1.51 1.39

0.81 0.66

0.54

0.45

0.17

0.54 1.30

0.83 0.76 1.28

1.03 1.11 1.01

0.59 0.47

0.31

0.27

0.11

0.36 0.88

0.52 0.47 0.72

1.11 2.05 1.91

1.11 0.92

0.94

0.78

0.27

0.81 1.92

1.31 1.23 2.30

1.05 1.58 1.23

0.82 0.66

0.43

0.43

0.17

0.45 1.19

0.86 0.84 1.54

1.02 1.17 0.90

0.61 0.48

0.26

0.25

0.11

0.30 0.83

0.55 0.53 0.89

1.08 2.13 1.66

1.11 0.91

0.70

0.72

0.27

0.67 1.71

1.32 1.33 2.68

Reference category for education = less than high school, race = non-Hispanic white/other, family structure = two biological married parents.

Note: Intended includes births that were “on time” and “wanted.” Sample does not include respondents whose pregnancies ended in an abortion or stillbirth. For education, omitted category is < high school degree. For race/ethnicity, omitted category is non-Hispanic white. For family structure, omitted category is two married biological parents.

0.90 1.00 0.21

0.57 0.40

0.12 0.31

0.27 0.64

African American Hispanic

0.45 0.56 1.03

0.94 1.35 4.27

High school graduate Some college College graduate

Variables



2




1



Intended vs. unintended

95% 95% 95% 95% 95% 95% Odds ratio CI lower CI upper Odds ratio CI lower CI upper Odds ratio CI lower CI upper




On-time vs. mistimed

Wanted vs. unwanted

table 3.3. Effects of Parents’ Characteristics on the Odds of Pregnancy Intention Status

Preparing for Parenthood

67

firstborn children were more likely than other parents to report that their pregnancy was wanted, on-time, and intended. The odds of reporting a wanted pregnancy were more than twice as high for first births. Finally, the sex of the parent affects pregnancy intentions, although not in a consistent fashion. Mothers are significantly less likely than fathers to report that a pregnancy was wanted, but they are more likely than fathers to report that it occurred at the preferred time. It is difficult to say whether the gender difference is due to real differences in men’s and women’s pregnancy intentions or to differences in their propensity to reinterpret the past after the baby is born. Participation in Parent Education Classes Consistent with the bivariate results in Table 3.2, education is an important predictor of class attendance (Table 3.4). The likelihood of attending a childbirth or parenting class increased with each level of education. African American and Hispanic parents were less likely than white parents to attend a childbirth class but equally likely to attend a parenting class. Since African Americans and Hispanics are disproportionately represented among teen parents and low-income parents, their chances of being targeted by interventions aimed at these populations may be higher than those of other parents. Parent characteristics had few other significant effects on class attendance, and those effects were not consistent across type of class. Parents in unstable families and parents who were never married were less likely to attend childbirth classes than married parents. Parents who were abused as children were much more likely to attend parenting classes but not birthing classes. Finally, older, female respondents were more likely to attend childbirth classes, while parents of firstborn children were less likely to attend parenting classes. Access to Social Support Education is strongly related to having grandparents nearby (Table 3.5). Parents who completed high school or attended college were more than twice as likely as those who did not finish high school to live within an hour of a grandparent. In contrast, completing college was not associated with proximity to grandparents. Since increased education is associated with greater mobility, especially at higher education levels, these results are not surprising. Education had no effect on having many friends or relatives to help in an emergency. African Americans were less likely than whites to report having relatives or friends to help out, and Hispanic parents were less likely than whites to report proximity to a grandparent or having many friends or relatives

0.39

0.27

0.67

0.44

1.07 0.87

1.04 0.94 1.47

Raise child differently Respondent was abused

Age of respondent at birth Child is firstborn Female respondent




3





1.07 1.27 1.97

1.42 1.20

0.72

1.13

0.98

3.85 8.52 12.68 0.76 0.94

0.99 0.74 1.14

1.00 1.45

1.14

0.89

1.31

1.34 2.16 3.39 0.94 1.15

0.96 0.56 0.88

0.76 1.08

0.74

0.51

0.88

0.89 1.41 2.14 0.65 0.83

1.01 0.97 1.48

1.31 1.94

1.77

1.52

1.94

2.00 3.29 5.36 1.34 1.60

1.02 0.87 1.26

1.03 0.07

0.80

0.67

1.09

2.13 4.64 6.54 0.49 0.70

1.00 0.64 0.94

0.77 0.77

0.50

0.40

0.71

1.45 3.00 3.85 0.33 0.50

1.05 1.18 1.69

1.38 1.49

1.28

1.15

1.67

3.12 7.16 11.10 0.72 0.97

Note: For education, omitted category is < high school degree. For race/ethnicity, omitted category is non-Hispanic white. For family structure, omitted category is two married biological parents.

1.01 0.70 1.09

0.81 0.63

0.42

0.64

Never-married single parent Two biological cohabiting parents Unstable family structure

1.75 3.57 4.56 0.35 0.49

2.59 5.52 7.61 0.52 0.68

High school graduate Some college College graduate African American Hispanic

Variables



2




1



Attend either class

95% 95% 95% 95% 95% 95% Odds ratio CI lower CI upper Odds ratio CI lower CI upper Odds ratio CI lower CI upper




Attend parenting class

Attend childbirth class

table 3.4. Effects of Parents’ Characteristics on the Odds of Parent Education Class Attendance

0.89

0.51

1.08

1.54

0.90

1.86

Never-married single parent Two biological cohabiting parents Unstable family structure




2






0.88 0.74

1.22 0.99

1.68 1.33

0.97

1.52 1.43

3.19

1.57

2.67

1.29 0.52

4.40 3.65 1.89

1.23 0.72

0.95

0.66 0.74

1.23

1.27

1.16

0.57 0.71

1.43 1.28 0.91

0.93 0.55

0.93

0.50 0.55

0.81

0.71

0.76

0.39 0.51

0.96 0.84 0.57

1.64 0.95

0.98

0.87 1.00

1.86

2.78

1.76

0.82 0.99

2.13 1.96 1.45

1.36 0.71

0.95

1.00 0.81

1.31

1.27

1.82

0.87 0.59

2.40 2.04 1.01

0.91 0.49

0.93

0.67 0.53

0.73

0.58

0.85

0.49 0.37

1.33 1.09 0.54

95% CI lower

3





2.03 1.05

0.99

1.49 1.24

2.37

2.79

3.91

1.55 0.96

4.33 3.81 1.92

95% CI upper

Note: For education, omitted category is < high school degree. For race/ethnicity, omitted category is non-Hispanic white. For family structure, omitted category is two married biological parents.

0.92

0.95

Age of respondent at birth Child is firstborn Female respondent

0.80 0.70

1.10 1.00

Raise child differently Respondent was abused

0.49 0.26

0.79 0.36

African American Hispanic

1.80 1.38 0.69

2.82 2.24 1.14

High school graduate Some college College graduate

Variables

1



Many relatives /friends to help Either grandparent or relatives /friends

95% 95% 95% 95% Odds ratio CI lower CI upper Odds ratio CI lower CI upper Odds ratio




Grandparent nearby

table 3.5. Effects of Parents’ Characteristics on the Odds of Social Support

70

Gager, McLanahan, Glei

to call on for support. Results for African Americans and Hispanics do not support our hypothesis about higher prevalence of kin support among minority populations, although this is not the first analysis to find that support systems of African American parents may have been exaggerated in previous research. The lower level of support among Hispanics may be due to immigrant status. Only one family structure variable was associated with social support. Parents in unstable family structures – a category that includes separated, divorced, and remarried parents – were more likely to live near grandparents than parents in stable relationships. Unfortunately, there is no way to determine from the data whether these parents moved closer to their own parents before or after they entered unstable family situations. The results indicate that parents in nontraditional families are no less likely than others to have many friends and relatives to help in an emergency. Parents’ own childhood experiences were not significantly related to proximity to a grandparent but did reduce the odds of having friends and relatives to help out. Parents who want to raise their child differently were only two-thirds as likely to report having many people to rely on, and parents who said they were abused were only three-quarters as likely to report having help. Unfortunately, the wording of the question about available support confounds help from family and friends. We might expect respondents who had negative experiences growing up to get less help from grandparents, yet we might also expect those parents to seek more help from friends. Since the question does not distinguish between the two sources of help, it is not possible to isolate lack of support from relatives, perhaps self-imposed by the parent, from more general problems in establishing close relationships with other people. Not surprisingly, age is negatively related to both types of social support. Older parents are less likely than younger parents to have a grandparent living nearby and less likely to have a large social network to turn to for support. Mothers are less likely than fathers to report having many people to turn to in an emergency. Psychological Resources The survey data do not support the hypothesis that parents from disadvantaged backgrounds would report fewer psychological resources than parents from more advantaged backgrounds (Table 3.6). Education had no effect on coping and was negatively correlated with confidence. Perhaps highly educated parents have a greater sense of their own limitations and therefore set unusually high expectations for themselves. African Americans were more likely than whites to say they are coping very well and to report positive feelings on the combined measure of coping and confidence. It is difficult to know whether these results reflect real or perceived differences in ability to cope.

0.76

0.79

0.65

1.10

1.34

0.98

0.85 0.80

1.03 1.46 0.91

Never-married single parent Two biological cohabiting parents Unstable family structure

Raise child differently Respondent was abused

Age of respondent at birth Child is firstborn Female respondent




3





1.06 1.88 1.16

1.10 1.05

1.48

2.26

1.57

2.43 1.38

1.07 1.15 1.59

1.01 0.17 0.88

1.09 0.68

0.90

0.88

1.15

1.19 1.16

0.69 0.59 0.46

0.99 0.13 0.67

0.82 0.49

0.53

0.49

0.76

0.77 0.82

0.45 0.37 0.28

1.04 0.23 1.16

1.44 0.93

1.54

1.58

1.74

1.83 1.65

1.07 0.93 0.76

1.03 0.55 0.98

0.97 0.68

0.96

1.24

1.13

1.62 0.97

0.65 0.66 0.67

1.01 0.42 0.77

0.75 0.51

0.62

0.74

0.77

1.14 0.70

0.44 0.44 0.43

1.05 0.71 1.26

1.25 0.90

1.50

2.07

1.66

2.29 1.34

0.96 1.00 1.05

Note: For education, omitted category is < high school degree. For race/ethnicity, omitted category is non-Hispanic white. For family structure, omitted category is two married biological parents.

1.01 1.13 0.71

0.66 0.60

1.21 0.74

1.72 1.01

African American Hispanic

0.51 0.51 0.65

0.74 0.77 1.02

High school graduate Some college College graduate

Variables



2




1



Coping well and confident

95% 95% 95% 95% 95% 95% Odds ratio CI lower CI upper Odds ratio CI lower CI upper Odds ratio CI lower CI upper




Confident at childbirth and now

Coping well

table 3.6. Effects of Parents’ Characteristics on the Odds of Coping with Parenthood

72

Gager, McLanahan, Glei

Parents who were abused as children were less likely than others to report feeling very confident about caring for their child or to score well on the combined measure. Older parents were more likely to report that they are coping well. Parents of firstborn children are more likely to report that they are coping well and less likely to report feeling confident. These results suggest that additional children in the home increase tension in the family, and that new parents are less confident than more experienced parents. Readiness Index Table 3.7 shows the effects of background characteristics on the two summary indices, the first index ranging from 0 to 4 and the second from 0 to 1. The former gauges the effects of background variables on the entire distribution of the readiness index, whereas the latter focuses on the effects at the bottom of the distribution. More specifically, the second measure helps identify characteristics that distinguish parents who are not ready. The numbers reported are based on OLS regressions (column 1) and logistic regressions (column 2) and are coefficients rather than odds ratios. Thus, a positive number means that a background variable is associated with an increase in the indices, and a negative number means the variable is associated with a reduction in the indices. The overall index indicates that education is positively related to “readiness to parent.” Parents with some college and parents with a college degree were much better prepared to assume their responsibilities than parents without a high school degree. African American parents, never-married single mothers, and parents in unstable family arrangements scored lower on the overall index, as did parents who were abused as children. Older parents scored higher on the overall index, showing that age increases readiness. The dichotomous index reveals only four major risk factors: Parents who were younger, African American, lacking a college degree, or abused as children were all less ready to be parents. Readiness and “Good Parenting” To confirm the ability of the readiness indicators to predict parental behavior, we estimated several regression models where parenting behavior was the dependent variable and each indicator was an independent variable. Measures of “good parenting” included whether the child was breastfed for more than a month, whether the parent spanks the child, whether the parent yells at the child, how often the parent reads to the child, and whether the child has regular bedtime routines. Pregnancy intentions were not related to parenting behaviors (Table 3.8), a surprising result that is inconsistent with some previous

−0.51 −0.42 −0.43 −0.14 −0.30 0.00 −0.16 −0.08

−0.33

−0.19

−0.23

−0.03 −0.18 0.01 −0.05 0.03

Never married single parent Two biological cohabiting parents Unstable family structure

Raise child differently Respondent was abused Age of respondent at birth Child is firstborn Female respondent

b Coded

ranging from 0–4. 1 if overall index is 2 or more, otherwise, coded 0.

a Variable

−0.36 −0.27

−0.19 −0.12

African American Hispanic

95% CI lower 1.91 −0.05 0.07 0.10

OLS coefficient

0.07 −0.05 0.02 0.06 0.15

−0.04

0.04

−0.15

−0.03 0.03

2.65 0.32 0.45 0.51

95% CI upper

0.19 −0.57 0.05 −0.06 −0.11

−0.10

−0.57

−0.11

−0.65 −0.40

0.84 −0.08 0.59 0.93

Logit coefficient

−0.22 −1.02 0.01 −0.48 −0.59

−0.72

−1.28

−0.69

0.60 −0.13 0.09 0.35 0.36

0.52

0.15

0.47

−0.10 0.11

2.23 0.49 1.23 1.77

−0.55 −0.64 −0.05 0.09 −1.19 −0.91

95% CI upper

95% CI lower

(2)

(1)

2.28 0.13 0.26 0.30

Intercept High school graduate Some college College graduate

Variables

Dichotomous indexb

Overall indexa

table 3.7. Effects of Parents‘ Characteristics on Overall Readiness

Gager, McLanahan, Glei

74

table 3.8. Effects of Readiness Indicators on Parenting Behaviors a

Mistimed Unwanted Birth class Parent class Grandparent nearby Many to help Very confident Coping well a Whether

Breastfed

Reading

−.10 −.23 .44 .26 −.46 −.20 .02 −.13

−.11 .05 .20 .02 .05 −.07 .18 .04

b

Spanking −.003 .06 −.02 −.02 .03 −.08 .03 −.15

c

d

Yelling .05 .16 .04 −.18 .14 −.19 −.06 −.31

Bedtimee routine −.10 −.17 .28 .05 −.02 .41 .50 .06

the child was breastfed for more than one month. Logit coefficients are shown.

b How often the parent reads to the child; ordinal variable ranging from 1–5 (1 = “Not at all,”

5 = “More than once a day”). OLS coefficients are shown.

c How often the parent spanks the child; ordinal variable ranging from 0–3 (0 = “Never,” 3 =

“Often”). OLS coefficients are shown. often the parent yells at or raises voice with the child; ordinal variable ranging from 0–3 (0 = “Never,” 3 = “Often”). OLS coefficients are shown. e Whether the child has a regular bedtime routine. Logit coefficients are shown. d How

research. In contrast, attending classes was positively associated with several measures of “good” parenting. Childbirth classes were associated with higher incidences of breastfeeding, reading, and regular bedtime routines, and parenting classes were associated with breastfeeding and refraining from yelling. Having friends and relatives to call on in case of an emergency reduced harsh punishment and increased the likelihood of regular routines. Surprisingly, having grandparents close by was not associated with positive parenting practices; indeed, this variable is negatively related to breastfeeding and refraining from yelling. Predictably, parents who said they were coping well were less likely to use harsh punishment (yelling and spanking), and parents who said they were confident were more likely to read to their children and institute consistent bedtime routines.

discussion and policy implications Overall, the results of these analyses should be encouraging to policymakers and practitioners who are concerned about child health and development. Even so, the findings should be taken with some caution, since the indicators are rather crude and may not accurately measure the domain of interest. To give just two examples, we have no information on the relationship between the mother and the father, which may have a stronger effect on “good parenting” than family structure per se. Similarly, the nature of the relationship between parents and grandparent is probably a more

Preparing for Parenthood

75

important indicator of social support than proximity. A grandparent who is punitive and critical may undermine parents’ confidence and their ability to be good parents to a new baby. Risk Factors for Readiness to Parent With respect to birth intentions, the key risk factors are being African American, living in a nontraditional family, coming from an abusive family, and being young. This information will not be news to people who work in the area of family planning, although until now the evidence for childhood abuse as a risk factor has not been documented on a nationally representative dataset. For not attending parenting or childbirth classes, the major risk factors are low education and being African American or Hispanic. Again, these results do not come as a surprise. More surprising, however, are the risk factors for social support. The results indicate that Hispanic and African American parents are less likely than white parents to have friends and relatives to turn to for help. Although lack of social supports among Hispanics is probably due to the high percentage of recent immigrants, the finding for African Americans is inconsistent with conventional wisdom regarding access to large kinship networks. The major lesson here is that practitioners should be careful not to romanticize the support networks of African American families. Finally, the finding that highly educated and white parents are at higher risk of being psychologically unprepared for parenthood is surprising. Less educated parents typically have access to fewer economic resources than college graduates, and African Americans are much more likely than whites to be exposed to economic stress, which has been linked to depression by previous research. Educated parents may read more books and magazines on parenting, which may raise their expectations and undermine their confidence. The other risk factors for not being psychologically prepared are having been abused as a child, being a young parent, and being a parent for the first time. The new-parent effect is due primarily to the fact that new parents said they were not confident at the time of the child’s birth, although they were more likely to report they were doing well by the time of the survey. These findings suggest that parenting programs should try to reach new parents as soon as possible after their child’s birth and continue to target young parents at least until the child is three years old. To illustrate the size of some of these effects, we calculated predicted probabilities for six family types, distinguished by parents’ marital status and education. The results, reported in Table 3.9, show that education and marital status have large effects on whether or not a pregnancy was intended. A never-married mother with only a high school degree has a 34 percent chance of reporting an intended birth, whereas a married mother

Gager, McLanahan, Glei

76

table 3.9. Predicted Likelihood of Being Ready for Parenthood, by Mother’s Marital Status and Education Level Never-married mothers

Birth was intendeda Class participationb Social supportc Psychological resourcesd

Two-parent married

High school (%)

Some college (%)

College graduate (%)

High school (%)

Some college (%)

College graduate (%)

32

32

45

71

71

81

71

84

88

69

83

87

95 40

94 41

89 41

91 37

90 38

81 38

a Intended

indicates that the pregnancy was wanted and occurred at the right time. participation indicates that respondent participated in either a childbirth class or a parenting class. c Social support indicates that respondent either had grandparents nearby or many friends/relatives to help. d Psychological resources indicate that respondent is coping well and is/was confident about his/her parenting skills at birth and now. b Class

with a college degree has a 77 percent chance of doing so – a 43 percentage point difference. Marital status accounts for most of the difference between the two groups: The gap between a never-married mother with a high school degree and one with a college degree is only 13 percentage points, whereas the gap between a never-married mother with a college degree and a married mother with a college degree is 36 percentage points. Participation in childbirth or parenting classes also varies by family type. A never-married mother with a high school degree has a 70 percent chance of attending a childbirth or parenting class, whereas a married mother with a college degree has an 89 percent chance of attending one or the other. In this instance, education is the critical factor. Indeed, there is virtually no difference in class attendance between a never-married mother with a high school degree (71 percent) and a married mother with a high school degree (69 percent). Similarly, although a parent with a college degree was nearly 20 percentage points more likely to have attended class than a parent with a high school degree, the difference between college-educated never-married and married mothers is nil. Social support also varies by parents’ education and marital status. A never-married mother with a high school degree is most likely to have access to social supports, whereas a married college graduate is least likely. Neither education nor marital status is related to psychological resources. Analyses of connections between parental readiness and several behaviors that are viewed as indicators of “good parenting” – including

Preparing for Parenthood

77

breastfeeding, reading to the child, setting regular bedtime routines, and avoiding harsh punishments – both confirm and cast doubt on some of the principles behind current strategies for helping new parents. On the positive side, attending childbirth classes is positively associated with breastfeeding, reading, and regular bedtime routines, and attending parenting classes is associated with breastfeeding and less yelling at the child. Although we cannot be sure that classes have a causal effect on these parenting behaviors, the results are at least consistent with this interpretation. Also, having friends and relatives to call on in case of an emergency was associated with less use of harsh punishment and more use of regular routines, while having confidence in oneself as a parent was associated with less yelling and spanking and more reading and regular bedtime routines. Thus, policies that attempt to increase the social support and psychological skills of new parents would appear to be on the right track. Less encouragingly, we found no relationship between pregnancy intentions and any of the “good parenting” behaviors, and we found a negative relationship between having grandparents nearby and breastfeeding and refraining from yelling. The finding about grandparents is counterintuitive yet, as previously noted, we are not the first to find such an effect (ChaseLansdale et al. 1994). It raises questions about recent policies that encourage grandparents to become foster parents to their grandchildren or childcare providers when parents go to work. At a minimum, the findings suggest that policymakers should not assume that support from grandparents is always a good solution to foster care or childcare problems, and should make sure that new parents have options outside their kin networks.

references Adler, N., & Mathews, K. (1994). Health psychology: Why do some people get sick and some stay well? Annual Review of Psychology, 45, 229–59. Baydar, N., & Grady, W. (1993). Predictors of Birth Planning Status and Its Consequences for Children. Seattle, WA: Battelle Public Health Research and Evaluation Center. Belsky, J. (1984). The determinants of parenting: A process model, Child Development 55, 83–96. Blomberg, S. (1980). Influence of maternal stress during pregnancy on postnatal development. Acta Psychiatr Scand, 62, 402–17. Boyer, D., & Fine, D. (1992). Sexual abuse as a factor in adolescent pregnancy and child maltreatment. Family Planning Perspectives, 42, 4–19. Briere, J. (1992). Child Abuse Trauma: Theory and Treatment of the Lasting Effects. Newbury Park, CA: Sage. Brown, S., & Eisenberg, L. (1995). The Best Intentions. Washington, D.C.: National Academy Press. Butler, J., & Burton, L. (1990). Rethinking teenage childbearing: Is sexual abuse a missing link? Family Relations, 39, 73–80.

78

Gager, McLanahan, Glei

Chase-Landsdale, L., Brooks-Gunn, J., & Zamsky, E. (1994). Young AfricanAmerican multigenerational families in poverty: Quality of mothering and grandmothering. Child Development, 65, 373–93. Cohen, S., & Willis, T.A. (1985). Stress, social support and the buffering hypothesis. Psychological Bulletin, 98, 310–57. Coletta, N. (1979). Support systems after divorce: Incidence and impact. Journal of Marriage and the Family, 41, 837–46. Conger, R., McCarty, J., Yang, R., Lahey, B., & Kropp, J.D. (1984). Perception of child, childrearing values, and emotional distress as mediating links between environmental stressors and observed maternal behavior. Child Development, 54, 2234–47. Crnic, K., Greenberg, M., Ragozin, A., Robinson, N., & Baskam, R. (1983). Effects of stress and social support on mothers and premature and full-term infants. Child Development, 54, 209–17. Crockenberg, S.B. (1987). Support for adolescent mothers during the postnatal period: Theory and research. In Boukysis, C. (ed.), Research on Support for Parents and Infants in the Postnatal Period. Norwood, NJ: Ablex. Daniel, J., Hampton, R., & Newberger, E. (1983). Child abuse and accidents in black families. American Journal of Orthopsychiatry, 53, 645–53. Deutsch, F.M., Ruble, D.N., Fleming, A., Brooks-Gunn, J., & Stangor, C. (1988). Information-seeking and maternal self-definition during the transition to motherhood. Journal of Personality and Social Psychology, 55, 420–31. Forrest, J. D. (1994). Epidemiology of unintended pregnancy and contraceptive use. American Journal of Obstetrics and Gynecology, 170, 1485–89. Forrest, J.D., & Singh, S. (1990). The sexual and reproductive behavior of American women 1982–1988. Family Planning Perspectives, 22, 206–14. Forssman, H., & Thuwe, I. (1981). Continued follow-up study of 120 persons born after refusal of application for therapeutic abortion. Acta Psychiatr Scand, 64, 142–49. Forssman, H., & Thuwe, I. (1966). One hundred and twenty persons born after application for therapeutic abortion refused. Acta Psychiatr Scand, 42, 71–88. Gershenson, H.P., Musick, J.S., Ruch-Ross, H.S., Magee, V., Rubino, K.K., & Rosenberg, D. 1989. The prevalence of coercive sexual experience among teenage mothers. Journal of Interpersonal Violence, 4, 204–19. Goldberg, W.A., & Easterbrooks, M.A. (1988). Maternal employment when children are toddlers and kindergartners. In Gottfried, A.E., & Gottfried, A.W. (eds.), Maternal Employment and Children’s Development: Longitudinal Research. New York, NY: Plenum. Gomby, D.S., Larson, C.S., Levit, E., & Behrman, R.E. (1993). Home Visiting: Analysis and Recommendations. The Future of Children, 3, 6–22. House, J.S., Umberson, D., & Landis, K.R. (1988). Structures and processes of social support. Annual Review of Sociology, 14, 293–318. Kendrick, J.S., Gargiullo, P.M., Williams, L.B., & Bruce, F.C. (1990). Unintended births and the risk for low birthweight: Data from the 1988 National Survey of Family Growth. Paper presented at the 1990 annual meeting of the American Public Health Association. Matejcek, Z., Dytrych, Z., & Schuller V. (1978). Children from unwanted pregnancies. Acta Psychiatr Scand, 57, 67–90.

Preparing for Parenthood

79

McLoyd, V.C. (1990). The impact of economic hardship on black families and children: psychological distress, parenting and socioemotional development. Child Development, 61, 311–46. McLoyd, V., Jayrante, T., Ceballo, R., & Borquez, J. (1994). Unemployment and work interruption among African-American single mothers: Effects on parenting adolescent socioemotional functioning. Child Development, 65, 562–89. McCubbin, H.I. et al. (1983). CHIP-Coping Health Inventory for Parents: An assessment of parental coping patterns in the care of the chronically ill child. Journal of Marriage and the Family, 5, 359–70. Morgan, S.P. (1996). Characteristic features of modern American fertility. Population and Development Review, 22 supplement, 19–63. Musick, J.S. (1993). Young, Poor and Pregnant: The Psychology of Teenage Motherhood. New Haven, CT: Yale University Press. Myhrman, A. (1988). Family relation and social competence of children unwanted at birth. Acta Psychiatr Scan, 77, 187–91. Newburger, E., Reed, R.B., Daniel, J.H., Hyde, J.N., & Kotelchuck, M. (1977). Pediatric social illness: Toward an etiologic classification. Pediatrics, 60, 178–85 Oates, R.K., Forest, D., & Peacock, A. (1985). Self-esteem and abused children. Child Abuse and Neglect, 9, 159–63. Olds, D., & Kitzman, H. (1993). Review of research on home visiting for pregnant women and parents with young children. Future of Children, 3, 53–92. Patterson, G. (1986). Performance models for antisocial boys. American Psychologist, 41, 432–44. Patterson, G. (1988). Stress: A change agent for family process. In Garmezy, N., & Rutter, M. (eds.), Stress, Coping and Development in Children. Baltimore MD: Johns Hopkins University Press. Polansky, N.A., Ammons, P.W., & Gaudin, J.M. (1985a). Loneliness and isolation in child neglect. Social-Casework, 66, 38–47. Polansky, N.A., Gaudin, J.M., Ammons, P.W., & Davis, K.B. (1985b). The psychological ecology of the neglected mother. Child Abuse and Neglect, 9, 265–75. Reichman, N., & McLanahan, S. (1997). Self-sufficiency, parenting adequacy, and child wellbeing: Lessons from New Chance, the Teenage Parent Demonstration Project, and LEAP. Paper presented at the Conference on Synthesizing the Results of Demonstration Programs for Teen Mothers, November 1997, Chicago. Ruble, D.N., Brooks-Gunn, J., Fleming, A.S., Fitzmaurice, G., Stangor, C., & Deutsch, F. (1990). Transition to motherhood and the self: Measurement, stability and change. Journal of Perspectives of Social Psychology, 58(3), 450–463. Simmons R.L., Whitbeck, L.B., Conger, R.D., & Chyi-In, W. (1991). Intergenerational transmission of harsh parenting. Developmental Psychology, 27, 159–71. Stack, C. (1974). All Our Kin. New York: Harper and Row. Stevens, J. (1984). Black grandmothers’ and black adolescent mothers’ knowledge about parenting. Developmental Psychology, 20, 1017–25. St. Pierre, R.G., Layzer, J.I., & Barnes, H.V. (1993). Two-generation programs: Design, cost and short-term effectiveness. The Future of Children, 5, 76–93. Taylor, R., & Roberts, D. (1995). Kinship support and maternal adolescent wellbeing in economically disadvantaged African-American families. Child Development, 66, 1585–97.

80

Gager, McLanahan, Glei

Umberson, D., & Gove, W.R. (1986). Parenting status and psychological well-being: social integration, social support and stress. Paper presented at American Sociological Association, 1986. Wasik, B.H., Bryant, D., Lyons, C., Sparling, J.J., & Ramey, C.T. (1997). The LBW, premature infant. In Helping Low Birthweight, Premature Babies. Stanford, CA: Stanford University Press. Wakschlag, L., Chase-Landsdale, L., & Brooks-Gunn, J. (in press). Not just ghosts in the nursery: Contemporaneous intergenerational relationships and parenting in young African-American families. Child Development. Webster-Stratton, C. (1997). From parent training to community building. Families in Society, 78, 156–71. Wilson J.B., Ellwood, D.T., & Brooks-Gunn, J. (1997). Welfare-to-work through the eyes of children. In Escaping from Poverty: What Makes a Difference for Children? Cambridge: Cambridge University Press. Wilson, W.J. (1987). The Truly Disadvantaged. Chicago: University of Chicago Press. Zuravin, S. (1991). Unplanned childbearing and family size: Their relationship to child neglect and abuse. Family Planning Perspectives, 23, 155–61.

part

II

CHILD-REARING PRACTICES

4 Meeting the Challenges of New Parenthood: Responsibilities, Advice, and Perceptions Allison Sidle Fuligni and Jeanne Brooks-Gunn

How do American parents respond to the changes involved in becoming a new parent? In this chapter, we consider three main questions: How do parents divide the responsibilities of daily care for the very young child? What different sources do parents use to obtain information and advice about parenting? and How well do parents feel they are handling their parenting responsibilities? All parents who participated in the Commonwealth Fund Survey of Parents with Young Children had a child under the age of three, so all had recently experienced the transition to being a new parent. The data from the Commonwealth Fund survey offer some of the only nationally representative data on these issues and provide a picture of child rearing in American families today.

the challenge of new parenthood The transition to parenthood is a life-changing event. New parents are suddenly responsible for the care of a virtually helpless infant, who requires feeding, changing, and soothing 24 hours a day. Caring for a very young child is physically demanding, emotionally intense, and continuous; the transition to being a new parent is therefore both physically and emotionally challenging (Antonucci and Mikus 1998). Lack of sleep alone, part of the lives of all new parents, and often a new experience for them, may affect parents’ perceptions of how well they are doing in their new roles and the help or advice they feel they need. Becoming a parent for the first time involves reorganizing perceived roles and relationships for mothers, fathers, and the couple as a unit (Antonucci and Mikus 1998; Cowan and Cowan 1988). It tends to involve more negative than positive life changes, at least early in the life of the new child (Cowan and Cowan 1988). The transition that occurs with a secondor later-born child may affect parents less because they have already experienced some of the drastic changes that becoming a parent entails and 83

84

Fuligni, Brooks-Gunn

negotiated some of the role changes – from husband or wife or single person to father or mother. Even so, expectations about what the new baby will be like or concerns about whether parents will be able to love a second child as much as the first can make later transitions to parenthood challenging as well (Goldberg and Michaels 1988). The Division of Parenting Responsibilities The new role of parent requires the restructuring of family life among married couples, as they negotiate which partner will take on which of the new responsibilities. Much research over the last 30 years has documented a decline in marital satisfaction following the birth of a first child, which is associated with new parents’ level of satisfaction with the division of household responsibilities (Cowan and Cowan 1988) but not the division of actual child care responsibilities (Ruble et al. 1988). This decline is relatively small, however, and marital satisfaction (along with characteristics such as emotional well-being, depression, and self-esteem) remains highly correlated across the transition to parenthood, illustrating that couples and individuals who were adapting well before the birth of a child continue to fare better than those who had troubled marriages or other problems before becoming parents (Heinecke 1995). The conceptualization of the role of fathers in families has evolved in recent years. Rather than primarily being responsible for material support of the family, fathers are now considered to have an important role to play in the daily lives of their children through direct interaction, accessibility to the child, and responsibility for caring for the child (Lamb et al. 1987). Across a number of studies, data suggest that the amount of time fathers spend engaged with their children has been increasing since as early as 1924, and certainly from the 1960s to the present (Pleck 1997). Even so, and despite an ideal and an actual increase in the engagement of fathers in their children’s lives, mothers typically provide the majority of daily care for young children. Researchers in the 1970s and 1980s reported that fathers’ engagement with children was about a third that of mothers (Lamb et al. 1985; Goldberg et al. 1985); in the 1990s, fathers’ engagement rose to about 43 percent of mothers’ (Pleck 1997). Pleck’s (1997) review of paternal involvement studies conducted in the 1980s and 1990s summarizes child and father characteristics that are related to levels of paternal involvement. Generally, studies find that fathers spend more time with sons than with daughters, although this gender difference is not documented for very young children. Fathers’ involvement declines as children get older, but their involvement relative to mothers’ actually increases over time. Some studies find that fathers’ proportional share of interactions with children is higher in families with more children, and fathers tend to be more involved with their firstborn children.

Challenges of New Parenthood

85

The relationship between fathers’ characteristics of race, education, occupation, and income and their involvement with young children is not straightforward. Some characteristics of mothers are related to levels of father involvement – in particular, fathers take on a larger share of involvement with their children when mothers are employed outside the home – but no relationship has been found between mothers’ educational attainment and fathers’ involvement. The consequences of having an involved father (when his involvement is positive rather than punitive) can be great for the child. When fathers perform 40 percent or more of the total family child care, preschool children exhibit positive outcomes in cognitive and social-emotional domains (Radin 1994). In addition, paternal involvement with a new baby can be beneficial for the father; for example, it is associated with better adjustment to parenthood and more positive marital outcomes (Goldberg et al. 1985). Mothers who are responsible for the vast majority of the care of the new baby report more marital difficulties (Goldberg et al. 1985). When couples share equally in the care of their children, there are positive consequences for the family, including more successful parent-child relationships and enhanced child development, supported by the stability of commitment from both parents (Biller 1993). Of course, child rearing occurs in all types of families, not only those with two parents living together. Single parents must also fit the care of a young child into their lives, which may include care of other children, employment outside the home, and many other responsibilities. Single and married parents may spend large or small proportions of their time actually caring for their young children, and they may be satisfied or dissatisfied with the amount of time they devote to child rearing. Therefore, as an additional indicator of how new parents are weathering the transition to parenthood, we analyze parents’ reports of their satisfaction with the amount of time they spend with their young child. We look at the relationship between shared caregiving (responsibility shared equally by mothers and fathers) and satisfaction for two-parent families, as well as other predictors of satisfaction for one- and two-parent families. Sources of Parenting Information, Support, and Advice The transition to parenthood generally begins with the decision to become pregnant or the discovery of pregnancy, and the first challenges are associated with adapting to the pregnancy. It is common for expectant parents to seek information about what to expect from pregnancy and childbirth, either from professional sources (such as childbirth classes or a doctor or nurse) or from informal sources (such as kin or non-kin members of their personal support networks).

86

Fuligni, Brooks-Gunn

In addition, expectant mothers and fathers may feel they are unprepared in the skills required of new parents, such as feeding, bathing, changing, and soothing a newborn baby (Gottlieb and Pancer 1988). Parents of young children, no matter how experienced or advantaged they are, can always use some form of help, support, or advice. Advice may be obtained informally, by observing and talking to family members, friends, and community members, or it may be obtained through more formal information channels. Information on parenting is available in many forms, including pamphlets, books, magazines, and parenting classes. The social support networks of new parents may include friends and community members living or working nearby and family members, either nearby or accessible by phone or written communication. These sources of support may help new parents with specific parenting needs or with the physical and psychological needs associated with the transition to parenthood, such as the mother’s physical post-partum symptoms, depression, fatigue, need for social interaction, and need for time away from the new baby. Gottlieb and Pancer (1988) describe four types of support that may be provided by an individual’s support network: (1) emotional support, such as expressions of closeness; (2) cognitive guidance, including advice and information given either formally or obtained informally through observation and social comparison; (3) tangible aid, such as financial assistance or other practical aid; and (4) coherence support, which helps recipients assess the meaning of the transition to parenthood for their ongoing stability and relationships. The most influential of these are cognitive guidance and tangible aid, which serve as a referral system for new parents and can have a positive influence on the use of prenatal medical services and reduce stressful experiences with infants. Lack of support is associated with higher rates of post-partum depression. Support from both within and outside the household is associated with development of a secure infantmother attachment bond. Although formal sources of parenting information and support may seem to offer the most authoritative or “expert” information, many parents rely on informal sources of advice. In particular, the mother’s own parents may be more influential in affecting her parenting attitudes and practices than professional advice, television and films, books, parent education classes, or personal experience (Cook and Ballenski 1981). Among black parents, grandmothers are universally cited as a primary source of parenting support (Hunter 1997). However, medical and parenting professionals do make other sources of parenting information available. Formal sources of parenting information can provide new parents with the most up-to-date information about child development and parenting practices that enhance development. Seeking such information can positively affect children, as well as parents. Maternal knowledge about

Challenges of New Parenthood

87

normal child development is associated with the quality of the child’s home environment, behavior, and IQ at age 3 (Benasich and Brooks-Gunn 1996). Research comparing groups of women who were not pregnant, currently pregnant, and post-partum has suggested that the transition through these stages affects individuals’ self-perceptions and styles of seeking information about parenting (Deutsch 1988). Pregnant women used parenting information to help construct views of themselves as mothers, and postpartum women used their own experiences in caring for their child to further develop their parental self-perceptions. Perceptions of Parenting As women and men negotiate the transition to parenthood, they experience fluctuations in their perceptions of how well they are doing in their new role. Parents may feel overall confidence in their abilities as parents that differs from their day-to-day experiences of frustration or aggravation when caring for their children. Levels of coping, confidence, and frustration may be affected by issues such as whether or not parents share the daily responsibilities of child care and their patterns of seeking information and advice on parenting. For instance, receiving support from peers, whether in a support group or through one’s informal support network, can help new parents view their experiences as more normal, which in turn affects their self-perceptions (Gottlieb and Pancer 1988).

the commonwealth survey sample Parenting Variables Questions from the Commonwealth Survey of Parents with Young Children correspond to a number of child-rearing variables, including division of child-care responsibilities in the household; types of parenting information provided to new parents by the hospital at the time of the child’s birth; other sources of advice, such as media sources and parenting classes; parents’ satisfaction with the amount of time they spend with their child; and confidence, coping, and frustration. Regarding the division of child care responsibilities, respondents living in two-parent families were asked, “Do you and your partner share the daily care for your child equally, do you usually provide more of his/her daily care, or does your partner usually provide more of his/her daily care?” Parents in both two-parent and single-parent households were asked, “Do you feel you spend about the right amount of time, or would you like to spend a lot more, a little more, a little less, or a lot less time with your child?”

88

Fuligni, Brooks-Gunn

Questions about sources of parenting information and advice include a general query about whether or not the parent received information about newborn care from hospital staff at the time of the child’s birth and specific inquiries about receiving information at the hospital about how to bathe the baby, breastfeed or bottle feed the baby, calm the baby, or position the baby for sleep. Parents were also asked to name one or two people with whom they can comfortably talk about how to raise their child; their use of books, magazines, newspapers, television, or videos to get information about how to raise their child; and their attendance at a class or a discussion about child rearing or parenting. Parents’ perceptions of themselves as parents are explored in questions asking them to rate how well they feel they are coping with the demands of parenthood and how confident they feel about caring for the child now. The survey also asked, “In a typical day, how many times would you say you feel frustrated or aggravated with your child’s behavior or that he/she gets on your nerves?” Independent Variables A strength of the Commonwealth Survey sample is its inclusion of both mothers and fathers, allowing for comparison of child-rearing practices and attitudes by gender.1 We also consider other demographic factors, such as the age of the parent, education, race/ethnicity, and income, which have been found to be associated with parenting attitudes and practices (see, for example, Brooks-Gunn et al. 1999; Brooks-Gunn and Chase-Lansdale 1995; Brooks-Gunn et al. 1996; Garcia-Coll and Vazquez Garcia 1995; McLoyd 1995). Regarding age of the parent, respondents are divided into those who were very young (under age 21) at the time of the child’s birth and those who were 21 years of age or older. Education is broken down into two groups: parents who have a high school education or less, and those who have some education beyond high school. Total family income for 1994 is grouped into three categories: low income (less than $20,000), middle income ($20,000 to less than $60,000), and high income ($60,000 or more). Where appropriate, we also consider whether the child’s age or gender are related to parenting patterns. Relationships between the parenting variables and demographic factors are first explored individually, in bivariate analyses (chi-squares and one-way analyses of variance), then in multivariate analyses that consider a range of demographic predictors.

1 Due to the lower response rate of fathers, the sample may not be fully representative of all new fathers nationally. The findings reported here are suggestive of possible differences between mothers and fathers, but caution should be used when generalizing to all fathers.

Challenges of New Parenthood

89

division of child-rearing responsibilities Parents answered two questions about the time they spend in caring for their child, one having to do with how parents in two-parent households divide responsibility for child care between them, and the other having to do with parents’ satisfaction with the time they spend with their child.

Parents’ Division of Daily Care The Commonwealth Survey asked parents to answer the question, “Do you and your partner share the daily care for your child equally, do you usually provide more of his/her daily care, or does your partner usually provide more of his/her daily care?” This question was asked of two-parent families only, so the discussion considers only their responses. Overall, 36 percent of survey respondents in two-parent households say they share responsibility for child care equally with their spouse or partner. When the sample is split according to the gender of the respondent, mothers and fathers seem to be in agreement: 37 percent of fathers and 36 percent of mothers say they share the daily care of their child about equally. Of the mothers, 62 percent claim they have primary responsibility and 2 percent say their spouse does. Consistent with those figures, 3 percent of fathers report providing the majority of care, and 60 percent say their spouse does. These figures support the common belief that mothers are the primary caretakers of very young children, with 61 percent of the sample reporting that the mother is the primary caretaker. However, fathers are not completely uninvolved: Overall, 2.6 percent of fathers in two-parent families are the primary providers of child care, and 36 percent of fathers share the job of caring for their child equally with their partners. An interesting relationship exists between the age of the respondent and the way caregiving duties are assigned (Table 4.1). Parents who were under age 21 when the child was born are most likely to say they share caregiving duties equally with their partner. Among parents who were 21 or older when the child was born, the mother is most likely to be the primary caregiver. Lower levels of education are also associated with higher levels of sharing child care responsibilities – not a surprising result, since younger parents are more likely to have completed fewer years of schooling. Parents are more likely to share caregiving responsibility when the child is the firstborn in the family. This finding may reflect a tendency among fathers to devote more time to older children upon the birth of later-born children. In other words, although fathers may not share equally in the caregiving of the new arrival, they may still be contributing to the family’s child care. Thus, even if a father provides a substantial portion of childcare in the family, the firstborn child may be most likely to experience his involvement, both at birth and when siblings arrive.

Fuligni, Brooks-Gunn

90

table 4.1. Division of Child Care Responsibilities in Two-Parent Families (%) Mother does more

Shared equally

By child’s agea 0–11 months 12–23 months 24–36 months

62 62 59

35 35 38

3 3 3

By child’s gender a Son Daughter

60 62

37 36

3 2

By child’s birth order∗∗∗∗ Firstborn Later-born

56 65

42 32

2 3

By parent’s age at child’s birth∗ Younger than 21 years of age 21 years of age or older

47 62

51 35

2 3

By parent’s educational attainment∗ High school or less More than high school

57 65

40 33

3 2

By family employment status∗∗∗ Both full-time One full-time, one part-time One full-time, one not employed Both not employed

38 65 82 41

60 32 16 49

2 3 2 10

Note: 0 –11 month olds N = 457. 12–23 month olds N = 506. 24–36 month olds N = 444. Sons N = 727. Daughters N = 692. Firstborn N = 535. Later-born N = 850. Parents younger than 21 years N = 81. Parents 21 years and older N = 1330. High school education or less N = 649. More than high school N = 764. Both full-time N = 498. One full-time, one part-time N = 275. One full-time, one not employed N = 556. Both not employed N = 39. a ns. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001. ∗∗∗∗ p < .0001.

Father does more

Challenges of New Parenthood

91

It is not surprising that caregiving responsibilities are related to the parents’ employment. The highest rates of shared caregiving are evident in groups in which both parents have similar levels of employment: 60 percent of families with both parents employed full time and 49 percent of families with neither parent employed report sharing caregiving equally. In families with one parent employed full time and one employed part time or not employed, the mother is more likely to be the primary caregiver. The highest incidence of fathers as primary caregivers is found in two-parent families with both partners unemployed. These analyses reveal a pattern of parenting behaviors suggesting that higher socioeconomic status is associated with the more traditional pattern of caregiving in which the mother is the child’s primary caregiver. Yet parents’ employment patterns affect the time they have for caregiving activities, so full-time employment is often associated with mothers being less likely to be the child’s primary caregiver, especially when she is employed full time. In addition to the bivariate comparisons just described, a logistic regression analysis was conducted to determine the joint effects of family characteristics on patterns of shared caregiving of young children. Responses to the caregiving question were grouped into two categories: mother as the primary caregiver, and caregiving shared equally. The thirty-seven cases in which the father was the primary caregiver were dropped from this analysis. Variables entered into the regression model as independent variables were child age, child gender, parent employment, parent education, parent gender, parent age, race, income, and birth order of the child (firstborn or not). The regression analysis showed that parental employment, income, race, and birth order are all significantly related to patterns of shared caregiving, while controlling for all the other independent variables (Table 4.2). Families in which at least one parent is employed part time are almost six times more likely than families with both parents employed full time to have the mother as the primary caregiver. Again, this result is not surprising: When one parent is working only part time (most often the mother), it seems likely that she would assume more of the child-rearing responsibilities. Race of the family is also important: Black families were half as likely as white families to have the mother as the primary caregiver rather than sharing responsibilities equally. It is important to remember that because this analysis includes only two-parent families, single black mothers are not included here. Shared caregiving is more prevalent in low-income homes (less than $20,000 per year) than in middleincome homes ($20,000–$60,000) and in families in which the child is the firstborn.

Fuligni, Brooks-Gunn

92

table 4.2. Logistic Regression Predicting Incidence of Mother as Primary Caregiver Independent variables Child’s age Child’s gender: boy Part-time employeda Not employeda Parent’s education beyond high school Parent’s age: 21 and over Race: blackb Race: Hispanicb Firstborn Parent-gender: mother Low incomec High incomec

Odds ratio

95% CI lower

95% CI upper

0.94 0.86 5.99∗∗∗∗ 1.59 1.27

0.79 0.65 4.53 0.73 0.94

1.11 1.12 7.94 3.50 1.71

1.73 0.53∗∗ 0.73 0.75∗ 1.02 0.67∗ 0.95

0.95 0.34 0.52 0.57 0.77 0.48 0.66

3.16 0.84 1.04 1.00 1.33 1.00 1.38

Note: Model X2 (12) = 215.10, p < .0001, N = 1143. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001. ∗∗∗∗ p < .0001. a Employment categories are compared with full-time employment. b Race categories are compared with white respondents. c Income was dummy-coded, with low income ( < $20,000/year) and high income ($60,000/year or more) each contrasted with middle income ($20,000–<$60,000/year).

Parents’ Satisfaction with the Time Spent with Their Child When asked about their satisfaction with the amount of time they currently spend with their child, most parents in two-parent families say they would like more time. Sixty percent of respondents say they would like “a little more” or “a lot more” time with their children, and 37 percent are satisfied with the current amount of time they spend with their child. Feelings vary with the amount of time parents actually do spend with children: parents who are not primary caregivers are less likely to be satisfied with the amount of time they spend with their children and more likely to desire more time together (Table 4.3). Interestingly, a substantial proportion of primary caregivers (39 percent) desired more time with their children. This finding suggests that even parents who report being primarily responsible for the daily care of the child or sharing this role equally with their spouse have other responsibilities that reduce the amount of time they spend with their children below levels they consider optimal. Fathers are more likely than mothers to desire more time with the child. Seventy-four percent of fathers wanted more time, compared with 49 percent of mothers. Six percent of mothers report that they would like to spend less time with their child, whereas only one father (0.2 percent) gave this

Challenges of New Parenthood

93

table 4.3. Satisfaction with Time Spent with Child in Two-Parent Families Want more time (%)

Satisfied (%)

Want less time (%)

By caregiver status∗∗∗ Primary caregivers Equal caregiving Non-primary caregiver

39 64 81

52 35 19

9 1 0

By parent gender∗∗∗∗ Father Mother

74 49

26 45

0 6

By child gender∗∗ Son Daughter

57 63

38 35

5 2

By family employment status∗∗∗ Both full-time One full-time, one part-time One full-time, one not employed Both not employed

77 51 49 53

22 46 45 42

1 3 6 5

Note: Primary caregivers N = 509. Equal caregiving N = 517. Non-primary caregivers N = 380. Sons N = 785. Daughters N = 746. Fathers N = 658. Mothers N = 873. Both full-time N = 497. One full-time, one part-time N = 273. One full-time, one not employed N = 556. Both not employed N = 38. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001. ∗∗∗∗ p < .0001.

response. The desire to spend more time with the child is also predictably related to parents’ employment status. In two-parent families, parents employed full time are most likely to desire more time with their children (77 percent); the highest rates of desiring less time are in families with at least one unemployed parent. Parents are slightly more likely to desire more time with daughters than with sons (p < .01). In particular, more parents of daughters report wanting more time with their child, and more parents of sons say they would like less time, although the differences are not large. To assess the combined effects of family characteristics on parents’ satisfaction with the amount of time they spend with their children, a regression analysis was conducted. Predictors included in the regression model were

Fuligni, Brooks-Gunn

94

table 4.4. Linear Regressions Predicting Satisfaction with Time Spent with Childa Two-parent families Independent variable Child’s age Child’s gender: boy Part-time employmentb Unemployedb Education beyond high school Parent’s age: 21 and over Race: blackc Race: Hispanicc Firstborn Parent gender: mother Incomed

β

SE

−.00 .00 −.08 .05 − .56∗∗∗ .05 −.59∗∗∗ .16 −.13 .06 .04 .12 .00 .09 .10 .07 −.08 .06 −.51∗∗∗ .05 −.05∗ .02 F (11, 1158) = 19.68∗∗∗

Single-parent families β

SE

.01 .00 −.03 .10 −.42∗∗ .16 −.73∗∗∗ .12 −.06 .12 .07 .12 .08 .12 −.02 .15 .00 .11 −.40 .22 −.04 .04 F (11, 373) = 5.44∗∗∗

∗p

< .05. < .01. ∗∗∗ p < .001. a Satisfaction was coded with lower scores representing wanting less time with the child, and higher scores indicating a desire to spend more time with children. Thus, positive regression coefficients suggest less satisfaction for the group in question. b Employment categories are compared with full-time employment. c Race categories are compared with white respondents. d Income was an ordinal variable ranging from 1 = < $10,000 to 6 = $60,000 or more per year. ∗∗ p

child age, child gender, parental employment, parent age, parent gender, parent education, income, race, and birth order of the child. Responses to the satisfaction question were rated using a 5-point scale, with higher scores indicating stronger desire for more time with the child. After accounting for the effects of the other variables, parental employment, income, and parent gender emerge as significant predictors of satisfaction (Table 4.4). Compared with families in which both parents are employed full time, unemployed families and families with at least one member employed only part time are significantly less likely to desire more time with their children. Mothers are less likely than fathers to desire more time. Interestingly, higher levels of income are associated with lower levels of desiring more time with the child. Overall, these findings suggest a strong trend in two-parent families toward desiring more time to spend with children. Substantial numbers of parents, even those who are primary caregivers or who are not burdened with the time constraints of full-time employment, report wanting more time with children in substantial numbers. Patterns of satisfaction were similar in single-parent families. Compared with two-parent families, single-parent families were slightly less likely

Challenges of New Parenthood

95

table 4.5. Satisfaction with Time Spent with Child in Single-Parent Families

By parent gender ∗∗ Father Mother By child age∗∗ 0–11 months 12–23 months 24–36 months Parent gender × child age∗∗ First year Mother Father Second year Mother Father Third year Mother Father By employment status∗∗∗ Full-time Part-time Not employed

Want more time (%)

Satisfied (%)

Want less time (%)

87 53

13 41

0 6

43 59 64

52 37 28

5 4 8

42 75

53 25

5 0

57 100

38 0

5 0

60 89

31 11

9 0

75 52 40

23 46 50

2 2 10

Note: Fathers N = 30. Mothers N = 425. First year N = 143 (Mothers N = 139, Fathers N = 4). Second year N = 160 (Mothers N = 153, Fathers N = 7). Third year N = 151 (Mothers N = 133, Fathers N = 18). Full-time N = 178. Part-time N = 63. Not employed N = 212. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001.

to desire more time with their children (55 percent, versus 60 percent in two-parent families) and slightly more likely to desire less time (6 percent versus 4 percent), although this difference was not statistically significant (p = .06). Again, fathers were more likely than mothers to desire more time with their children (Table 4.5), a somewhat surprising finding in a group of single fathers raising their children alone, yet it is probably explained by fathers’ higher rates of full-time employment. In single-parent families, there is no association between child gender and parents’ desire to spend more time with them, but the relationship between the age of the child and parents’ satisfaction is significant.

96

Fuligni, Brooks-Gunn

An interesting pattern emerges here, with rates of desiring more time increasing for each year of the child’s life. Forty-three percent of parents with children under 12 months old wanted more time with their child, compared with 59 percent of parents of 1-year-olds and 64 percent of parents of 2-year-olds. Such a pattern may be explained by the decreasing amount of time parents actually spend with children over the first three years of life, as mothers return to employment and children enter alternative day care situations. However, an opposite trend was also found: 5 percent of single parents with children under 12 months old and 4 percent of parents of 1-year-olds say they want less time with their children, but the figure rises to 8 percent of single parents of 2-year-olds. For parents who spend large quantities of time with their children, the third year of life can be a trying time, characterized by the increasing willfulness of the “terrible twos” stage. Interestingly, this relationship between the child’s age and parent satisfaction applies only to single mothers. Single fathers almost universally desire more time with their children, especially in the second and third years of life, yet mothers exhibit increasing dissatisfaction (in both positive and negative directions) with the amount of time they spend with their children over the first three years. Employment status among single parents is associated with satisfaction in a pattern similar to that observed in two-parent families. Single parents employed full time are most likely to be dissatisfied with the amount of time they spend with their child, with 75 percent saying they want more time. Satisfaction with the amount of time spent with the child is not significantly associated with parents’ age or level of educational achievement in the bivariate comparisons. Linear regression using family characteristics to predict satisfaction among single parents indicates that employment status is the only significant predictor when child age, child gender, parent employment, education, gender, income, and race are considered together (see Table 4.4). In single-parent families, parents employed full-time are more likely to desire more time with their children than parents employed part-time or not at all. Taken together, the findings regarding daily child-care responsibilities and satisfaction with the amount of time spent with the child reveal that many parents of very young children have competing responsibilities. Although a substantial proportion of fathers share in child-rearing, it is still the norm for the mother to be the primary caregiver. Shared caregiving is most common among younger and less-educated couples. Employment seems to play a role by leaving less time for either parent to spend with children. Many parents, even those who are not employed full-time, wish they had more time to spend with their young children.

Challenges of New Parenthood

97

sources of information and advice about parenting Survey respondents answered several questions about the availability of information and their own use of specific information sources, including the hospital where their child was born, individuals, media sources, and parenting classes. Information from the Hospital Eighty-eight percent of parents say that they received information on newborn care from the hospital or birthing center at the time of the child’s birth. Specific topics included bathing a newborn (received by 85 percent of those whose hospitals provided information), breastfeeding or bottle feeding (94 percent), tips for calming a newborn (66 percent), and infants’ sleeping positions (81 percent).2 Virtually all demographic characteristics are related to the types and sources of information and advice parents receive (Table 4.6). Married parents were more likely to have received parenting information at the hospital (91 percent) than single parents (86 percent). Among full-time employed families (two-parent families with both parents employed full time, or single parents employed full time), 92 percent received information from the hospital, compared with 89 percent of families with at least one parent working part time and only 85 percent of unemployed parents. In a logistic regression (Table 4.7) determining likelihood of receiving information from the hospital about caring for a newborn, the only significant predictor (after controlling for marital status, employment status, parent education, parent age, race, and income) was whether or not the child was a firstborn. Families were more than one-and-a-half times as likely to receive information from the hospital when the child was their first. Parenting Advice from Family, Health Professionals, and Others As has likely been the case throughout history, a very common source of advice for parents is simply talking to other parents or experts. More than 97 percent of the parents surveyed report that they talk to at least one person about how to raise their children. Among the small number 2 Sleeping position is an important topic that has received much recent attention because Sudden Infant Death Syndrome (SIDS) has been associated with infants who sleep in a face-down position. Generations-old conventions (and, often, medical advice) previously suggested that face-up sleeping was risky because of the possibility of the child spitting up during sleep and inhaling or choking on the vomit. Recent public information campaigns have increased the incidence of side- and back-sleeping positions among healthy newborns, and there has been a reduction in the number of SIDS deaths.

Fuligni, Brooks-Gunn

98 table 4.6. Parents Receiving Parenting Information from Hospital % By education High school or less More than high school

88 91

By marital status Single Married

86 91

By employment status Full-time Part-time Not employed

92 89 85

of parents who reported that they do not talk to anyone, no significant patterns could be found regarding age of the parent, education, marital status, or employment status. The most commonly cited sources of advice and support are the parent’s own mother or mother-in-law (50 percent), the parent’s spouse (43 percent

table 4.7. Logistic Regression Predicting Information Given at the Hospital Independent variables Marital status: single Part-time employeda Not employeda Parent’s education beyond high school Parent’s age: 21 and over Race: blackb Race: Hispanicb Firstborn Low incomec High incomec

Odds ratio

95% CI lower

95% CI upper

0.77 0.71 0.69 1.21

0.44 0.47 0.38 0.81

1.32 1.08 1.24 1.79

1.44 1.12 0.81 1.58∗ 0.80 0.75

0.83 0.66 0.52 1.07 0.52 0.44

2.50 1.87 1.25 2.31 1.25 1.28

Note: Model X2 (10) = 22.79, p < .05, N = 1143. ∗ p < .05. a Employment categories are compared with full-time employment. b Race categories are compared with white respondents. c Income was dummy-coded, with low income (< $20,000/year) and high income ($60,000/year or more) each contrasted with middle income ($20,000–< $60,000/year).

Challenges of New Parenthood

99

50 45 40 35 30

% 25 20 15 10 5 0 Mother/mother- Other relative Spouse/Partner Child's non- Doctor/Nurse/ Minister/Priest/ in-law residential other Health Rabbi parent Professional

Friend or neighbor

Note: Other categories, chosen by fewer than 1% of respondents, were child care worker, teacher, social worker/psychologist or other counselor, and community organization. Percentages do not add to 100% because respondents could indicate up to two sources of advice.

figure 4.1 Which person are you most comfortable talking with about how to raise your children?

overall and 56 percent of parents living with a partner), another relative (17 percent), and a doctor, nurse, or other health professional (16 percent) (Figure 4.1). (Other responses were chosen by fewer than 15 percent of respondents, and few showed significant associations.) Mothers and fathers tend to choose different people. Among the top choices, mothers are more likely than fathers to say they talk to their mother or mother-in-law, another relative, or a doctor, nurse, or other health professional. Fathers are more likely than mothers to indicate that they seek advice from their spouse or partner (Table 4.8). Parents’ responses vary by age, marital status, education, and employment status. Choosing one’s mother or mother-in-law as a main source of advice about child-rearing issues is most common among parents who were under 21 when the child was born, have a high school degree or less, are single, or are not employed. Seeking advice from another relative is also more common among younger parents and single parents. Among twoparent families, seeking parenting advice from one’s spouse or partner is more common for older parents (21 and older) and for those with some education beyond a high school degree. Consulting a health professional varies only according to marital status, with single parents more likely to talk to a health professional about child-rearing issues. We are particularly interested in characteristics that predict which parents will consult health professionals for parenting advice, and so a logistic regression model was tested. When marital status, employment status, education, parent age, race, parent gender, income, and birth order are all accounted for, black parents are the only group for whom the likelihood of seeking parenting advice from a health professional is significantly

Fuligni, Brooks-Gunn

100 table 4.8. Top Four Sources of Parenting Advice Mother/ mother-in-law (%) By parent’s gender Fathers Mothers By parent’s age Younger than 21 years of age 21 years of age or older By educational attainment High school or less More than high school By employment status Full-time Part-time Not employed By marital status Single Married

Spouse/ Other Health partnera relative professional (%) (%) (%)

39 56

59 34

13 19

12 18

75 46

30 58

22 16

(16) (17)

55 44

50 61

(19) (15)

(17) (16)

50 46 64

(61) (61) (55)

17 14 25

(16) (15) (21)

65 45

— —

27 14

19 15

Note: Numbers enclosed in parentheses within a demographic category are not significantly different from one another at the .05 level. See Appendix A for results of chi-squared analyses in each category. a Rates shown are for two-parent families only.

different (Table 4.9). Black parents are about half as likely as white parents to consult a health professional for child-rearing advice. Other predictors that were marginally significant were birth order, with parents of firstborns less likely to consult health professionals; parent gender, with mothers slightly more likely than fathers to consult health professionals; and income, with parents in households earning less than $20,000 or more than $60,000 per year more likely than middle-income parents to consult a health professional. Advice and Information from Media Sources A large majority of survey respondents (71 percent) report having used media sources, such as books, magazines, newspapers, television, and videos, to obtain information about rearing their children. Although the majority of parents in all categories of age, education, gender, marital status, employment status, and birth order report consulting these sources, rates are highest among older parents, women, those with education beyond high school, married parents, employed parents, and those reporting on a firstborn child (Table 4.10). Interestingly, the age of the child was not related to

Challenges of New Parenthood

101

table 4.9. Logistic Regression Predicting Information-Seeking from Health Professionals

Independent variables Marital status: single Part-time employeda Not employeda Parent education beyond high school Parent’s age: 21 and over Race: blackb Race: Hispanicb Firstborn Parent gender: mothers Low incomec High incomec

Odds ratio

95% CI lower

95% CI upper

1.13 0.87 0.93 0.95

0.84 0.63 0.58 0.69

2.00 1.19 1.52 1.30

1.20 0.57∗ 1.21 0.73 1.38 1.43 1.46

0.73 0.37 0.85 0.54 0.99 1.00 0.97

1.98 0.89 1.73 1.00 1.91 2.06 2.22

Note: Model X2 (11) = 28.10, p < .01, N = 1544. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001. a Employment categories are compared with full-time employment. b Race categories are compared with white respondents. c Income was dummy-coded, with low income (< $20,000/year) and high income ($60,000/year or more) each contrasted with middle income ($20,000–< $60,000/year).

use of media sources of advice, with parents of children in each of the first three years of life consulting such sources at similarly high rates. A logistic regression analysis predicting parents’ likelihood of consulting various media sources of advice was conducted, including child characteristics of age, gender, and birth order, and parent characteristics of marital status, employment status, education, age, race, gender, and income as predictors. Several characteristics emerged as significant (Table 4.11). Parents of male children and firstborn children are more likely to consult media sources. Those with at least some education after high school are more than twice as likely as those with only a high school degree or less. Also, mothers are more likely than fathers to use media sources of advice, and white parents are twice as likely as black parents to use such sources. Families in which the parents are not employed are more likely than those with parents employed full-time to consult media sources, but families in the lowest income category are less likely to use those sources than middle-income families. Information from Parenting Classes Attending a class or discussion group about parenting, a less common way of obtaining information, was used by 34 percent of parents in the survey

Fuligni, Brooks-Gunn

102 table 4.10. Use of Media Sources for Parenting Advice Category

%

Parent’s age∗∗ Under 21 21 and over

64 73

Child’s agea 1st year 2nd year 3rd year

74 71 69

Parent’s gender ∗ Mother Father

73 68

Parent’s education∗∗∗ High school or less More than high school

63 81

Marital status∗∗∗ Single parent Two-parent families

63 74

Employment status∗∗∗ Full-time Part-time Not employed

72 75 62

Birth order∗∗∗ Firstborn Later-born

77 67

a ns.

∗ p < .05.

∗∗ p < .01.

∗∗∗ p < .001.

sample. Women are more likely to say they have attended a parenting class or discussion (35 percent) than men (31 percent). Parents with higher education are more likely to have attended a parenting class (40 percent) than those with a high school education or less (28 percent). There are no significant differences in rates of parenting class attendance by parent age, child age, employment, or marital status. However, significant interaction was found between marital status and education, such that married parents with some education beyond high school are significantly more likely than married parents with a high school education or less to attend a parenting class (Table 4.12). Educational attainment had no effect on parenting class attendance among single parents. A logistic regression using child and family characteristics jointly to predict attendance at parenting classes found two variables to be significant

Challenges of New Parenthood

103

table 4.11. Logistic Regression Predicting Parent Seeking Information from Media Sources

Independent variables Child’s age Child’s gender: boy Marital status: single Part-time employeda Not employeda Parent education beyond high school Parent’s age: 21 and over Race: blackb Race: Hispanicb Firstborn Parent gender: mother Low incomec High incomec

Odds ratio

95% CI lower

95% CI upper

0.94 1.34∗∗ 0.91 1.26 1.55∗ 2.36∗∗∗∗

0.81 1.06 0.62 0.96 1.03 1.81

1.09 1.70 1.32 1.64 2.33 3.08

1.44 0.51∗∗∗∗ 0.93 1.80∗∗∗∗ 1.66∗∗∗ 0.62∗∗ 1.15

0.97 0.36 0.69 1.38 1.27 0.47 0.77

2.13 0.70 1.26 2.34 2.17 0.83 1.72

Note: Model X2 (13) = 158.64, p < .0001, N = 1569. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001. ∗∗∗∗ p < .0001. a Employment categories are compared with full-time employment. b Race categories are compared with white respondents. c Income was dummy-coded, with low income (<$20,000/year) and high income ($60,000/year or more) each contrasted with middle income ($20,000– < $60,000/year).

predictors (Table 4.13). First, parents with some education beyond high school are almost twice as likely as those with less education to have attended a parenting class or discussion. Second, birth order is significant in a direction opposite to what might have been expected: Parents are less likely to have attended a parenting class when the child is their first. This finding may indicate that parents take classes or attend discussions when their parenting burden becomes greater (that is, when they have more than one child) or when their children are older (since parents of later-born children by definition have older children in the household). table 4.12. Parents Attending Parenting Classes, by Education and Marital Status (percentage of parents in each education category)

High school or less More than high school

Single-parent families

Two-parent families

36 36

25 41

Fuligni, Brooks-Gunn

104

table 4.13. Logistic Regression Predicting Attending Parenting Class

Independent variables Child’s age Child’s gender: boy Marital status : single Part-time employeda Not employeda Parent education beyond high school Parent age: 21 and over Race: blackb Race: Hispanicb Firstborn Parent’s gender: mother Low incomec High incomec

Odds ratio

95% CI lower

95% CI upper

0.95 1.16 1.21 1.08 1.05 1.93∗∗∗∗

0.83 0.94 0.85 0.85 0.70 1.52

1.09 1.45 1.72 1.38 1.56 2.47

0.85 0.95 0.97 0.77∗ 1.20 0.92 1.24

0.58 0.70 0.72 0.61 0.94 0.69 0.91

1.25 1.30 1.30 0.97 1.53 1.22 1.69

Note: Model X2 (13) = 52.22, p < .0001, N = 1567. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001. ∗∗∗∗ p < .0001. a Employment categories are compared with full-time employment. b Race categories are compared with white respondents. c Income was dummy-coded, with low income (<$20,000/year) and high income ($60,000/year or more) each contrasted with middle income ($20,000– < $60,000/year).

Overall, patterns of information-seeking and information availability for parents of young children show relationships with several indicators of socioeconomic status. Employed, educated, married, and older parents are more likely to receive information when their children are born and are often more likely to seek others sources of parenting information. Health professionals are consulted about parenting issues more frequently by highincome and low-income parents, suggesting that government-provided health care programs for low-income parents may be increasing their access to professional advice on parenting. In addition, there are variations in the type of information parents seek out. While a majority of young, single, and unemployed parents seek advice from their children’s grandmothers, only a small percentage of parents from these groups is likely to consult health professionals or attend parenting classes. In fact, only about one-third of the overall sample

Challenges of New Parenthood

105

70 Equal/Shared Caregiving

60

Primary Caregivers Non-primary Caregivers

50 40

% Responses 30 20 10 0 Very Well

About Average

Not Too Well

Not Well at All

figure 4.2 In general, how well do you feel you are coping with the demands of parenthood?

reported attending parenting classes or discussions. Such groups can be helpful for parents in all demographic groups in many ways: They serve to introduce parents to other parents who may have advice to share; they allow families to share and compare their experiences, thus normalizing parents’ perceptions of their own experiences; and they are often run by professionals who can provide up-to-date information on child rearing and development, identify signs of dysfunction, and offer referrals as necessary.

parents’ perceptions of their abilities as parents Although the job of caring for an infant or toddler is notoriously demanding, survey respondents report in general that they are coping with the challenge quite well. Overall, 56 percent say they are coping “very well” with the demands of parenthood and 43 percent say they are doing an “average” job of coping. Fewer than 1 percent report not coping well with these demands. Being active in the daily care of the child seems to play a role in how well parents feel they are coping (Figure 4.2). In two-parent families, parents who report being primary caregivers or equal partners in caring for their children are more likely to say they are coping “very well” (57 percent) than parents who are not primary or equal caregivers (54 percent). Even so, primary caregivers in two-parent families are also the only ones to report problems coping: Five said they are coping “not too well” and one said “not well at all.” Because there were so few parents who reported problems with coping, further analysis in this area was not conducted. Ninety-three percent of parents report feeling “very confident” about caring for their child, and an additional 6 percent say they are “somewhat

Fuligni, Brooks-Gunn

106 45

Equal/Shared Caregiving

40

Primary Caregiver

35

Non-primary Caregiver 30 25

% Responses

20 15 10 5 0 Never

Once a day

2-3 Times

4-5 Times

6 or More Times

figure 4.3 In a typical day, how many times would you say you feel frustrated or aggravated with your child’s behavior or that he/she gets on your nerves?

confident,” leaving less than 1 percent of respondents who feel only “a little” or “not at all” confident. Again, no further analyses were conducted, owing to the lack of variation in responses to this question. The daily reality of caring for a young child can be demanding, and at times frustrating. Infants may cry or fuss and be difficult to soothe, and older babies and toddlers may throw tantrums, constantly get into mischief or dangerous situations, and require lots of “no, no, no’s” from their caretakers. Add the fact that caring for a young child is an aroundthe-clock job for many parents, and it is not difficult to imagine that some parents feel overwhelmed or frustrated. Parents were asked to indicate how frequently, on a daily basis, they felt frustrated by their child’s behavior, or that the child “got on their nerves.” The most common response to this question, chosen by 42 percent of respondents, was that they feel frustrated once a day (Figure 4.3). An additional 31 percent report that their young child “never” gets on their nerves, although this response is more common response among parents who are not the primary caregivers (39 percent) than among those who are (25 percent). Responses to this question were scored on a 5-point scale, with higher numbers indicating more frequent frustration, ranging from 1 for never frustrated to 5 for frustrated six or more times a day. In order to determine patterns in frequency of parental frustration, one-way analyses of variance (ANOVAs) were conducted (see Appendix A). Child gender is related to frustration, with parents of boys being frustrated more often than parents of girls. Child age is also a significant factor: For each year of the child’s life, frustration is more frequent, averaging between none and once a day in the first year of life, slightly over once a day in the second year, and between once and twice a day in the third year. Mothers also report slightly higher levels of frustration than fathers, probably because more of them are

Challenges of New Parenthood

107

primary caregivers. There is no relationship between levels of frustration and parent age, employment, education, marital status, or birth order of the child. Regression analysis was performed to predict levels of frustration using child and family characteristics jointly. Again, age and gender of the child are related to frustration, with parents of older children and parents of boys being frustrated more often. Similarly, parent gender is a significant predictor, with mothers experiencing more frustration than fathers. Finally, Hispanic race emerged as a significant negative predictor of frustration: Even after controlling for the other demographic variables, Hispanic parents report lower levels of frustration than white parents. The reasons for this are unclear, and we did not find significant effects of Hispanic ethnicity in other analyses. Garcia-Coll and Vazquez Garcia (1995) have discussed two characteristics of Hispanics that may be relevant to this finding. The first is the claim by some researchers that Hispanics tend to exhibit less self-disclosure than other groups, suggesting that frustration is not actually different but is reported differently. The second is the strong belief among Hispanics in a supernatural realm, either magical or religious, which may help them to cope with life’s difficulties by promoting a passive acceptance, suggesting that frustration is actually lower because it is defused by cultural beliefs. A correlation between depressive symptoms and levels of parenting frustration (r = .16, p < .001) led us to consider further the effect of depression (Table 4.14). Parents’ number of reported depressive symptoms was added as a predictor to the frustration regression model described. After controlling for the other factors, depression significantly adds to the prediction of frustration. The addition of depression to the model did not seem to decrease the significance of child age, child gender, parent gender, or Latino race, although it did raise the value of the coefficient for black race to a significant level (p < .05), suggesting a suppression effect between race and depression. White parents report higher levels of frustration, but have lower levels of depression than black or Latino parents. Once the positive effect of depression on frustration levels is accounted for, the lower frustration levels of black parents can be observed. Overall, the findings for parents’ reports of coping, confidence, and frustration suggest that although parents are occasionally frustrated and aggravated by their young children, they do not lack confidence in their parenting abilities nor do they feel they are having problems coping. Frustration is not systematically related to demographic characteristics of parents, but to child characteristics. Just as boys and 2-year-olds are generally considered to be more difficult behaviorally, they tend to elicit more feelings of frustration in their parents. In addition, mothers – who tend to spend more time daily with their children – experience more frustration than fathers.

Fuligni, Brooks-Gunn

108

table 4.14. Linear Regression Predicting Parental Frustration Levels Model 1: Child and family characteristics Independent variable Child’s age Child’s gender: boy Marital status: single Part-time employment a Unemployeda Parent education beyond high school Parent age: 21 and over Race: blackb Race: Hispanicb Firstborn Parent gender: female Incomec Depression

Model 2: With depressive symptoms

β

S.E.

β

.39∗∗ .09∗ −.03 .02 .06 .04

.03 .04 .07 .05 .07 .05

.38∗∗ .01∗ −.01 .02 .04 .07

−.08 −.11 −.20∗∗ .05 .24∗∗ .00

.07 .06 .05 .04 .05 .02

−.07 .07 −.13∗ .06 −.23∗∗ .05 .07 .04 .22∗∗ −.05 .01 .02 .21∗∗ .03 F(13, 1535) = 26.76∗∗

F(13, 1536) = 24.95∗∗

S.E. .03 .04 .07 .05 .07 .05

∗p

< .05. < .001. a Employment categories are compared with full-time employment. b Race categories are compared with white respondents. c Income was an ordinal variable ranging from 1 = < $10,000 to 6 = $60,000 or more per year. ∗∗ p

conclusions The data reported in this chapter provide a valuable picture of how families are dealing with the early years of parenting, in terms of how they divide daily child-rearing responsibilities, seek information and advice on parenting, and feel they are managing parenting responsibilities. The findings on how two-parent families divide caregiving suggest that, although over a third of couples share these duties equally, mothers continue to be the primary caregivers in the majority of families. Although the rate of shared caregiving rises dramatically (to 60 percent) when both parents are employed full time, a substantial portion of mothers who are employed full time are also the primary caregivers of their young children (38 percent). Rates of fathers providing the majority of child care are highest in two-parent families in which neither parent is employed. These patterns suggest somewhat higher rates of shared caregiving than other studies have done. Ruble et al. (1988) found that in a large but not representative sample of white, college-educated families, only 11.5 percent reported equal caregiving by fathers and mothers when children were 3

Challenges of New Parenthood

109

months old. None of the fathers in that sample were primary caregivers. In a national sample of two-parent families with children under 5 years, fathers reported spending an average of 2.3 hours per day caring for the needs of their young children, compared with an average of 5.9 hours per day among mothers (Aldous et al. 1998). This study, however, did not address shared caregiving directly. Both findings reflect the effect of employment patterns on families’ caregiving: The women in the Ruble sample were probably much less likely than their husbands to be working three months after the birth of a child. In the Aldous study, mothers averaged half the employment hours of fathers and almost three times the child care hours. The amount of time fathers spent caring for their children was both positively related to the work hours of their wives and negatively related to their own work hours. In families with young adolescent children, wives’ employment hours have been shown to be associated with husbands’ higher proportional share of child care duties in the family (Almeida et al. 1993). Overall, the Commonwealth Survey data suggest several observations about caregiving patterns: r Division of caregiving responsibilities in two-parent families is strongly related to parents’ employment patterns, but mothers tend to be the primary caregivers for young children, even in families with working mothers. r A substantial proportion of parents, even those who are not employed outside the home, express dissatisfaction with the amount of time they are able to spend with their children, and would like to have more time with them. r Parents with higher incomes report feeling more satisfied with the amount of time they spend with their children. These findings present a picture of family life in which mothers, along with a fair number of fathers, participate in the daily care of their young children. Many are combining child care responsibilities with employment and other demands and would like more time with their children. The findings have the following implications: r A trend toward higher rates of father involvement in caregiving may be emerging, as shown by the relationship between mothers’ employment and shared caregiving patterns and the fact that women are returning to the workforce in increasing numbers during their children’s first few years of life. Also, the finding that the highest rates of shared caregiving occur among very young parents may reflect emerging attitudes about the role of fathers in the current generation of parents and future parents. r Although we are not aware of other research that has linked income to parents’ satisfaction with the time they spend with their children, it may

110

Fuligni, Brooks-Gunn

be the case that wealthier families are able to afford services (such as house-cleaning) that enable them to have more time with their children or to spend available time more enjoyably. r Employers should be aware that many employees with children, even fathers, are going home to substantial child care responsibilities. Employers can be sensitive to the needs of working parents through a variety of policies, including flexible scheduling, time off for well-baby medical care, assistance in finding high-quality child care or parenting support and advice, and encouragement of fathers’ involvement in child rearing (see Galinsky and Bond 1998; Shore 1998). Regarding sources and types of advice and support used by parents of very young children, several findings stand out as especially important: r The majority of new parents receive some information on caring for their new baby from the hospital when the baby is born; however, parents with lower levels of education and employment and single parents are less likely to receive parenting information from the hospital. r Parents are more likely to seek advice on parenting from family members than from health professionals. Overall, only 16 percent of parents report that they turn to a doctor, nurse, or other health professional for advice on parenting issues. Black parents, regardless of income, education, employment, and marital status, are half as likely as white parents to turn to a health professional for parenting advice. r A substantial proportion of parents (71 percent) use books, magazines, newspapers, television, or videos as sources of advice on parenting issues. Use of these media sources is most common among parents of higher socioeconomic status (with higher education or higher employment rates or in two-parent families), older parents, women, and parents of firstborn children. Again, white parents are more likely than black parents to use media sources of advice. r Only about one-third of parents have attended a class or discussion group about parenting. Married parents with more than a high school education are most likely to attend a parenting group. Taken together, these findings reveal that most advice on parenting is sought in an informal way, through family members and media sources. Information from health professionals and hospitals or birth centers is most widely available to more advantaged families. The following points are indicated by those patterns: r Additional efforts should be made at the time of birth to provide parenting information to lower-income, less-educated, and single parents. These groups should be encouraged to seek information and parenting support from multiple sources.

Challenges of New Parenthood

111

r Health professionals and parenting experts should make specific efforts to encourage parents, starting in the hospital at the birth of their child, to consult with professionals on parenting issues as their children get older. They should also seek to involve extended family members in education and outreach efforts, since family members often provide advice and support for parents of young children. The importance of focusing on the interplay between social support networks and professional services for new parents has been recommended in the social support literature (Gottlieb and Pancer 1988). Furthermore, a study of the use of parenting support services by adolescent mothers has found that they think it would be helpful to have members of their informal support networks participate with them when they receive services (Crockenberg 1986). Finally, our analysis of survey findings on how well parents feel they are handling their parenting responsibilities uncovers several important findings: r The vast majority of parents feel they are coping quite well with parenting and feel confident in their parenting. r Levels of depression contribute significantly to the amount of frustration parents experience. r Parents who spend more time with their young children report feeling frustrated with them more regularly, especially parents of boys and parents of 2-year-olds. Although we did not find that income had an effect on parents’ reports of frustration, poverty can contribute to psychological stress, which may affect parenting (McLoyd 1995). Recent data from the Urban Institute’s National Survey of America’s Families indicate that 9 percent of children nationally live with a parent who feels highly aggravated with parenting. Among children in low-income families (below 200 percent of the poverty level), 14 percent have a highly aggravated parent, whereas only 6 percent of children in higher-income families have an aggravated parent. In addition, children in single-parent households are more likely (16 percent) than children in two-parent households (7 percent) to live with a highly aggravated parent (Ehrle 1999). Support programs for disadvantaged families should reflect an awareness of links between parenting practices and low income or psychological distress. Services should therefore combine social and emotional support with immediate material support (such as transportation, meals, and even financial assistance) if they are to have an impact on parenting and child developmental outcomes (McLoyd 1995). Today’s parents of young children are faced with many challenges. Those who combine employment and child-rearing responsibilities are

112

Fuligni, Brooks-Gunn

likely to feel the stress of these multiple demands. Those who are not employed and devote most of their time to caring for their young children are likely to experience frustration simply because of the tiring nature of caring for children all day. All parents, regardless of their economic status or amount of parenting experience, can benefit from advice and support, which may help to alleviate stress and frustration. Outreach efforts to provide support and advice to parents of very young children should seek to target all types of families, develop a range of media sources of information, encourage participation in support groups, and educate extended family members about parenting. The Commonwealth Survey illustrates that particular groups of parents are less likely than others to receive or seek out parenting advice from health professionals or other “authoritative” sources. Specific, targeted outreach efforts are clearly called for. In particular, parents who are very young at the time their child is born tend to rely on their mothers or mothers-in-law for parenting advice and are less likely to consult media sources of advice. Young married parents are more likely to share the care of their child equally with their spouse. In patterns similar to those of very young parents, black parents and low-income parents exhibit high levels of shared caregiving in two-parent families and relatively low levels of consulting health professionals or media sources for information on parenting. Education also emerged as an important factor in the picture of parents’ access to and use of support and information. Parents who have had some education beyond high school are more likely to receive information about parenting from the hospital, attend parenting classes, and consult media sources for parenting information. They are more likely to consult their spouses for advice and less likely to turn to their mothers. Similarly, employed parents are more likely than unemployed parents to receive information from the hospital and to consult media sources for parenting information and less likely to turn to their mothers for advice. Thus, very young parents, those with low incomes or low education, and those who are unemployed could all benefit from targeted outreach efforts to provide them with parenting information and support. Several avenues for this outreach are suggested: hospitals and birthing centers, parenting classes, and outreach to grandmothers, since they are often primary sources of advice to new parents in these groups. In the current national climate of welfare reform, in which many young, single, unemployed mothers are losing their welfare benefits, support for parents with very young children is even more critical. Mothers formerly receiving public assistance are now required to find employment, even when their infants are very young. This change has implications for the time these mothers have available to spend with the child and their psychological and physical resources when they are at home. Support

Challenges of New Parenthood

113

services will have to balance the provision of parenting support with the needs of these parents for child care and other tangible support. The health care system may represent an ideal entry point for reaching such families, as public health care assistance is generally extended even as other welfare benefits are terminated. If former welfare recipients are able to obtain highquality health care for their children, information may be provided through the doctor’s office or clinic on child development, parenting techniques, and resources for additional support.

references Aldous, J., Mulligan, G.M., & Bjarnason, T. (1998). Fathering over time: What makes the difference? Journal of Marriage and the Family, 60, 809–20. Almeida, D.M., Maggs, J.L., & Galambos, N.L. (1993). Wives’ employment hours and spousal participation in family work. Journal of Family Psychology, 7(2), 233– 44. Antonucci, T.C., & Mikus, K. (1998). The power of parenthood: Personality and attitudinal changes during the transition to parenthood. In Michaels, G.Y., & Goldberg, W.A. (eds.), The Transition to Parenthood: Current Theory and Research. Cambridge: Cambridge University Press. Benasich, A.A., & Brooks-Gunn, J. (1996). Maternal attitudes and knowledge of child-rearing: Associations with family and child outcomes. Child Development, 67, 1186–1205. Biller, H.B. (1993). Fathers and Families: Paternal Factors in Child Development. Westport, CT: Auburn House/Greenwood Publishing Group. Brooks-Gunn, J., Britto, P.R., & Brady, C. (1999). Struggling to make ends meet: Poverty and child development. In Lamb, M.E. (ed.), Parenting and Child Development in “Nontraditional” Families. Mahwah, NJ: Lawrence Erlbaum Associates. Brooks-Gunn, J., & Chase-Lansdale, L. (1995). Adolescent parenthood. In Bornstein, M. (ed.), Handbook of Parenting: Vol. 3, Status and Social Conditions of Parenting. Mahwah, NJ: Lawrence Erlbaum Associates. Brooks-Gunn, J., Klebanov, P.K., & Duncan, G. (1996). Ethnic differences in children’s intelligence test scores: Role of economic deprivation, home environment, and maternal characteristics. Child Development, 67, 396–408. Cook, A.S., & Ballenski, C.B. (1981). Perceived influences on parenting. Family Perspectives 15, 11–14. Cowan, P.A., & Cowan, C.P. (1988). Changes in marriage during the transition to parenthood: Must we blame the baby? In Michaels, G.Y. & Goldberg, W.A. (eds.), The Transition to Parenthood: Current Theory and Research. Cambridge: Cambridge University Press. Crockenberg, S.B. (1986). Professional support for adolescent mothers: Who gives it, how adolescent mothers evaluate it, what they would prefer. Infant Mental Health Journal, 7, 49–58. Deutsch, F.M., Ruble, D.N., Fleming, A., Brooks-Gunn, J., & Stangor, C. (1988). Information-seeking and self-definition during the transition to motherhood. Journal of Personality and Social Psychology, 55(3), 420–31.

114

Fuligni, Brooks-Gunn

Ehrle, J., Moore, K., & Brown, B. (1999). Adults’ environment and behavior: Parental aggravation. In Snapshots of America’s Families. Washington, D.C.: Urban Institute. Galinsky, E., & Bond, J.T. (1998). The 1998 Business Work-Life Study: A Sourcebook. New York: Families and Work Institute. Garcia-Coll, C., & Vazquez Garcia, H.A. (1995). Hispanic children and their families: On a different track from the very beginning. In Fitzgerald, H.E., Lester, M. B., & Zuckerman, B. (eds.), Children of Poverty: Research, Health, and Policy Issues. New York, NY: Garland Publishing. Goldberg, W.A., & Michaels, G.Y. (1988). The transition to parenthood: Synthesis and future directions. In Michaels, G.Y. & Goldberg, W.A. (eds.), The Transition to Parenthood: Current Theory and Research. Cambridge: Cambridge University Press. Goldberg, W.A., Michaels, G.Y, & Lamb, M.E. (1985). Husbands’ and wives’ adjustment to pregnancy and first parenthood. Journal of Family Issues, 6, 483–503. Gottlieb, B.H., & Pancer, S.M. (1988). Social networks and the transition to parenthood. In Michaels, G.Y., & Goldberg, W.A. (eds.), The Transition to Parenthood: Current Theory and Research. Cambridge: Cambridge University Press. Heinicke, C.M. (1995). Determinants of the transition to parenting. In Bornstein, M. (ed.), Handbook of Parenting: Vol. 3, Status and Social Conditions of Parenting. Mahwah, NJ: Lawrence Erlbaum Associates. Hunter, A.G. (1997). Counting on grandmothers: Black mothers’ and fathers’ reliance on grandmothers for parenting support. Journal of Family Issues, 18, 251–69. Lamb, M.E., Pleck, J.H., Charnov, E.L., & Levine, J.A. (1987). A biosocial perspective on paternal behavior and involvement. In Lancaster, J.B., Altmann, J. Rossi, A.S. & Sherrod, R.R. (eds.), Parenting across the Lifespan: Biosocial Perspectives. Hawthorne, NY: Aldine. Lamb, M.E., Pleck, J.H., & Levine, J.A. (1985). The role of the father in child development: The effects of increased paternal involvement. In Lahey, B., & Kazdin, A. (eds.), Advances in Clinical Child Psychology, vol. 8. New York, NY: Plenum. McLoyd, V.C. (1995.) “Poverty, parenting, and policy: Meeting the support needs of poor parents. In: Fitzgerald, H.E., Lester, B.M. & Zuckerman, B. (eds.), Children of Poverty: Research, Health and Policy Issues. New York, NY: Garland. Pleck, J.H. (1997). Paternal involvement: Levels, sources, and consequences. In Lamb, M.E. (ed.), The Role of the Father in Child Development, 3rd ed. New York, NY: John Wiley. Radin, N. (1994). Primary-caregiving fathers in intact families. In Gottfried, A.E., & Gottfried, A.W. (eds.), Redefining Families: Implications for Children’s Development. New York, NY: Plenum. Ruble, D.N., Fleming, A.A., Hackel, L.S., & Stangor, C.S. (1988). Changes in the marital relationship during the transition to first time motherhood: Effects of violated expectations concerning division of household labor. Journal of Personality and Social Psychology, 55(1), 78–87. Shore, R. (1998). Ahead of the Curve: Why America’s Leading Employers Are Addressing the Needs of New and Expectant Parents. New York, NY: Families and Work Institute.

[Single-parent families] Parent gender: 12.91∗∗ Parent age: .49 Child gender: 1.10 Child age: 18.35∗∗ Fathers only: 1.75 Mothers only: 15.44∗∗ Parent education: 1.40 Parent employment: 52.22∗∗∗

[Health professional] Parent gender: 10.28∗∗ Parent age: 0.01 Education: 0.48 Marital status: 4.00∗ Employment: 5.28

[Other relative] Parent gender: 12.61∗∗∗ Parent age: 5.20∗ Education: 7.27∗∗ Marital status: 44.95∗∗∗ Employment: 15.71∗∗∗

[Spouse or partner] Parent gender: 115.87∗∗∗ Parent age: 32.19∗∗∗ Education: 16.09∗∗∗ Employment: 0.57

Satisfaction with time spent with child (chi-squares): Marital status: 5.62 (p=.06)

[Two-parent families] Caregiver status: 182.53∗∗∗ Parent gender: 110.60∗∗∗ Parent age: 1.55 Child gender: 10.62∗∗ Child age: 8.34 Parent education: 1.82 Parent employment: 103.92∗∗∗

Sources of parenting advice (chi-squares): [Mother or mother-in-law] Parent gender: 50.78∗∗∗ Parent age: 74.33∗∗∗ Education: 21.45∗∗∗ Marital status: 52.29∗∗∗ Employment: 28.14∗∗∗

Division of care (chi-squares) by: Parent age: 7.70∗ Child gender: 2.45 Child age: 1.42 Parent education: 8.92∗ Firstborn child: 15.56∗∗∗ Parent employment: 239.53∗∗∗

appendix a. Chi-Square and ANOVA Results

(continued)

Parenting class attendance (chi-squares) by: Parent age: 1.08 Child age: 0.95 Parent gender: 4.60∗ Parent education: 32.97∗∗∗ Marital status: 1.22 Parent employment: 0.35 Parent education × Marital status: [Two-parent families]: 46.05∗∗∗ [Single-parent families]: 0.00

Media sources of advice (chi-squares) by: Parent age: 9.06∗∗ Child age: 3.43 Parent gender: 4.82∗ Parent education: 77.22∗∗∗ Marital status: 19.79∗∗∗ Parent employment: 15.07∗∗∗ Firstborn child: 22.04∗∗∗

2nd year: 2.13

Part-time: 1.99

More than high school: 2.03

Later-born: 1.99

Female: 2.07

Two parents: 1.98

Child age 1st year: 1.55 148.58∗∗∗

Employment Full-time: 2.02 0.22

Education High school or less: 1.96 3.38

Birth order First born: 2.00 0.07

Parent gender Male: 1.84 30.43∗∗∗

Marital status Single: 2.03 1.07

∗∗ p

< .05. < .01. ∗∗∗ p < .001. ∗∗∗∗ p < .0001.

∗p

Female: 1.92

21 and over: 1.98

Child gender Male: 2.06 12.01∗∗∗

Frustration (means): Results (F) Parent age Under 21: 2.06 1.55

appendix a. (Continued)

Not employed: 2.01

3rd year: 2.31

5 Reading, Rhymes, and Routines: American Parents and Their Young Children Pia Rebello Britto, Allison Sidle Fuligni, and Jeanne Brooks-Gunn

The past few years have witnessed a renewed interest in early child development. The Carnegie Corporation’s Starting Points (1994) and Years of Promise (1996) reports, the creation of a National Goals Panel and the Goals 2000 legislation of 1994, and President Clinton’s early childhood initiative (1997) all provide evidence of increasing awareness of the importance of early experiences. Two White House conferences have focused on early childhood development and on child care. Media attention – including an entire issue of Newsweek devoted to the early years, a prime-time television documentary on early development, and numerous television news shows, newspaper articles, and magazine pieces – have conveyed the message that what happens in children’s early years is strongly associated with their school readiness, achievement, and adolescent functioning. During the early years, children make great strides in emotional regulation and the acquisition of gross motor, fine motor, language, cognitive, and social skills. Parents and committed caregivers are the primary providers of experiences associated with those developments. For example, parents provide cognitive and linguistic experiences through activities such as looking at books, encouraging communication, and exposing children to a range of auditory and visual stimuli (Bradley 1995; Snow 1993). When parents exhibit warmth through actions such as hugging and cuddling, they influence their children’s development of relationships and emotional well-being (Barnard and Martell 1995). Finally, through regularity and consistency in daily routines, parents provide continuity and stability, conditions thought to be important to children (Boyce et al. 1983). In brief, parents provide, at the very least, three critical domains of experience for young children: cognitive and linguistic stimulation, warmth and love, and daily routines. Like the so-called three Rs of early education, these three domains – which we refer to with a bit of poetic license as “reading, rhymes, and routines” – are fundamental to early child development. 117

118

Britto, Fuligni, Brooks-Gunn

This chapter examines the frequency with which parents of children from birth to age three provide these experiences, using data from the Commonwealth Fund Survey of Parents with Young Children. Specifically, the survey asked parents about their engagement in four signal activities: shared book reading, playing, hugging and cuddling, and adherence to daily routines. The survey is one of the only nationally representative data sets to include information on parenting practices in these areas (for exceptions see the Panel Study of Income Dynamics-Child Supplement (PSIDCS), The Early Childhood Longitudinal Study-Birth Cohort (ECLS-B); Brooks-Gunn, Berlin, Leventhal, and Fuligni, 2000).

parent-child activities and child development Shared Book Reading Reading to infants is a highly interactive and complex activity (Resnick et al. 1987; Teale 1984). Some parents delay shared book reading until their children are preschoolers, since reading to an infant is not always rewarding for the parent (Bus and Van IJzendoorn 1997). Yet the benefits of early reading to infants and toddlers are evidenced in several domains of child functioning, such as language acquisition and school achievement. Although formal literacy training may not begin until a child enters school, educators today are focusing closely on “emergent literacy” and the skills children acquire before formal schooling begins. Young children acquire literacy skills not only as a result of direct instruction but also through contact with a stimulating and responsive environment (Teale 1986), where they are exposed to print, observe the functionality and uses of print, and are motivated and encouraged to engage with print. Emergent literacy consists of skills, knowledge, and attitudes that are developmental precursors to conventional forms of reading and writing (Whitehurst and Lonigan 1998). Shared book reading between parents and young children promotes language acquisition, encourages the comprehension and interpretation of texts, and teaches children how to participate in a range of discourse forms (Heath 1982; Snow 1991, 1993). A positive association has been demonstrated between being read to from infancy onward and subsequent language development. In addition, investigators from numerous disciplines agree that children who are read to regularly in early childhood enter school with a head start on reading and school achievement (Adams 1990; Bus et al. 1995; Lonigan 1994). Shared book reading has also been cited as a possible contributor to children’s social development (Resnick et al. 1987). For instance, children may incorporate routines they learn during shared book reading, such as

Reading, Rhymes, and Routines

119

the point-and-say routine, into other activities, such as playing with their peers. Bus and van IJzendoorn (1997), through several years of research, have demonstrated that weak-affective or insecurely attached mother-andchild dyads are less inclined to share books, that the children are less interested in book reading, and that the mothers tend to spend more time disciplining their children during book reading. The influence of shared book reading on children’s social development is a relatively new area of study, and has yet to be explored fully. Playing Play is central to young children’s experiences. Several forms of parentand-child play have been studied (Uzgiris and Raeff 1995), including faceto-face interactions, social games or routines, rough-and-tumble play, and object play. Face-to-face interactions are playful in nature, show characteristics of enjoyable engagement, and combine repetitive behavior and creative facial expressions, vocalizations, and mannerisms. As children get older and move into later stages of infancy and toddlerhood, social games such as peek-a-boo and pat-a-cake become more common, as do rough-and-tumble sorts of play. In contrast, object play is characterized by less mutual engagement and greater emphasis on objects, such as toys and other materials. Hugging and Cuddling Parental warmth and emotional support are often expressed through physical expressions of touching, stroking, patting, kissing, and cuddling with their children. Parents also sometimes use touch to soothe, orient, or alert an infant (Barnard and Martell 1995). Parental warmth has been conceptualized as a specific aspect of parenting (MacDonald 1992). The dimension of warmth ranges from, at the high end, strong shows of affection, positive reinforcement, and sensitivity to the child’s needs and desires to, at the low end, rejection and hostility. Warm interactions create and maintain bonds of affection between parents and children. Children’s psychological wellbeing and relationships during the early years are also affected by warm and sensitive parenting (Maccoby and Martin 1983; Oatley and Jenkins 1996; Pettit and Bates 1989). Following Daily Routines Family routines have been defined as observable, repetitive behaviors that involve two or more family members and occur with predictable regularity in the ongoing life of the family (Boyce et al. 1983). They characterize the day-to-day and week-to-week existence of a family.

120

Britto, Fuligni, Brooks-Gunn

In addition to indicating the quality of family functioning, family routines and planning children’s daily activities are believed to be important for children’s well-being, although few studies have tested this premise. Parents, through predictable behaviors and routines, are believed to provide continuity and stability to family members, especially during periods of change, stress, and transition. For example, families who adopt a morning routine (such as eating breakfast together) and eliminate morning decisions (concerning issues such as television or dress) can transform the “getting-ready battleground” into a “congenial team effort” (Duffy 1996). In situations that involve changes in life circumstances, family routines act as effective symbols of permanence. It has been suggested that the maintenance of stable household routines may help children adapt successfully to parental divorce (Hetherington et al. 1978). In families with newborns, parents are more likely to magnify the necessity of order and the complexity inherent in attaining it. Incongruities in family routines and infant rhythmicity (that is, sleep-patterns and other cyclic biological functions) have been associated with difficulties in family adjustment and adaptation to a new baby (Sprunger et al. 1985). The predictability of daily life is an important aspect of children’s developmental experiences that may be established early in the life of the child.

the commonwealth survey data The Commonwealth Fund Survey of Parents with Young Children included questions concerning the frequency with which parents read to, play with, and hug or cuddle their children and the consistency with which they maintain daily routines. Parents were asked to say how often in the past week they had engaged in book reading, playing, and hugging or cuddling: not at all, sometimes but not daily, or daily. Because each respondent reported only on his or her own participation in each activity, the survey does not necessarily indicate the frequency of parent-child activities in two-parent families. The survey also asked questions about parents’ adherence to bedtime, naptime, and mealtime routines, and answers were coded to indicate consistency in following the three routines from day to day. After determining the percentage of parents who engaged in each of the three activities and kept to daily routines, we looked at the association between those variables and certain child and family characteristics, including the age of the child, parent’s gender, parent’s age at the birth of the child, martial status, education, employment status, income, and race or ethnicity. For parent age, we divided the respondents into two groups: those who were younger than 21 years of age at the time the child was born and those who were 21 or older at the birth of the child. For education, we

Reading, Rhymes, and Routines

121

divided the parents between those who had completed high school or less and those who had some education beyond high school. Employment status was categorized for single- and two-parent households into three groups: full time (both parents in a two-parent household employed full time, or single parents employed full time), part time (at least one parent employed part time), and not employed (both parents not employed or a single parent not employed). Family income was grouped into three categories: low income (less than $20,000), middle income ($20,000 to $59,999), and high income ($60,000 or more). Associations between activities and routines and these sociodemographic factors were initially examined using descriptive and bivariate analyses. Logistic regression analyses were then conducted to understand the combined effects of child and family characteristics on activities and routines. Finally, we looked at the additional effects of parental depression on activities and routines and at the influence of attending parenting or childbirth education classes on parenting behaviors. Parental depression was coded on a three-point scale: 2 = no depressive symptoms, 1 = one to two depressive symptoms, and 0 = three or more depressive symptoms.

frequency of parent-child activities Shared Book Reading In their responses, 18 percent of parents in the Commonwealth Survey say they have not read to their child during the previous week, 45 percent say they read at least once, and 37 percent say they read to their child daily. As expected, the frequency of parental book reading varies significantly by the age of the child (Table 5.1). About 37 percent of parents with infants (12 months of age or less) report no reading with the child in the past week. In contrast, only about 8 percent of the parents of toddlers (13–36 months of age) did not read to their child in the past week. The share of parents who engage in daily reading is only 22 percent among parents of infants, but rises to 45 percent among parents of toddlers ( p < .0001). Mothers are more likely than fathers to read to their child. Fifteen percent of mothers and 22 percent of fathers say they did not read to their child during the past week, and 41 percent of mothers and 29 percent of fathers say they read with their child daily ( p < .0001). The child’s gender does not appear to be an important factor in shared book reading: Parents are equally likely to read to their sons and daughters. Parental educational attainment is modestly associated with frequency of book reading. Daily reading is more common among parents with education higher than high school (41 percent) than among those with a high school education or less (33 percent). Additionally, parents with less

Britto, Fuligni, Brooks-Gunn

122 table 5.1. Frequency of Shared Book Reading

No reading Weekly reading Daily reading ∗∗∗∗

By age of child 12 months or less 13–24 months 25–36 months

37 8 8

41 48 47

22 44 45

By gender of parent∗∗∗∗ Mother Father

15 22

44 49

41 29

Note: Less than 12 months N = 658. 13 to 24 months N = 710. 25 months and older N = 624. Mothers N = 1314. Fathers N = 691. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001. ∗∗∗∗ p < .0001.

education are more likely not to read to their children at all (22 percent) than are parents with more education (14 percent) (p < .0001). Parents’ marital status is not associated with frequency of book reading, and no significant differences are apparent between single-parent and two-parent homes. However, for parents with lower educational achievement, marital status is associated with frequency of reading: the percentage of loweducation parents who do not read to their child at all is higher in twoparent homes (24 percent) than in single-parent homes (16 percent) (p < .05), perhaps because the other, non-interviewed parent is reading to the child in two-parent households. This difference was not noted among parents with higher educational attainment, possibly indicating that parents with more education engage in shared reading regardless of marital status. Other characteristics, such as parent’s age and employment status, are not significantly associated with frequency of book reading. Playing Eighty-three percent (N = 1,661) of parents say they play with their child daily; of those, 83 percent (N = 1,378) do so more than once a day. Very few parents – indeed, virtually none – say they have not played with their child in the past week, and 17 percent (N = 338) report playing with their child only once in the past week. Parents’ participation in play activities appears to be associated with child age (Table 5.2). Daily play is reported by a higher percentage of the parents of infants and toddlers ages 13–24 months than parents of toddlers ages 25–36 months (p < .01). Parents’ educational attainment is

Reading, Rhymes, and Routines

123

table 5.2. Frequency of Playing with Child No playing

Weekly playing

Daily playing

By age of child 12 months or less 13–24 months 25–36 months

1 1 1

14 14 21

85 85 78

By educational attainment∗∗∗∗ High school or less More than high school

1 1

20 13

79 86

∗∗

Note: 12 months or less N = 660. 13–24 months N = 710. 25–36 months N = 628. High school or less N = 1047. More than high school N = 953. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001. ∗∗∗∗ p < .0001.

also associated with engagement in play activities, with 86 percent of parents with greater than high school education reporting daily play compared with approximately 79 percent of parents with a high school degree or less ( p < .0001). No differences are associated with gender of the child, parent’s age, parent’s gender, employment, or marital status. Hugging and Cuddling Less than 1 percent of parents (N = 5) report that they did not hug or cuddle their child in the past week. A little over 11 percent (N = 229) say they had done so at least once in the past week. The rest (89 percent) state that they hug or cuddle their child daily: Of these, almost all say they do so more than once a day (N = 1,660). Bivariate analysis shows that parent’s age at the birth of the child and parent’s educational attainment are associated with reported frequency of hugging or cuddling. Parents who were 21 years of age or older at the time of the child’s birth are slightly more likely to report hugging their child daily (89 percent) than are parents who were younger than 21 at the birth of their child (82 percent) (p < .05). Parent’s educational attainment is also associated with reported warmth (hugging or cuddling) displayed toward the child: 91 percent of parents with education above high school report hugging and cuddling their child daily, compared with 85 percent of parents with a high school education or less (p < .0001). No differences were noted in association with child’s gender, parent’s gender, employment status, or household composition.

124

Britto, Fuligni, Brooks-Gunn

Following Daily Routines Parents’ answers regarding the regularity with which they follow daily mealtime, naptime, and bedtime routines were coded on a four-point scale: all three routines followed each day, routines mostly the same (two routines followed daily), routines somewhat the same (one routine followed daily), and routines change from day to day. Infants below 12 months of age were omitted from these analyses because routines such as naptime and mealtime may not apply to an infant’s schedule. Among parents of 1and 2-year-olds, 51 percent report that their child’s bedtime, naptime, and mealtime routines remain the same, 28 percent say that two of the three routines remain the same, 16 percent say that one routine is followed each day, and 5 percent say that the routines change daily. Bivariate analyses show significant effects for child age and gender (Table 5.3). Parents of children ages 12–23 months appear more likely to keep to all three daily routines (55 percent) than do parents of children ages 24–36 months (46 percent) (p < .05). Inversely, more parents report daily changes in all routines with children ages 24–36 months (6 percent) than with children ages 12–23 months (4 percent) (p < .05). Although these differences are small, they suggest age-related patterns in the maintenance of daily routines. In terms of child gender, more parents of sons report maintaining daily routines (54 percent) than do parents of daughters (47 percent) (p < .05). Parents’ age and gender are associated with the maintenance of daily routines. Daily routines are maintained by a higher percentage of fathers (55 percent) than mothers (49 percent) (p < .05). Interestingly, mothers are more likely than fathers to maintain most or some of the daily routines: 17 percent of mothers and 12 percent of fathers report maintaining one routine daily, and 29 percent of mothers and 27 percent of fathers report maintaining two daily routines. Although statistically significant, these differences are not big, and they suggest that both mothers and fathers are for the most part managing to follow daily routines. Parents 21 or older at the child’s birth are more likely to maintain daily routines (52 percent) than are younger parents (41 percent) (p < .01). Family sociodemographic characteristics associated with the maintenance of routines are marital status, educational attainment, and employment status. Daily routines are maintained by 54 percent of parents in two-parent families and 41 percent of single parents. Five percent of parents in two-parent households and 8 percent of single parents say their child’s routines change daily (p < .001). These discrepancies could be attributable to shared caregiving responsibilities, and therefore greater ease in maintaining daily routines, in two-parent families. Parents with educational attainment beyond high school are more likely to report keeping to daily routines (58 percent) than are parents with a high school

Reading, Rhymes, and Routines

125

table 5.3. Maintenance of Family Routines Changes Sometimes Mostly Same every day the same the same every day By child age∗ 12–23 months 24–36 months

4 6

14 17

27 31

55 46

By child gender ∗ Son Daughter

5 5

15 16

26 32

54 47

By parent’s gender ∗ Father Mother

6 5

12 17

27 29

55 49

By parent’s age at time of child’s birth∗∗ Younger than 21 years of age 21 years of age or older

3 6

23 14

33 28

41 52

By marital status∗∗∗ Single-parent household Two-parent household

8 5

20 13

31 28

41 54

By educational attainment∗∗∗ High school or less More than high school

6 4

19 11

31 27

44 58

By parent’s employment status∗ Full-time Part-time Unemployed

5 4 8

13 15 21

31 27 28

51 54 43

Note: 12–23 months N = 694. 24–36 months N = 576. Sons N = 695. Daughters N = 575. Fathers N = 441. Mothers N = 829. Parents younger than 21 years N = 153. Parents 21 years or older N = 1111. Single-parent N = 303. Two-parent N = 967. High school education or less N = 653. More than high school N = 612. Full-time N = 452. Part-time N = 593. Unemployed N = 156. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001.

126

Britto, Fuligni, Brooks-Gunn

degree or less (44 percent) (p < .001). Conversely, 6 percent of parents with less education and 4 percent of parents with more education say their child’s routine changes daily. Finally, maintaining daily routines is more common among parents employed full time (51 percent) or part time (54 percent) than among unemployed parents (43 percent) (p < .05).

correlates of parent and child activities To assess the combined effects of family characteristics on parent and child activities, logistic regression analyses were conducted separately for shared book reading, playing, hugging and cuddling, and following daily routines. Variables included child’s age, gender, and birth order; parent’s age, gender, marital status, educational attainment, and employment status; race/ethnicity; and family income. Shared Book Reading Two logistic regression models were built to assess the combined effects of family characteristics on parental reports of shared book reading. The first model compares parents who do not read to their child at all (N = 358) with parents who read to their child weekly or daily (N = 1,647). Child age, birth order, parent’s gender, race/ethnicity, and income emerge as significant correlates (Table 5.4). Parents are nearly eight times as likely to read to a toddler as to an infant, and one and a half times as likely to read to a firstborn as to a later-born child. Mothers are twice as likely as fathers to read to their child. Parents who are not of Hispanic origin are nearly twice as likely as Hispanic parents to read to their child. Additionally, parents in the highest income bracket ($60,000 and above) are nearly twice as likely to read to their child as parents in lower income strata. The second model examines differences between parents who read to their children daily (N = 742) and parents who read to their children weekly (N = 905). Parents who did not read at all were not included. Once again, the child’s age and birth order and the parent’s gender and income emerge as significant correlates. In this model, parent’s age also emerges as a significant correlate. Parents are twice as likely to read daily to their toddlers as to their infants, and firstborn children are twice as likely as later-born children to be read to daily. Mothers are nearly twice as likely as fathers to read to their child daily. Parents age 21 years or older at the child’s birth are one and a half times as likely to read to their child daily as younger parents. Parents from higher income strata are nearly one and a half times as likely as lower income parents to read daily with their child.

Reading, Rhymes, and Routines

127

table 5.4. Comparison of Parents Who Read to Their Child Weekly and Daily

Variables Child age: toddler Child gender: girl Birth order: firstborn Parent age 21 or older at birth of child Parent gender: mother Two-parent household Education: more than high school Part-time employed Unemployed Race/ethnicity: black Race/ethnicity: not Hispanic High income Low income

Odds ratio

95% CI lower

95% CI upper

7.88∗∗∗∗ 0.95 1.51∗∗ 0.89

5.85 0.74 1.14 0.51

10.68 1.32 2.18 1.46

2.09∗∗∗∗ 1.04 1.28

1.48 0.64 0.93

2.85 1.71 1.81

1.26 1.39 1.19 1.93∗∗∗ 1.76∗∗ 1.33

0.54 0.42 0.78 1.32 1.20 0.90

1.60 1.22 1.88 2.67 2.69 1.91

Note: Model X2 (13) = 276.23, p < .0001, N = 1598. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001. ∗∗∗∗ p < .0001.

Playing The combined effects of family sociodemographic characteristics on the frequency with which parents play with their children were also explored through two logistic regression models. The first model compares parents who engage in play activities daily (N = 1,661) and parents who engage in play activities weekly (N = 338). Parents reporting no play activities were omitted from the regression model because the group is too small. Parents’ educational achievement, race/ethnicity, and income emerge as significant correlates (Table 5.5). Parents with education above the high school level are one and a half times as likely as parents with lower educational attainment to play with their children daily. White parents are nearly twice as likely as black parents and one and a half times as likely as Hispanic parents to play with their children daily. High-income parents are nearly one and a half times as likely to play with their children daily as are parents from middle and lower-income groups. The second logistic regression model compares parents who play with their child once a day (N = 283) and parents who engage in this activity more than once a day (N = 1,378). Of the variables entered, child age, parent gender, and employment status emerge as significant correlates. Parents

Britto, Fuligni, Brooks-Gunn

128

table 5.5. Comparison of Parents Who Play with Their Child Daily and Weekly

Variable Child age: infant Child gender: girl Birth order: firstborn Parent 21 or older at birth of child Parent gender: mother Single-parent household Education: more than high school Part-time employed Full-time employed Race/ethnicity: not black Race/ethnicity: not Hispanic High income Low income

95% CI lower

95% CI upper

1.26 0.99 1.10 1.18

0.94 0.76 0.83 0.76

1.68 1.28 1.47 1.82

1.23 0.94 1.49∗∗

0.91 0.62 1.11

1.65 1.44 2.01

1.03 0.79 1.84∗ 1.42∗∗ 1.56∗∗ 0.84

0.63 0.49 1.28 1.01 1.10 0.62

1.68 1.26 2.66 2.00 2.23 1.13

Odds ratio

Note: Model X2 (13) = 43.95, p < .0001, N = 1597. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001.

are nearly one and a half times as likely to play with an infant as with a toddler more than once a day. Mothers are nearly twice as likely as fathers to play more than once a day. Parents employed full time are twice as likely as part-time employed and unemployed parents to play with their child more than once a day. Hugging and Cuddling Logistic regression analyses were conducted to examine the combined effects of family characteristics on parental reports of hugging and cuddling behavior. The regression model was also used to compare parents who reported weekly (N = 229) and daily (N = 1,776) expressions of affectionate behavior. Parents who did not hug and cuddle at all in the last week were omitted from the analysis, since the group was so small. Parents’ age, educational attainment, race or ethnicity, and income are significant correlates of reported affectionate behavior (Table 5.6). As foreshadowed in the bivariate analyses, parents 21 or older at the child’s birth report hugging and cuddling of their children more frequently than do younger parents. Parents in the older age group are nearly twice as likely

Reading, Rhymes, and Routines

129

table 5.6. Comparison of Parents Who Hug or Cuddle Their Child Daily and Weekly

Variable Child age: infant Child gender: girl Birth order: firstborn Parents 21 or older at birth of child Parent gender: mother Single-parent household Education: more than high school Part-time employed Full-time employed Race/ethnicity: not black Race/ethnicity: not Hispanic High income Low income

Odds ratio

95% CI lower

95% CI upper

0.81 0.91 0.94 1.73∗

0.57 0.67 0.66 1.08

1.15 1.25 1.32 1.32

1.40 0.90 1.48∗

0.97 0.54 1.03

2.00 1.48 2.11

0.82 0.69 1.93∗∗ 1.29 2.43∗∗ 0.80

0.46 0.40 1.25 0.86 1.53 0.54

1.46 1.21 2.98 1.94 3.87 1.18

Note: Model X2 (13) = 54.21, p < .0001, N = 1599. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001. ∗∗∗∗ p < .0001.

as younger parents to display warmth toward their child daily. Parents with education beyond high school are nearly half again as likely as less educated parents to hug and cuddle their child daily, while black parents are half as likely as other parents to hug and cuddle their child daily. Finally, parents from higher-income groups are nearly two and a half times as likely as lower-income parents to display warmth toward their child. Following Daily Routines Two logistic regression models were used to examine the combined effects of family characteristics on adherence to daily routines. The first model compared parents who maintain daily routines always or mostly (N = 1,012) and parents who maintain routines sometimes or not at all (N = 258). The second model compared parents who keep all three routines every day (N = 648) and parents who follow one of the three routines every day (N = 192). Once again, parents of infants below 12 months of age were omitted from the analyses. In the first model, child age, birth order, parent’s educational attainment, and race/ethnicity emerged as significant correlates (Table 5.7). Parents of

Britto, Fuligni, Brooks-Gunn

130

table 5.7. Comparison of Parents Who Do and Do Not Follow Daily Routines

Independent variables Child age Child gender: boys Birth order: later born Parent 21 or older at birth of child Parent gender: mothers Two-parent household Education: More than high school Full-time employed Unemployed Race/ethnicity: not black Race/ethnicity: not Hispanic High income Low income

Odds ratio

95% CI lower

95% CI upper

1.57∗∗ 1.11 1.52∗∗ 0.71

1.14 0.80 1.07 0.42

2.17 1.53 2.14 1.20

0.78 1.59∗∗ 1.59

0.53 0.97 1.11

1.14 2.61 2.28

0.75 1.25 1.80∗∗ 2.01∗∗ 1.08 0.84

0.42 0.71 1.14 1.34 0.70 0.55

1.34 2.18 2.84 3.02 1.65 1.27

Note: Model X2 (13) = 52.88, p < .0001, N = 1042. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001.

12-to-23-month-olds are nearly one and a half times as likely to keep to daily routines as parents of 24-to-36-month-olds. Parents are also one and a half times as likely to maintain daily routines if the child is not the first born – in other words, if there is more than one child in the family. Parents with education beyond high school are one and half times as likely to keep to daily routines as less-educated parents. White parents are nearly twice as likely as Hispanic or black parents to keep to the same schedule of daily routines. In the second model, which excludes parents who “mostly” keep routines or who do not keep routines at all, child age, birth order, parent’s educational attainment, and race/ethnicity emerge as significant correlates. Parents are one and a half times as likely to follow all three daily routines with their 12-to-23-month-olds as with their 24-to-36-month-olds. Also, parents are more than one and a-half times as likely to keep to consistent daily routines if the child is not the first born. Parent’s education level also appears to matter. Parents with some education beyond high school appear to be nearly twice as likely to maintain daily routines as parents with lower educational attainment. Finally, Hispanic parents are about half as likely to follow daily routines as non-Hispanic parents.

Reading, Rhymes, and Routines

131

mediators of parent-child activities The negative association between parental depression and child academic and school achievement outcomes is well documented. Depressed mothers are often unresponsive, disengaged from their children, less affectionate, and emotionally unavailable (Crnic and Greenberg 1987; Patterson 1986; Teti et al. 1990). Parental depression has also been linked to lower levels of literacy activities in the home, such as shared book reading, and lower literacy competence in children (Brody et al. 1994). In the Commonwealth Study, 59 percent of the respondents report no depressive symptoms, 32 percent report one or two symptoms, and 9 percent report from three to five symptoms. Results from logistic regression models, controlling for sociodemographic characteristics, indicate that parents with greater depressive symptoms appear to be engaging less frequently in reading and playing activities with their children. Increasing levels of depressive symptoms are associated with lower rates of daily parent-child book reading (Table 5.8). Similarly, after controlling for demographic factors, parents reporting no depressive symptoms are significantly more table 5.8. Effects of Parental Depression on Daily Shared Book Reading

Variables Child age Gender: boy Birth order: later born Parent 21 or older at birth of child Parent gender: mother Two-parent household Education: more than high school Part-time employed Unemployed Black Hispanic High income Low income Depressive symptoms

Odds ratio

95% CI lower

95% CI upper

1.54∗∗∗∗ 1.05 1.74∗∗∗∗ 1.60∗

1.32 0.84 1.36 1.07

1.79 1.32 2.22 2.40

1.87∗∗∗∗ 1.01 0.93

1.44 0.70 0.72

2.43 1.46 1.20

1.42 1.19 0.82 0.74 1.34∗ 0.84 0.75∗∗

0.92 0.92 0.59 0.54 1.01 0.62 0.63

2.18 1.54 1.14 1.02 1.79 1.15 0.91

Note: Model X2 (14) = 93.84, p < .0001, N = 1319. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001. ∗∗∗∗ p < .0001.

Britto, Fuligni, Brooks-Gunn

132

table 5.9. Effects of Parental Depression on Daily Playing with Child

Variable Child age Gender: boy Birth order: later born Parents 21 or older at birth of child Parent gender: mother Two-parent household Education: more than high school Part-time employed Unemployed Black Hispanic High income Low income Depressive symptoms

Odds ratio

95% CI lower

95% CI upper

0.68∗∗∗∗ 1.13∗ 1.50∗∗∗ 1.18

0.59 1.01 1.18 0.79

0.79 1.58 1.91 1.77

2.85∗∗∗∗ 1.00 1.29∗

2.23 0.70 1.00

3.64 1.45 1.66

1.00 1.58∗ 1.04 0.98 1.36∗ 1.21 0.77∗∗

0.78 1.02 0.74 0.73 1.03 0.90 0.65

1.28 2.44 1.45 1.32 1.82 1.63 0.93

Note: X2 (14) = 157.93, p < .0001, N = 1589. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001. ∗∗∗∗ p < .0001.

likely to play with their child daily than parents reporting a few or many depressive symptoms (Table 5.9). Parents’ mental health also appears to be associated with the maintenance of routines in children’s daily life (Table 5.10). Parents reporting no depressive symptoms are significantly more likely than parents with such symptoms to follow the same daily routines. Even after controlling for the significant effects of birth order, parent education, unemployment, and ethnicity, higher levels of depression are associated with less consistent daily routines. We also used regression analyses to examine parent information-seeking as a potential correlate of engagement in reading and playing activities. Thirty-four percent of parents in the Commonwealth Study report having attended parenting classes or a discussion group about raising children. One model investigated the effects of attendance at parenting classes or discussion groups. A second model looked at attendance at childbirth classes to assess their long-term effects on parenting behaviors, over and above later parenting class attendance. Parents who have attended some form of parenting class or discussion are one and a half times as likely to read to their child on a regular basis,

Reading, Rhymes, and Routines

133

table 5.10. Effects of Parental Depression on Following Daily Routines

Variables Gender: boy Birth order: later born Parents 21 or older at birth of child Parent gender: mother Two-parent household Education: more than high school Part-time employed Unemployed Black Hispanic High income Low income Depressive symptoms

95% CI lower

95% CI upper

0.91 0.65∗ 0.72

0.66 0.46 0.43

1.25 0.92 1.21

0.83 1.52 1.52∗

0.57 0.192 1.06

1.21 2.49 2.18

0.87 0.61∗ 0.56∗ 0.50∗∗∗ 1.05 1.19 0.74∗

0.50 0.42 0.36 0.33 0.69 0.79 0.59

1.52 0.90 0.89 0.75 1.61 1.80 0.95

Odds ratio

Note: X2 (13) = 50.84, p < .0001, N = 1042. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001.

even after controlling for the significant effects of child age, birth order, parent gender, ethnicity, and income (Table 5.11). After adding childbirth education to the model, the effect of parenting classes is no longer significant, suggesting that childbirth classes may set parents on a trajectory of positive parenting behaviors even before the child is born. A similar set of regressions was performed to differentiate parents who read daily with their child from those who read less than daily or not at all. Here, we found that attendance at parenting classes is not a significant predictor of daily book reading, although parents who attended childbirth classes are significantly more likely to read daily (Table 5.12). Also, the addition of childbirth education to the model caused income to become a nonsignificant predictor, suggesting that the relationship between high income and daily reading may be partially mediated by attendance at childbirth classes. Results regarding play activities were slightly different. Parents who attended parenting classes were more likely than others to play with their child daily. The addition of parenting class attendance to the regression reduced the regression coefficient for parent education to nonsignificance, suggesting that the effects of parenting classes may partially mediate the

2.93 0.73 1.16 0.53 1.58 0.64 0.93 0.98 0.87 0.74 0.37 1.19 1.19 — —

3.62∗∗∗∗ 0.97 1.60∗∗ 0.90 2.20∗∗∗∗ 1.07 1.30 1.36 1.48 1.15 0.53∗∗∗ 1.78∗∗ 0.75 — — 256.29∗∗∗∗ (13)

Child age Child gender: boy Birth order: later born 21 or older at child’s birth Parent gender: mother Two-parent household More than high school Part-time employed Unemployed Black Hispanic High income Low income Parenting class Childbirth education X2 (df)

Note: N = 1568. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001. ∗∗∗∗ p < .0001.

95% CI lower

Odds ratio

Variables 4.48 1.30 2.20 1.52 3.07 1.78 1.80 1.90 2.53 1.79 0.76 2.66 2.66 — —

95% CI upper

Model I: Demographics only

2.98 0.73 1.16 0.54 1.56 0.66 0.89 0.97 0.87 0.74 0.37 1.17 0.52 1.09 —

95% CI lower

Model II: Parenting class

3.69∗∗∗∗ 0.98 1.60∗∗ 0.90 2.18∗∗∗∗ 1.10 1.23 1.35 1.49 1.15 0.53∗∗∗ 1.75∗∗ 0.76 1.50∗ — 262.51∗∗∗∗ (14)

Odds ratio

table 5.11. Effects of Information-Seeking on Shared Book Reading

4.56 1.30 2.21 1.52 3.05 1.83 1.73 1.88 2.56 1.79 0.75 2.62 1.10 2.07 —

95% CI upper 3.74∗∗∗∗ 0.98 1.60∗∗ 0.87 2.17∗∗∗∗ 1.04 1.17 1.37 1.49 1.23 0.55∗∗∗ 1.66∗ 0.78 1.37 1.50∗ 268.61∗∗∗∗ (15)

Odds ratio 3.02 0.74 1.16 0.52 1.55 0.63 0.83 0.99 0.87 0.79 0.39 1.11 0.53 0.98 1.08

95% CI lower

4.64 1.31 2.21 1.47 3.03 1.74 1.63 1.91 2.56 1.93 0.79 2.49 1.13 1.90 2.08

95% CI upper

Model III: Childbirth education

1.30 0.82 1.45 1.03 1.50 0.83 0.75 0.93 0.93 0.58 0.51 1.01 0.63 — —

1.52∗∗∗∗ 1.04 1.85∗∗∗∗ 1.55∗ 1.96∗∗∗∗ 1.23 0.97 1.20 1.44 0.80 0.71∗ 1.35∗ 0.86 — — 90.14∗∗∗∗ (13)

Child age Child gender: boys Birth order: later born 21 or older at child’s birth Parent gender: mothers Two-parent household More than high school Part-time employed Unemployed Black Hispanic High income Low income Parenting class Childbirth education X2 (df)

1.77 1.30 2.37 2.31 2.57 1.81 1.26 2.23 2.23 1.12 0.97 1.81 1.18 — —

95% CI upper

Note: N = 1294. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001. ∗∗∗∗ p < .0001.

95% CI lower

Odds ratio

Variable

Model I: Demographics only

1.52∗∗∗∗ 1.04 1.87∗∗∗∗ 1.56∗ 1.95∗∗∗∗ 1.23 0.96 1.20 1.45 0.80 0.70∗ 1.35∗ 0.86 1.11 — 90.89∗∗∗∗ (14)

Odds ratio 1.31 0.83 1.46 1.04 1.49 0.83 0.74 0.93 0.93 0.58 0.51 1.01 0.63 0.87 —

95% CI lower 1.78 1.31 2.39 2.33 2.56 1.82 1.24 1.55 2.24 1.12 0.97 1.80 1.18 1.42 —

95% CI upper

Model II: Parenting class

table 5.12. Effects of Information-Seeking on Daily Shared Book Reading

1.53∗∗∗∗ 1.04 1.88∗∗∗∗ 1.52∗ 1.94∗∗∗∗ 1.21 0.93 1.21 1.46 0.84 0.72∗ 1.29 0.88 1.05 1.33∗ 95.10∗∗∗∗ (15)

Odds ratio 1.33 0.83 1.47 1.01 1.48 0.82 0.71 0.93 0.94 0.60 0.52 0.96 0.64 0.82 1.01

95% CI lower

1.79 1.30 2.41 2.28 2.55 1.79 1.20 1.56 2.27 1.18 1.00 1.72 1.21 1.35 1.75

95% CI upper

Model III: Childbirth education

0.59 1.01 1.23 0.78 2.22 0.80 1.03 0.80 1.09 0.72 0.69 1.04 0.91 — —

0.67∗∗∗∗ 1.26∗ 1.57∗∗∗ 1.17 2.86∗∗∗∗ 1.18 1.32∗ 1.03 1.69∗ 1.00 0.93 1.39∗ 1.23 — — 148.03∗∗∗∗ (13)

Child age Child gender: boy Birth order: later born 21 or older at child’s birth Parent gender: mother Two-parent household More than high school Part-time employed Unemployed Black Hispanic High income Low income Parenting class Childbirth education X2 (df)

0.78 1.58 2.00 1.75 2.69 1.74 1.70 1.32 2.64 1.41 1.25 1.86 1.67 — —

95% CI upper

Note: N = 1560. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001. ∗∗∗∗ p < .0001.

95% CI lower

Odds ratio

Variable

Model I: Demographics only

0.68 1.28∗ 1.59∗∗∗ 1.18 2.83∗∗∗∗ 1.19 1.27 1.02 1.70∗ 1.01 0.93 1.37∗ 1.23 1.33∗ — 153.09∗∗∗∗ (14)

Odds ratio

table 5.13. Effects of Information-Seeking on Daily Playing with Child

0.59 1.02 1.25 0.79 2.19 0.81 0.98 0.79 1.09 0.72 0.69 1.03 0.91 1.04 —

95% CI lower 0.78 1.60 2.04 1.77 3.66 1.76 1.64 1.31 2.65 1.41 1.25 1.83 1.67 1.70 —

95% CI upper

Model II: Parenting class

0.68∗∗∗∗ 1.28∗ 1.60∗∗∗ 1.17 2.82∗∗∗∗ 1.18 1.25 1.02 1.70∗ 1.03 0.95 1.35∗ 1.25 1.29∗ 1.13 153.91∗∗∗∗ (15)

Odds ratio 0.59 1.02 1.25 0.78 2.19 0.80 0.97 0.79 1.09 0.73 0.70 1.01 0.92 1.00 0.87

95% CI lower

0.78 1.60 2.04 1.76 3.64 1.74 1.62 1.31 2.65 1.44 1.28 1.80 1.69 1.66 1.46

95% CI upper

Model III: Childbirth education

Reading, Rhymes, and Routines

137

relationship between higher levels of education and daily playing. Adding childbirth education as a predictor did not significantly improve the logistic regression model (Table 5.13). Overall, the findings seem to link attendance at parenting and childbirth education classes and engagement in reading and playing activities. It may be that parents who elect to attend parenting and childbirth classes are also more interested than others in providing their children with growthfostering experiences, or that parenting classes increase parents’ knowledge of and engagement in stimulating activities with their children. Several intervention efforts, especially with low-income families, appear to be having short-term positive effects on maternal knowledge, attitudes, and behavior (see Fuligni and Brooks-Gunn, this volume). Since the study is cross-sectional, however, the association between parents’ attendance at classes and engagement in activities with their children may be attributable to other, unmeasured characteristics.

conclusions Activities such as reading, singing and playing, and hugging and cuddling are all opportunities for parents to stimulate cognitive, social, physical, and emotional growth in their young children. Several common threads emerge from our investigation of associations between family sociodemographic factors and the frequency with which parents engage in these activities and maintain daily routines. First, the age of the child appears to be an important determinant of what parents do with their children. Parents are far more likely to engage in reading with their toddlers than with their infants. Very few parents report reading to their infants, a pattern that has not been documented previously. Earlier work has indicated that reading to an infant is quite difficult and not very rewarding (Bus and van IJzendoorn 1997). Further, parents may put off reading until the child’s second or third year in the belief that younger children will not comprehend the activity. In terms of playing, however, a reverse age pattern was noted, with parents reporting more playful activities during infancy than during toddlerhood. Indeed, as children grow older, parents seem to be reducing the time they spend playing while increasing book reading, perhaps because parents perceive their toddlers as being ready for preschool tasks. These findings indicate a need to disseminate information that stresses the importance of reading and playing interactions throughout infancy and toddlerhood (Rubin, Bukowski, & Parker, 1998). Parents need to be better informed, for example, that shared reading can help even very young children acquire rudimentary behaviors – such as handling a book and knowing how to turn pages – that are important to emerging literacy skills.

138

Britto, Fuligni, Brooks-Gunn

Second, mothers rather than fathers appear be the parents who primarily engage in these activities. This finding was expected, given earlier work indicating that mothers assume the “lioness’s” share of child-rearing responsibilities (see Fuligni and Brooks-Gunn, this volume) and development of children’s literacy life (Pellegrini et al. 1997). Even so, the survey found relatively high involvement of fathers, with 30 percent reading and 55 percent playing daily with their children. Third, it appears that parents who were 21 or older at the child’s birth engage in more activities with their children than younger parents, who seem less likely to read daily or to display warmth toward their child. Past research has shown that teenage mothers are less responsive than non-teenage mothers to their children, have more unrealistic expectations, and provide a less stimulating verbal environment (BrooksGunn and Chase-Lansdale 1995). The Commonwealth findings add to the existing body of literature on teenage parenthood, and present contributive implications for policies and programs aimed at preventing teenage parenthood. Educational attainment also appears to be a factor in parents’ activities and routines with their children. Parents with education beyond high school are more likely to read, play, hug and cuddle, and keep to daily routines with their children. Much previous research has demonstrated the strong association between maternal education and reading and literacy activities in the home (Bradley et al. 1989; Gottfried 1984; Hart and Risley 1995; Klebanov, Brooks-Gunn, and Duncan 1994; Klebanov et al. 1998; Snow 1993). The links between children’s IQ scores and the quality of learning opportunities and emotional support in the home are stronger for children of more educated mothers than for children of mothers who have not completed high school (Brooks-Gunn et al. 1996). Mothers themselves state that education gives them a strong sense of their own ability to assist their children (Gadsden 1995). The findings have strong implications for educational policy by highlighting the connection between the education of parents and early stimulation and nurturing for young children. In addition, parents with less education deserve special attention in efforts to increase parental understanding of the importance of providing stimulating environments for young children, perhaps through GED preparation classes or programs for young parents. Another sociodemographic factor associated with parental activities is income. Parents from higher income groups report more frequent book reading, playing, and hugging and cuddling with their children. A report from the Urban Institute’s National Survey of America’s Families reports alarming figures regarding the low incidence of parent-child reading and storytelling activities among low-income families (Ehrle and Moore 1999). Using data from families with children ages one to five years, that survey found that 17 percent of all children nationally and 24 percent of children

Reading, Rhymes, and Routines

139

in families with income below 200 percent of the federal poverty level were read to fewer than three days per week. Low income has been associated with detrimental effects on children’s health, cognition, and school outcomes, especially during the early years (Brooks-Gunn and Duncan 1997). Pathways through which poverty operates include parenting behavior and the home environment (Brooks-Gunn et al. 1999). The stresses associated with low income often have a negative influence on child rearing and the creation of a stimulating home environment. Under new welfare reform legislation, most states now apply work requirements to mothers of children as young as one year old. Some states are more lenient, granting exemptions until a child’s third birthday, while others are more stringent, allowing mothers to stay at home full time only until a child is 3 months old. The entry of these mothers into the workforce will likely create additional economic hardships as a result of child care and transportation costs. Financial constraints may force some parents to enroll their children in substandard child care. The findings therefore indicate a critical need for more family-friendly policies and programs aimed at enhancing and supporting low-income parents’ engagement in stimulating activities with their young children. The final sociodemographic factor associated with activities and routines is the family’s race or ethnicity. Non-white parents in the Commonwealth sample are less likely to engage in activities such as book reading, playing, and hugging and cuddling, and less likely to maintain consistent daily routines with their young children. These findings may be explained in two ways: as true differences in parenting practices or as differences attributable to characteristics not controlled for in this study. It seems likely that the racial or ethnic differences found are at least somewhat overestimated, owing to other confounding factors. For instance, blacks are more likely to be persistently poor (or poor for longer periods of time) and to have a greater depth of poverty (Brooks-Gunn and Duncan 1997; Phillips et al. 1998). In addition, of adults with education beyond high school, more whites than African-Americans or Hispanics complete college (BrooksGunn et al. 1992). The effects of intergenerational education are also linked to child outcomes, and blacks of earlier generations have much less education than whites of today (Phillips et al. 1998). These important aspects for understanding the effects of poverty on family functioning were not measured. Further, it is important to consider that some parent-child activities may be culturally laden (Auerbach 1989). For instance, blacks may tend to have a stronger oral history tradition, which may in turn lead them to engage in fewer conventional literacy activities with their children (Heath 1983). Since racial differences in parenting practices have been shown to explain some of the racial disparity in young children’s cognitive scores (Phillips et al. 1998), it is important to consider the implications of the racial

140

Britto, Fuligni, Brooks-Gunn

differences found here when designing programs to promote literacy activities in families with very young children. The Commonwealth data offers insights into the family lives of children through age three. To extend our view of daily shared reading to children through age five, we employed national data reported by the federal government (U.S. Department of Health and Human Services, 1998). According to the Commonwealth data, approximately 25 percent of infants and 45 percent of toddlers through age 3 are read to daily by their parents. Federal data indicate that 57 percent of children ages 3 to 5 participated in shared reading with their parents. There appears to be an increase in frequency of shared book reading as children move from infancy to toddlerhood to the preschool years, with the biggest increase from infancy to toddlerhood. Similar trends apply to gender, race or ethnicity, income, maternal education, and employment (Table 5.14). Neither data set shows differences in daily patterns of reading to sons or daughters, except during infancy. White parents in both samples report the highest frequency of daily reading, and Hispanic parents the lowest: Indeed, no black or Hispanic parents report daily reading with their infants. Parents at or above the poverty level report higher frequencies of daily reading than do parents living below the poverty level. Nearly half of mothers with some education beyond high school report daily reading. For children in all age groups, unemployed mothers are slightly more likely than full-time or part-time employed mothers to read to their children daily. The home environment has been cited as a portentous influence on the development of literacy and school readiness skills (Snow et al. 1991). Parental book reading, playing, and hugging and cuddling are important activities in the home environment. Further, the potent influence of learning and literacy experiences upon child well-being appears to be the strongest during the early childhood years (Brooks-Gunn et al. in press). Parents use the environment to provide stimulation and structure the environment via their interactions with their children to facilitate children’s development (Bradley 1995). The parenting activities investigated in the Commonwealth Study occur in conjunction with each other. For instance, reading is often accompanied by cuddling and hugging, since parents tend to take toddlers onto their laps to read. Some parents also include these activities as part of their daily routine through, for example, bedtime stories. A significant association was noted in the data between singing and playing activities and warmth displayed by the parent. Parents who engage in frequent playful games and activities with their children also tend to hug and cuddle their children more often, reinforcing the view that parent and child activities occur in conjunction with one another to provide cognitive, linguistic, and affective developmental experiences for young children.

Reading, Rhymes, and Routines

141

table 5.14. Daily Shared Book Reading: Results from Two Studies National data (1996)

Commonwealth Fund data Birth to 1 yr N = 664 22

1–2 yrs N = 712 44

2–3 yrs N = 629 45

3–5 yrs

Gender Boys Girls

45 55

44 44

46 45

56 57

Race/ethnicity White (not Hispanic) Black (not Hispanic) Hispanic

20 27 24

52 36 32

51 43 32

64 44 39

Incomea At or above poverty Below poverty

21 21

50 32

49 38

61 46

Family Structure One-parent Two-parent

28 20

35 46

41 47

46 61

Maternal education Less than high school High school and more

21 21

41 47

37 54

37 49

Maternal employment Full-time Part-time Unemployed

21 23 28

49 47 48

47 57 51

54 59 59

total

57

a In

the Commonwealth study, families with income below $20,000 were considered to be below the poverty level.

The seemingly low rates of parent-child linguistic activities and provision of stable routines – especially in families with children under 1 year of age, those with lower levels of income and education, very young parents, and black and Hispanic families – suggest the need for concerted and focused efforts to increase parents’ awareness of the relationship between early home experiences and later school readiness and success. Other important figures in the lives of children may also play a role in improving children’s early experiences. Child care providers should be educated about the importance of reading to very young children and encouraged to provide such experiences for the children in their care. Child care providers and other professionals working with families should also strive to help parents feel comfortable reading with their young children. Pediatric offices represent another arena that could be exploited to

142

Britto, Fuligni, Brooks-Gunn

urge parent-child book reading. The Reach Out and Read (ROR) program, in which pediatricians provide books and encourage parents to read them with their children, has been particularly successful in increasing book reading among mothers enrolled in public assistance programs (see Zuckerman, this volume). Numerous intergenerational literacy programs, such as the Home Instruction Program for Preschool Youngsters (HIPPY) and the Federal Even Start Family Literacy Program, also aim to promote parent-child book reading in low-income families. Even so, the data reported here suggest that all parents could benefit from extra support and information. Greater understanding and encouragement could help many more parents improve the environments they provide for their young children, and therefore their children’s development. Acknowledgments We wish to thank the Commonwealth Fund and the MacArthur Network on Family and Work for their support in the writing of this Chapter. The support of the NICHD Research Network on Child and Family Well-Being and the Spencer Foundation is appreciated. We also wish to thank Neal Halfon, Kathryn McLearn, and Mark Schuster for their comments.

references Adams, M.J. (1990). Beginning to Read. Cambridge, MA: MIT Press. Auerbach, E. (1989). Toward a social-contextual approach to family literacy. Harvard Educational Review, 59(2), 165–81. Barnard, K.E., & Martell, L.K. 1995. Mothering. In Bornstein, M. (ed.), Handbook of Parenting: Vol 3. Status and Social Conditions of Parenting. Mahwah, NJ: Lawrence Erlbaum. Boyce, W.T., Jensen, E.W., James, S.A., & Peacock, J.L. (1983). The family routines inventory: Theoretical origins. Social Science Medicine, 17 (4), 193–200. Bradley, R.H. (1995). Environment and parenting. In Bornstein, M. (ed.), Handbook of Parenting, Vol. 2. Hillsdale, NJ: Erlbaum. Bradley, R.H., Caldwell, B.M., Rock, S.L., Ramey, C.T., Barnard, K.E., Gray, C., Hammond, M. A., Mitchell, S., Gottfried, A.W., Siegel, L., & Johnson, D. (1989). Home environment and cognitive development in the first three years of life: A collaborative study involving six sites and three ethnic groups in North America. Developmental Psychology, 25, 217–35. Brody, G.H., Stoneman, Z., Flor, D., McCary, C., Hastings, L., & Conyers, O. (1994). Financial resources, parent psychological functioning, parent co-caregiving, and early adolescent competence in rural two-parent African-American families. Child Development, 65, 590–605. Brooks-Gunn, J., Berlin, L.J., Leventhal, T., & Fuligni, A. (2000). Depending on the Kindness of Strangers: Current National Data Initiatives and Developmental

Reading, Rhymes, and Routines

143

Research. Special Issue, “New Directions for Child Development in the TwentyFirst Century.” Child Development, 71(1), 257–267. Brooks-Gunn, J., Britto, P.R., & Brady, C. (1999). Struggling to make ends meet: Poverty and child development. In Lamb, ME (ed.), Parenting and Child Development in “Nontraditional” Families. Mahwah, NJ: Lawrence Erlbaum. Brooks-Gunn, J., & Chase-Lansdale, L. (1995). Adolescent parenthood. In Bornstein, M. (ed.), Handbook of Parenting: Vol 3. Status and Social Conditions of Parenting. Mahwah, NJ: Lawrence Erlbaum. Brooks-Gunn, J., & Duncan, G.J. (1997). The effects of poverty on children. The Future of Children, 7 (2), 55–71. Brooks-Gunn, J., Gross, R.T., Kraemer, H.C., Spiker, D., & Shapiro, S. (1992). Enhancing the cognitive outcomes of low-birth-weight, premature infants: For whom is the intervention most effective? Pediatrics, 89(8), 1209–15. Brooks-Gunn, J., Klebanov, P.K., & Duncan, G. (1996). Ethnic differences in children’s intelligence test scores: Role of economic deprivation, home environment, and maternal characteristics. Child Development, 67, 396–408. Bus, A.G., & van IJzendoorn, M.H. (1997). Affective dimension of mother-infant picturebook reading. Journal of School Psychology, 35(1), 47–60. Bus, A.G., van IJzendoorn, M.H., & Pellegrini, A.D. (1995). Joint book reading makes success in learning to read: A meta-analysis on intergenerational transmission of literacy. Review of Educational Research, 65, 1–21. Carnegie Task Force on Learning in the Primary Grades (1996). Years of Promise: A Comprehensive Learning Strategy for America’s Children. New York: Carnegie Corporation of New York. Crnic, K., & Greenberg, M. (1987). Maternal stress, social support, and coping: influences on early mother-child relationships. In Boukydis, C. (ed.), Research on Support for Parents and Infants in the Postnatal Period. Norwood, NJ: Ablex. Duffy, R. (1996). From a parent’s perspective: Hints for morning success. Child Care Information Exchange, 112, 30–31. Ehrle, J., & Moore, K. (1999). Children’s environment and behavior: Reading and telling stories to young children. In Snapshots of America’s Families. Washington, DC: Urban Institute. Gadsden, V.L. (1995). Representations of literacy: Parents’ images in two cultural communities. In Morrow, L.M. (ed.), Family Literacy: Connections in Schools and Communities. New Brunswick, NJ: Rutgers University Press. Gottfried, A.W. (1984). Home Environment and Early Cognitive Development. New York: Academic Press. Hart, B., & Risley, T.R. (1995). Meaningful Difference in the Everyday Experience of Young American Children. Baltimore: Brookes. Heath, S.B. (1982). What no bedtime story means: Narrative skills at home and at school. Language in Society, 11, 49–76. Heath, S.B. (1983). Ways with Words: Language, Life, and Work in Communities and Classrooms. New York: Cambridge University Press. Hetherington, E.M., Cox, M., & Cox, R. (1978). The aftermath of divorce. In Stevens, J.H., & Matthews, M.J. (eds.), Mother/Child Father/Child Relationships. Washington, DC: Association for the Education of Young Children.

144

Britto, Fuligni, Brooks-Gunn

Klebanov, P.K., Brooks-Gunn, J., & Duncan, G.J. (1994). Does neighborhood and family poverty affect mothers’ parenting, mental health, and social support? Journal of Marriage and the Family, 56(2), 441–55. Klebanov, P.K., Brooks-Gunn, J., McCarton, C., & McCormick, M.C. (1998). The contribution of neighborhood and family income to developmental test scores over the first three years of life. Child Development, 69, 1420–36. Lonigan, C.J. (1994). Reading to preschoolers exposed: Is the emperor really naked? Developmental Review, 14, 303–23. Maccoby, E.E., & Martin, J. (1983). Socialization in the context of the family: Parentchild interactions. In Mussen, P.H. (ed.), Handbook of Child Psychology. New York: Wiley. MacDonald, K. (1992). Warmth as a developmental construct: An evolutionary analysis. Child Development, 63, 753–73. Newsweek. Special 2000 Edition. Your Child. Fall/Winter 2000. New York: Newsweek Inc. Oatley, K., & Jenkins, J.M. (1996). Understanding Emotions. Cambridge, MA: Blackwell. Patterson, G. (1986). Performance models for antisocial boys. American Psychologist, 41, 434–44. Pellegrini, A.D., Galda, L., & Charak, D. (1997). Bridges between home and school literacy: Social bases for early school literacy. Early Child Development and Care, 99–109. Pettit, G.S., & Bates, J.E. (1989). Family interaction patterns and children’s behavior problems from infancy to 4 years. Developmental Psychology, 25, 413–20. Phillips, M., Brooks-Gunn, J., Duncan, G.J., Klebanov, P.K., & Jencks, C. (1998). Family background, parenting practices, and the black-white test score gap. In Jencks, C., & Phillips, M. (eds.), The Black-White Test Score Gap. Washington, DC: Brookings Institute. Resnick, M.B., Roth, J., Aaron, P.M., Scott, J., Wolking, W.D., Larsen, J.J., & Packer, A.B. (1987). Mothers reading to infants: A new observational tool. The Reading Teacher, 888–94. Rubin, K.H., Bukowski, W., & Parker, J.G. (1998). Peer interactions, relationships, and groups. In Eisenberg, N. (ed.), Handbook of Child Development: Social, Emotional, and Personality Development. New York: Wiley. Snow, C.E. (1991). The theoretical basis for relationships between language and literacy in development. Journal of Research in Childhood Education, 6, 5–10. Snow, C.E. (1993). Families as social contexts for literacy development. In Daiute, C. (ed.), The Development of Literacy through Social Interaction. San Francisco: JosseyBass. Snow, C.E., Barnes, W.S., Chandler, J. Hemphill, L., & Goodman, I.F. (1991). Unfulfilled Expectations: Home and School Influences on Literacy. Cambridge: Harvard University Press. Sprunger, L.W., Boyce, W.T., & Gaines, J.A. (1985). Family-infant congruence: Routines and rhythmicity in family adaptations to a young infant. Child Development, 56(3), 564–72. Teale, W.H. (1984). Reading to young children: Its significance for literacy development. In Goelman, H., Oberg, A.A., & Smith, F. (eds.), Awakening to Literacy: Writing and Reading. Norwood, NJ: Ablex.

Reading, Rhymes, and Routines

145

Teale, W.H. (1986). Home background and young children’s literacy development. In Teale, W.H., & Sulzby, E. (eds.), Emergent Literacy: Writing and Reading. Norwood, NJ: Ablex. Teti, D., Gelfand, D., & Pompa, J. (1990). Depressed mothers’ behavioral competence with their infants: demographic and psychosocial correlates. Development and Psychopathology, 2, 259–70. Trevathen, C. (1987). Sharing makes sense. In Steele, R., & Treadgold, T. (eds.), Language topics: Essays in Honor of Michael Halliday. Amsterdam: John Benjamin Publishing. U.S. Department of Health and Human Services. (1998). Trends in the Well-Being of America’s Children and Youth. Washington, DC: Office of Assistant Secretary for Planning and Evaluation Uzgiris, I.C., & Raeff, C. (1995). Play in parent-child interactions. In Bornstein, M. (ed.), Handbook of Parenting: Vol. 4. Applied and Practical Parenting. Mahwah, NJ: Lawrence Erlbaum. Whitehurst, G.J., & Lonigan, C.J. (1998). Child development and emergent literacy. Child Development, 69, 848–72.

6 Child Discipline in the First Three Years of Life Lawrence S. Wissow

Since the start of recorded history, students of human development have tried to explain why individual humans grow up to be distinct from each other in their attitudes, emotions, and behaviors. Most explanations have, to a greater or lesser extent, given a role to children’s interactions with others – and, in particular, to how parents treat their children. Scholars from many fields have examined child-parent interactions. Human rights and child welfare advocates have focused on children’s right to protection from exploitation and interpersonal violence and explored distinctions between fair and abusive parent disciplinary practices (Gil 1970). Political theorists have viewed child discipline as a reflection of strongly held beliefs about appropriate means of fostering human development, beliefs that are in turn reflected in ideals regarding the structure and role of government (Lakoff 1996). Neuroscientists of late have described how brain growth continues actively after birth in a process that relies on input from the child’s environment (Sameroff 1997). From their earliest days, children are learning and may be laying down powerful memories – in the form of nearly indelible neural pathways – related to fear, aggression, trust, and nurturing (LeDoux 1996).

approaches to parental discipline and punishment The word “discipline” has many meanings and connotations. In its broadest sense, discipline is the collection of acts and rules that parents use to “socialize” their children, to teach them the values and normative behaviors of the society in which they will have to function as adults (Cherlin 1996). A slightly more narrow definition sees discipline as targeted toward the control of specific behaviors that parents consider to be appropriate or inappropriate. Even in this more narrow sense, however, discipline includes a “system” of explicit teaching, modeling, encouragement, and consequences designed to help children learn how to behave (Committee 146

Child Discipline in the First Three Years

147

1998). Systems of discipline can include expressions of support, love, affection, nurturing, and acceptance, which function as rewards and positive motivators for desired behavior; explaining, reasoning, and modeling, as means of teaching children what behaviors are desired; and punishments, as consequences for children’s violation of rules (Cherlin 1996). For many people, however, the word “discipline” is synonymous only with punishment, the inflicting of some sort of penalty in response to an act that is considered wrong. Punishment can take many forms. Nonphysical punishment may include taking away attention, positive regard, privileges, or possessions; isolating a child for a period of time; or asking a child to perform a task that he or she regards as difficult or undesirable (for example, washing the dishes, cleaning a room, or writing an essay). It may also include scolding, shaming, or making negative attributions about the person who is being targeted for punishment in order to inflict emotional pain. Physical punishment involves inflicting physical pain or discomfort. Spanking is perhaps the most commonly used form of physical punishment, but other forms of hitting may be involved, as may forcing someone to perform a task (such as running a certain distance or standing in place for an extended period of time) that is calculated to evoke fatigue or discomfort. Physical punishment may be accompanied by varying degrees of harsh emotional communication, as well as varying degrees of fear. The boundary between acceptable and unacceptable punishment – physical or emotional – is not well defined and has varied over time and from culture to culture (Skolnick and Skolnick 1983). One set of criteria for defining “appropriate” discipline is related to parents’ view of the essential nature of children and humanity. At one extreme, people are seen as inherently vulnerable to evil and deviancy, their behavior shaped through a balancing of reward and punishment. The strength to resist internal and external temptations toward evil is best built by learning to obey and to endure hardship, and powerful temptations can be countered only by more powerful punishments. In such a system, appropriate discipline may depend heavily on punishments that are emotionally or physically painful. Alternatively, if children are seen as innately oriented toward caretakers who are loving and caring, behavior is seen as being shaped mostly by dialogue with and copying of those who provide one’s care. In that case, discipline may focus more on nurturing, teaching, and empathetic but firm response to transgressions rather than on punishment and reward (Lakoff 1996). A second set of criteria for judging discipline relates to the outcomes that parents desire for their children, the set of behavioral and temperamental characteristics that are seen as most adaptive for success as an adult. The scope and style of discipline, to be successful, must match the future social needs of the child (Kagan 1983). To use an example developed by Jerome Kagan, one archetype of American society is the individual who is

148

Wissow

recognized as extremely competent in some relatively narrow field, whose loyalty to himself or herself transcends loyalty to others, and who will be “successful” in heterosexual relationships. Parents who see this model as a goal for their children employ disciplinary strategies that reward public, individual achievement and independence from peers more than collaborative activities. For example, parents may discourage childhood exploration in favor of the acquisition of certain skills (especially athletic prowess), or they may reward aggressiveness, and punish passivity, in both social and academic spheres. To take just one culture as a point of contrast, in many traditional Native American communities, a very different set of cultural goals leads to different standards for discipline. Interdependence with peers is valued as much as or more than individual achievement, so cooperative behavior is rewarded more than self-assertion, which may even be punished. Achievement, in the sense that it is recognized by many in Western culture, may be rewarded less than behaviors that show respect for community ties and for traditional wisdom (Yazzie 1994). Psychiatry has proposed another way of looking at the goals of development that might provide criteria for judging the appropriateness of disciplinary approaches. Stern (1985) has defined acquisition of a “sense of self” as one of the key developmental goals of early childhood. Stern defines sense of self as the awareness that we have “a single, distinct, integrated body,” that we can be “the agent of action, the experiencer of feelings, the maker of intentions,” that we have an independent ability to take in outside experience, process it, and come to new understandings. Disruptions to the sense of self appear to result in serious distress during childhood and to cause long-term difficulties with functioning later in life. Failure to develop a sense of self involves developing, instead, a sense of legitimately existing only to the extent that we please others, meet their behavioral expectations, or develop world views identical to theirs (Suransky 1982; Epstein 1995). Parents who provide positive attention only intermittently, or who intrude harshly on children’s physical and emotional exploration, risk creating a child with a “false self” that is adaptive in the short term to the parents’ demands but inadequate in an adult life that requires self-direction and an internal sense of purpose and well-being. From observing a sample of American nursery school children, Diana Baumrind (1967, 1989) envisaged three major approaches to discipline in Western culture, each associated with a different pattern of outcomes. She proposed the label “authoritative” for an approach that combines high levels of emotional support with reasoning and requests for cooperation as the main method for enforcing limits. Authoritative parents can be demanding and controlling, but they are also warm and open to their children’s feelings and encouraging of independent behavior. Baumrind observed that this style is associated with children who are the most competent, most self-reliant, and most able to control themselves.

Child Discipline in the First Three Years

149

In contrast, an “authoritarian” style features low levels of emotional support and high levels of control enforced by punishment. Authoritarian parents tend to believe in obedience as a virtue in itself and do not place a positive value on engaging in reasoning or discussion with their children over behavioral issues. In Baumrind’s initial studies, children of these parents tended to be more unhappy, withdrawn, and distrustful of others. In subsequent studies, boys from these families tended to be more hostile and oppositional, while girls were relatively less independent than their peers. A third, “permissive” approach includes high levels of support but relatively little behavioral control, by any means. Permissive parents make relatively few behavioral demands on their children, and, while they are available to their children, they may not take a particularly active role in shaping their children’s lives. In Baumrind’s studies, children of these parents seemed most immature; they were the least self-reliant, self-controlled, and oriented toward achieving. In later writing, Baumrind has pointed out that parents’ choice of one of these broad styles, and the social adaptiveness of their outcomes, depends on many factors (Baumrind 1994). Some parents may deliberately choose relatively strict and cold approaches to child-rearing specifically because they believe that being tough, guarded, and respectful of dangerous social forces are essential traits (McLoyd 1990). Parents from other cultural groups may fall entirely outside American categories as they try to foster other combinations of traits.

the debate over physical punishment A particular type of child discipline – physical punishment – has long been an object of debate within U.S. society. In the years prior to the Civil War, and then again in the early twentieth century, concern arose about curbing severe physical punishment that could lead to serious injury or even death (Straus 1994). Reports of serious injuries to children (Maurer 1979) precipitated in part the movement to ban physical punishment in American schools, an action that has proceeded on a state-by-state basis. As of 1995, twenty-seven of the fifty states had made it illegal for teachers to strike children as a means of enforcing classroom behavioral rules (Grossman et al. 1995). In the United States, in the latter part of the twentieth century, the use of severe and frequent physical punishment seems to have declined, although upwards of 60 percent of American families with children under 18 continue to report using physical punishment (Gallup Organization 1995). At a national level, however, there has been little consensus about the advisability of physical punishment of children. TheU.S.SupremeCourtruled in 1977 that physical punishment in schools violated neither children’s due process rights nor their right to protection from cruel and unusual punishment

150

Wissow

(Ingraham v. Wright, 430 US 651, 1977). More recently, a panel convened by the American Academy of Pediatrics (1996) concluded that spankings “should not be the primary or only response to misbehavior used by a caregiver” and felt on balance that there was more evidence suggesting long-term ill effects from physical punishment than there was evidence of short- or long-term positive effects. Even so, the panel was able to agree only on recommending against physical punishment for children under age 2, primarily because the participants believed that young children were particularly vulnerable should the punishment “escalate” to a level that might cause physical injury (American Academy of Pediatrics 1996). Opponents of physical punishment make four general classes of arguments. First, they claim, there is no good evidence that physical punishment is any better than, or even as good as, other forms of punishment or means of developing discipline (Socolar et al. 1997). Second, the use of physical punishment is associated with increased rates of emotional distress during childhood and adulthood, including increased rates of alcoholism and depression (Holmes and Robins 1988; Seagull and Weinshank 1984). Third, physical punishment teaches that violence and force are acceptable means of interpersonal communication, and its use is associated with high rates of physically aggressive behavior in childhood and adulthood (Strassberg et al. 1994; Straus 1994). Fourth, when physical punishment is delivered at a time of heightened parental anger or extreme child fear, there is a danger of escalating levels of violence being directed toward the child or parent (Committee 1998). One of the major rebuttals to arguments against physical punishment is that studies showing its ill effects have failed to distinguish between punishment in the context of a loving family and punishment delivered in an emotionally harsh environment (Larzelere 1996). Gunnoe and Mariner (1997), for example, analyzing longitudinal data from the National Survey of Families in Households, found that spanking was associated with children’s aggressive behavior in some families but not others. Gunnoe and Mariner examined data for children ages 4 through 11, African American and white, living in two-parent or single-parent homes. For the majority of children, no statistically significant relationship was found between spanking and aggression. However, among African-American girls ages 4–7 living in two-parent homes, spanking was inversely related to later aggressive behavior, while for white boys ages 8–11, spanking was positively related to later aggression. The authors felt that these results supported the hypothesis that in some families, spankings could be perceived by children as legitimate expressions of parental authority and thus lead to improved behavior. One of the main barriers to consensus on the use of physical punishment has been the need for studies that examine its use in the context of other disciplinary techniques. While some parents – notably those who have been abusive to their children – use a relatively narrow and inflexible range

Child Discipline in the First Three Years

151

of disciplinary strategies (Chamberlain and Patterson 1995), most parents deploy a range of approaches that vary with the nature of the child’s behavior (Trickett and Kuczynski 1986). In families where abuse has not been reported, violations of day-to-day social rules, such as having bad manners or coming home late, tend to be responded to with reasoning and nonpunitive strategies, while “high arousal” behavior such as being physically rough or arguing may receive punishment. Scholars argue that the outcomes of physical punishment could easily be quite different in families where it is used as a primary or sole means of discipline and in families where it forms a part (probably small) of a range of disciplinary practices. A similar argument is made regarding harsh emotional tone: some believe that it is harsh emotions, not physical punishment, that harms children (Larzelere 1996). The Commonwealth Survey of Parents with Young Children does not allow us to look at differences in child outcomes that might be related to various disciplinary styles, but it does let us examine correlates of physical punishment. First, following on the recommendations of the American Academy of Pediatrics panel, it is important to learn about physical punishment directed at very young children, especially those who cannot yet walk or talk. Are certain children particularly likely to receive this form of punishment, or are certain parents particularly likely to employ it? Second, combinations of disciplinary styles are significant. Do some parents use predominantly harsh, negative, or physical punishment, and do others use a mix of techniques, including punishments and incentives, physical and non-physical? Finally, do parents feel they could use more information about child discipline? If so, it might be possible to provide help to those who want it or to those whose parenting styles might put their children at risk.

lessons from previous surveys Several studies in the past decade have attempted to describe patterns of child rearing and discipline among families in the United States. These studies have used different definitions of discipline and different timeframes (for example, the past week, the past year, ever) for measuring the frequency of disciplinary acts. The studies have involved widely varying numbers of respondents, and therefore they vary in the detail with which they can try to explain parent behaviors. In addition, although some studies examine the same families at different points in time, the analyses presented have all been based on information collected at a single interview. Thus, important questions about what comes first – for example, difficult child temperament or increased physical punishment – cannot be answered. A survey entitled Disciplining Children in America, conducted by the Gallup Organization during August and September 1995, gathered

152

Wissow

responses from 1,000 parents with children under 18 and was intended to be representative of parents nationally. The vast majority of parents (85 percent) said that they sometimes yelled at their children, and 57 percent reported using what the survey called “ordinary physical punishment,” including spanking with a hand only (47 percent), slapping on the hand, arm, or leg (37 percent), and nonabusive shaking (9 percent), in the past year. Twenty-one percent reported spanking their child in the past year with a belt or other object. Nearly all parents reported using either explanations, time out, a distraction, or taking away privileges. Child age was a factor for how parents applied physical but not verbal discipline. Parents were only slightly less likely to yell at children under age 5 (79 percent) than at teenagers (88 percent), for example, but children under 5 were more than three times as likely to be spanked, slapped, or shaken (74 percent, compared with 21 percent). Despite their frequency of yelling and hitting, parents were divided on the utility of these approaches to discipline. Two-thirds said that yelling does not work for children of any age, and about half doubted the efficacy of spanking. About a quarter of the parents surveyed said they do not believe that spanking is at all effective as a means of disciplining children under age 7; of those parents, however, 37 percent said they had slapped or hit their child in the past year. Among the parents surveyed by Gallup, income served to explain some – but not a lot – of the difference in frequency of spanking and slapping: 61 percent of the poorest parents had spanked, slapped, or shaken their child in the past year, compared with 51 percent of the wealthiest. There was much less variation in yelling; if anything, there was a trend for higher income families to yell more. Approaches to discipline also varied by whether the parent had experienced abuse as a child. Sexually abused parents were somewhat more likely to spank (61 percent) than were parents who had not been sexually abused (55 percent). In contrast, physically abused parents were less likely to spank (50 percent) than parents who had not been physically abused (55 percent). Parents who had been spanked as children, however, spanked their own children at a rate comparable with that of parents who had been sexually abused (61 percent). The National Family Violence Surveys, conducted in 1975 and 1985, collected information from a total of nearly 4,400 families across the United States (Straus 1994). Physical punishment was most common among children from age 2 through age 7: about 90 percent of parents of children within this age group said they had engaged in some form of physical punishment of their child in the past year, a rate about three times as high as the rate reported by parents of teens. Overall, among parents of children under 18, just over 60 percent said they had used some form of physical punishment. Of those, about 90 percent said they had slapped or spanked their child, and about half said they had shoved their child. Overall, parents reported an average of nine episodes of physical punishment a year.

Child Discipline in the First Three Years

153

Like the Gallup survey, the National Family Violence Survey also found that belief in the efficacy of hitting had relatively little to do with parents’ actual practices. Especially for younger children, parents hit regardless of their beliefs about the efficacy of what they were doing. The survey found no differences by income and slightly higher rates of hitting among white than among minority families. As in the Gallup survey, parents who were spanked as children were more likely to hit their own children. Straus also analyzed data from the National Longitudinal Study of Youth (Straus 1994), an ongoing “panel” survey of about 5,800 women of ages 14–21 in 1979. The survey participants were selected to include a higher percentage of minority and poor women than prevailed in the general population. When interviewed in 1988, women with children were asked about their use of spanking in the past week. Of those who had children under age 6, two-thirds said they had found it “necessary” (the wording of the survey) to spank their child at least once that week; the average number of times was three. For the last 12 years, the National Committee to Prevent Child Abuse (NCPCA) has also conducted annual surveys of attitudes toward various aspects of child rearing, child abuse, and parenting education, using random-digit telephone dialing to develop a national sample of about 1,200 homes. The 1998 NCPCA study asked about the prevalence of yelling and hitting, as well as opinions about the usefulness of these practices (Center on Child Abuse Prevention Research 1998). The survey found that about 45 percent of parents living with a child under 18 had spanked or hit their child in the past month – a substantial percentage, but lower than the 62 percent who reported spanking or hitting in 1988. Over the same decade, there was a similar drop, from 53 percent to 37 percent, in parents reporting that they had spoken harshly to their child (insulting or swearing). These declines applied across income, ethnic, and geographical lines, with one exception. Among the youngest parents, ages 18–24, the share who reported spanking increased from 45 percent to 59 percent. The survey also found a decline in the proportion of parents who felt that physical punishment could be risky for children. In 1988, 40 percent of parents were concerned that physical punishment could often lead to a child’s injury, compared with 30 percent in 1998. Day and colleagues (1998) analyzed data on the use of spanking from the National Survey of Families and Households (NSFH), a large-scale, 13,000– respondent study conducted in 1987–88. In thinking about the survey results, the researchers tried to address several issues regarding conclusions reached by other teams that had studied spanking. First, they were concerned that some prior work had classified families in too simplistic a fashion, by whether they had ever spanked or not. Day reasoned that there might be a difference between those who spanked rarely and those who spanked more frequently, or with more intensity, or with an intent to harm.

154

Wissow

The NSFH data allowed them to address the question of frequency, but not intensity or intent. Another concern was that prior studies had not considered the full range of differences in parents and children that might lead to a differential use of spanking, including the age and personality of the child, the parent’s mental health, and characteristics of the family in which the parent and child had to interact. The children involved in the NSFH ranged in age from 1 to 11 at the time of the survey. Parents were asked if, and how many times, they had spanked their child in the previous week. Mothers were more likely than fathers to have spanked in the last week: Among parents of boys, 44 percent of mothers and 31 percent of fathers said they had spanked in the last week; and among parents of girls, 37 percent of mothers and 20 percent of fathers said they had spanked in the last week. Over all age groups, boys tended to be spanked slightly more frequently than girls. The frequency of spanking was highest among toddlers and preschool children and decreased as children entered grade school. Day and colleagues found it difficult to arrive at any single predictor of parents’ use of spanking. Children who were seen as being “easier” were less likely to be spanked. Older, more educated parents who rated their own mental health as better were less likely to spank. Parents’ religious conservatism had an inconsistent relationship with spanking, but it was somewhat more predictive among unmarried than married parents. Similarly, family poverty was predictive only among African American married mothers with older children and white unmarried mothers with younger children. Among mothers of children ages 1–4, single white mothers reported slightly lower frequencies of spanking than married mothers, while the opposite was true among African-American mothers. The frequency of arguing between parent and child was the most consistent predictor of spanking: More frequent arguing was strongly correlated with more frequent spanking. The presence of a step-parent, social support for the parent, and larger family size influenced spanking, but in different ways among parents of different ethnic groups. Day and colleagues concluded that understanding the use of spanking might require consideration of specific combinations of parent age, marital status, mental health, and culture, as well as child age and temperament. Smith and Brooks-Gunn (1997) examined data from a group of 715 mothers selected because they had given birth to low-weight infants. The mothers, 39 percent of whom were single heads of household, were followed for three years. Questioned when their children were 3 years old, 41 percent said they had used more than one physical punishment with the child in the past week. Boys were not significantly more likely than girls to be hit more than once in the past week. For boys, family income level was not related to reports of the child being hit; for girls, however, parents in a middle income group (150–250 percent of the federal poverty level) were

Child Discipline in the First Three Years

155

more likely to report hitting more than once a week than poorer or more well-to-do parents. Given the different methods involved in these studies, what can we tentatively conclude? Children of preschool, kindergarten, and first-grade ages are the most likely to be hit or spanked. Studies are inconsistent on whether boys are more likely to be spanked than girls. There is some evidence that the use of physical punishment is declining, although it is probably still practiced in about half of American homes. Parents use physical punishment despite relatively little faith in its effectiveness. One motivator for its use may be the parent’s own experiences as a child, with those who were spanked more likely to spank their own children. It is difficult to paint a picture of the typical parent who uses physical punishment. Studies get different results when they look at variations by income or ethnicity, possibly because of differences in how researchers define physical punishment.

results from the commonwealth survey The Commonwealth Survey affords the opportunity to look at how parents discipline very young children, those under age three. The survey asked parents about their use of six types of disciplinary practices. Three practices could be classified as inherently negative in tone, by definition involving physically or emotionally aversive acts (Trickett 1986): yelling, spanking, and hitting. The other three practices – time out, taking away a toy or treat, and explaining – can be performed in a more neutral or positive manner. All parents were asked about their use of the three negative practices. Only parents of children 18–36 months were asked about the neutral/positive practices. All questions were asked without reference to a particular time period; instead, parents were asked to characterize the frequency with which they used particular practices: often, sometimes, rarely, or never. Overall, about 40 percent of parents say that they sometimes or often yell at their child, and 40 percent say they have spanked their child at least once (Table 6.1, and see Table 6.2). Only 7 percent say they have ever hit or slapped their child. These three practices become more common as children get older. Although 16 percent of parents say they sometimes or frequently yell at children ages 6–11 months, 63 percent yell at children ages 2–3 years. Similarly, 11 percent of children ages 6–11 months old have been spanked at least once, compared with 66 percent of 2- and 3-year-olds. Hitting and slapping also increase in frequency as children got older; 3 percent of parents say they have hit their 6–11-month-olds, compared with 11 percent of children ages 2–3 years. All three neutral/positive practices are used sometimes or often by about 50 percent or more of parents. Time out is the most common, used with 62 percent of 18–23-month-olds and 73 percent of 24–36-month-olds.

16 11 4 3 2 na na na

2 0

1 0

na

na

na

6–11 months

4

<6 months

na

na

na

8 1

36 12

40

12–17 months

58

48

62

10 4

59 18

51

18–23 months

71

55

73

11 2

67 25

63

24–36 months

66

53

69

7 2

40 14

40

All ages

p < .001

N.S.

p < .01

p < .0001 n.s.

p < .0001 p < .0001

p < .0001

Significancec

a Estimated

na = not asked in this age group. proportions of respondents with index child in each age group. Estimates derived from survey data adjusted for sampling methods. b Respondents were asked to indicate “never,” “rarely,” “sometimes,” or “often.” Results in the table are presented as “sometimes and often” versus “rarely and never,” except for hit/slap and spanking, where we also present “never” versus any reported use (“ever”). The “ever” cut points are presented and used in the text because, as can be seen from the table, relatively few parents reported hitting or spanking their children “sometimes” or “often.” c Significance test is for difference in percentage use of each technique across the age groups; significance values are based on Pearson chi-squared statistics, corrected for the survey design.

Yell (sometimes or often) Spank (ever) Spank (sometimes or often) Hit or slap (ever) Hit or slap (sometimes or often) Time out (sometimes or often) Take away (sometimes or often) Explain (sometimes or often)

Disciplinary practiceb

table 6.1. Reported Use of Disciplinary Practices by Child Agea (%)

7 6 11 6 4 2

Parent age (years) < 25 25–29 30–34 35 or more

10 5 9 4 1 2

Family income ($1000s per year) < 10 10 – < 20 20 – < 30 30 – < 40 40 – < 60 60 or greater

Parent gender Male Female

7 6 7

Parent depressive (symptom score) 0 1 or 2 3 or moreb

Factor

.017

.63

.05

.86

Significance level

Children under 12 months of age

Estimated percentage of parents who ever spanka

table 6.2. Correlates of Spanking by Child Age

60 62 54 44

57 56

49 58 66 60 60 44

52 65 54

Estimated percentage of parents who ever spanka

(continued)

p < .01

N.S.

p < .001

p < .01

Significance

Children 12–36 months of age

6 2 7 11 8 4 3 9 6 11 5 7 8 5 8 2 1

Parent education < High school High school/GED Some college College or more

Parent race Asian Black Hispanic White

Parent (respondent) employment Full time Part-time Unemployed

Parent abused as child Never Verbal or emotional Physical or sexual

Estimated percentage of parents who ever spanka

.003

.50

.18

.06

.52

Significance level

Children under 12 months of age

Parent marital status Married or live together Widowed, divorced, or separated Never married

Factor

table 6.2. (continued)

56 53 62

58 52 64

41 67 47 57

53 63 60 47

56 53 60

Estimated percentage of parents who ever spanka

(continued)

p < .05

N.S.

p < .05

p < .001

N.S.

Significance

Children 12–36 months

4 7

Child premature Yes No

0.43

0.92

0.0004

Significance level

b In

derived from survey data adjusted for sampling methods. the survey, three or more symptoms is considered the point at which a parent might be referred for further clinical follow-up.

a Estimates

6 6

3 8 18

Estimated percentage of parents who ever spanka

Children under 12 months of age

Child gender Male Female

Episodes/day of frustration with child None 1 >1

Factor

table 6.2. (continued)

57 56

59 53

34 57 68

Estimated percentage of parents who ever spanka

N.S.

N.S.

p < .001

Significance

Children 12–36 months

160

Wissow

Next most common is explaining, used with 58 percent and 71 percent of children in the two age groups. Taking things away is the least common of the neutral/positive methods, used with 48 percent of the younger and 55 percent of the older children. Boys are slightly but significantly more likely to be yelled at or spanked than girls. Forty-two percent of parents of boys say that they sometimes or often yell, compared with 37 percent of parents of girls (p < .05). Similarly, 43 percent of parents of boys say that they have spanked their child at least once, compared with 37 percent of parents of girls (p = .05). Parents of boys are no more likely than parents of girls to say they hit their child. There are no statistically significant differences by child gender for any of the other practices covered by the survey. Spanking Young Children To examine parents’ practices regarding spanking very young children, and because spanking infants is relatively uncommon among survey participants, the results have been divided by child age (Table 6.2). Overall, 6 percent of the parents of children under 1 year of age and 56 percent of the parents of children 12–36 months of age report having spanked their child. The single most obvious association between parent characteristics and reported spanking of children under 1 year old was parents’ frustration or aggravation with their child’s behavior. An estimated 18 percent of parents who say they are frustrated or aggravated more than once a day have spanked their infant, compared with 3 percent of parents who say that on a typical day, they are never frustrated or aggravated. Other parent characteristics that may be related to the use of spanking, but which fell short of statistical significance (with the level for statistical significance adjusted to .005 by the Bonferroni method because of the number of comparisons made) include education, income, and age. The less education parents or guardians have, the more likely they are to say they have spanked their infant. Similarly, lower income is associated with a greater proportion of children being spanked. The younger parents or guardians are, the more likely they are to spank. Parents who report having experienced abuse as children are significantly less likely to say they have spanked their infant: only 1 percent of parents who say they were physically abused report spanking their infant, compared with 8 percent of parents who report no history of abuse. Parental depression, gender, marital status, employment, and race are not significantly related to reports of spanking; neither are child gender or history of premature birth. Reports from parents of children 12–36 months of age indicate a different set of characteristics related to spanking. Parent frustration continues to be strongly associated with spanking: Parents reporting two or more episodes

Child Discipline in the First Three Years

161

of frustration or aggravation on a typical day are twice as likely as those reporting no episodes to say they have spanked their child (68 percent, compared with 34 percent) (p < .001). Among children in this age group, parent-depressive symptoms are also a predictor of spanking. Using the scale developed for the study, parents reporting one or two depressive symptoms are more likely to say they have spanked their child (65 percent) than are parents who have no symptoms (52 percent) or three or more symptoms (54 percent) (p < .01). In this age group, income and education are strongly associated with the use of spanking, but not in the “linear” fashion observed as a trend among the parents of infants. Families with incomes in the $20,000–$30,000 range report the greatest proportion of spanking (66 percent), with lesser percentages found among higher and lower income groups (p = .001). Parents with a high school education (or equivalent) are more likely to report spanking (63 percent) than parents with higher or lower education (p = .001). Younger parents, as among the parents of infants, are also more likely report that they have spanked their child. Spanking is reported by an estimated 60 percent of parents under age 25, 62 percent of parents ages 25–29, 54 percent of parents ages 30–34, and 44 percent of parents 35 and over (p = .01). Among children ages 12–36 months, a parent’s history of abuse has only a marginally significant effect on spanking, and in the opposite direction seen among infants. The trend among parents reporting that they were physically or sexually abused as children is to spank older children more, not less, than parents with no history of abuse. Parent race was also marginally significant, with African-American parents reporting a higher prevalence of spanking (67 percent) than whites (57 percent), Hispanics (47 percent), and Asians (41 percent) (p = .05). Neither parent gender, marital status, or employment were related to use of spanking, nor was child gender or history of premature birth. A logistic regression was used to analyze parent factors predicting the use of spanking among infants under 12 months (Table 6.3). Factors were chosen from those shown in Table 6.2 that were statistically associated with spanking at a p < .05 level or better, a very inclusive standard. Frustration and aggravation are associated with increased use of spanking. Parents who report more than one episode per day of frustration or aggravation with their infant are nearly eight times as likely to spank as those with no episodes, and parents reporting only one episode on a typical day also have an increased odds of spanking. A parental history of sexual or physical abuse is strongly associated with parents not spanking. With regard to income, only families in the $40,000–$60,000 income range have a significantly different (and lower) odds of spanking. Neither parent age nor education was independently associated with spanking infants. Another logistic regression analyzes parent factors predicting the use of spanking among children ages 12–36 months (Table 6.4). Among children

Wissow

162

table 6.3. Logistic Regression Analysis of Factors Associated with Ever Spanking an Infant

Factor

Coefficient

Parent abused as child Never Verbal or emotional Physical or sexual

Reference category −1.43 −2.57

Family income ($1000s per year) < 10 10 – < 20 20 – < 30 30 – < 40 40 – < 60 60 or greater

Reference category −0.82 −0.28 −1.11 −1.93 −1.41

Parent education < High school High school/GED Some college College or more

Reference category −0.02 −0.63 −0.57

Episodes/day of frustration with child None 1 >1

Reference category 1.22 2.06

Parent age (years) 35 or more < 25 25–29 30–34

Reference category 0.68 0.67 0.45

Constant

−2.90

∗p <

Odds ratio and 95% confidence limits (if significant)

0.08 (0.01 − 0.61)∗

0.14 (0.03 − 0.63)∗

3.4 (1.4 − 7.9)∗∗ 7.9 (2.8 − 22.1)∗∗∗

0.06 (.01 − 0.25)∗∗∗

.05. ∗∗ p < .01. ∗∗∗ p < .001.

in this age group, only two factors, both related to parental mental health, are associated with spanking. As with parents of younger children, daily frustration and aggravation increases the odds that a parent will spank. Parents who report more than one episode of frustration or aggravation are four times more likely to spank, and one episode a day more than doubles the odds. Parents reporting some but not the highest level of depressive symptoms are significantly more likely to report that they have spanked their child. Parental income, education, and age are not significantly related to spanking children in this age group.

Child Discipline in the First Three Years

163

table 6.4. Logistic Regression Analysis of Factors Associated with Ever Spanking a Child Age 12–36 Months

Factor

Coefficient

Parent depressive symptom score 0 1 or 2 symptoms 3 or more symptoms

Reference category 0.34 −0.26

Family income ($1000s per year) < 10 10– < 20 20– < 30 30– < 40 40– < 60 60 or greater

Reference category 0.09 0.35 0.06 0.26 −0.15

Parent education < High school High school/GED Some college College or more

Reference category 0.33 0.22 −0.22

Episodes/day of frustration with child None 1 >1

Reference category 0.87 1.35

Parent age (years) 35 or more < 25 25–29 30–34

Reference category 0.17 0.34 0.09

Constant ∗p <

−1.03

Odds ratio and 95% confidence limits (if significant)

1.4 (1.02–1.92)∗

2.38 (1.65–3.48)∗∗∗ 3.86 (2.61–5.70)∗∗∗

0.36 (.20–.63)∗∗∗

.05. ∗∗ p < .01. ∗∗∗ p < .001.

Parent Frustration Because frustration is such a strong independent predictor of spanking, we attempted to develop a profile of parents who report feeling this way about their children. An estimated 42 percent of parents say they are frustrated or aggravated with their child at least once on a typical day, and 26 percent say they feel that way more than once a day. Women are more likely to report frustration (74 percent) than men (61 percent) (p < .0001). White and African-American parents are more likely to report frustration (72 percent and 65 percent, respectively) than are Asians and Hispanics

164

Wissow

(59 percent and 55 percent) (p = .0001). Parent depressive symptoms are strongly related to the frequency of episodes of frustration or aggravation. Among parents reporting three or more depressive symptoms, 44 percent report being frustrated or aggravated with their child more than once a day, compared with 21 percent of parents reporting no depressive symptoms (p < .0001). Child gender and age are also strongly related to parents’ experience of frustration. Sixty-three percent of parents of boys report one or more episodes of frustration a day, compared with 53 percent of parents of girls (p < .0001). As children get older, parents are more likely to experience episodes of frustration or aggravation. Only 40 percent of parents of infants under 5 months old report one or more episodes a day, compared with 83 percent of parents of children ages 24–36 months (p < .0001). Frustration is not significantly associated with parent income, marital status, education, employment, or experience with abuse. The Context of Punishment In interpreting data about the use of physical or harsh verbal punishment, it is persistently difficult to differentiate situations in which punishment forms the sole or predominant basis for parent-child interaction from those that take place in the context of other, more positive parenting interactions. There are several ways to approach this problem. (The following analyses pertain only to children ages 18–36 months because questions about time out, explaining, and taking things away were not asked of parents of younger children.) First, we can ask the straightforward question of whether the six disciplinary practices covered in the survey, plus some additional aspects of parent-child interaction (which might be called nurturing interactions and include reading, listening to music, playing, and hugging), are more or less likely to occur together (Table 6.5). A first observation is that spanking is positively correlated with all the other forms of discipline, indicating that parents who spank their children use more of all kinds of discipline. The correlation is strongest for yelling and hitting, but it is present as well for time out, taking things away, and explaining. Spanking is also negatively correlated with reading to the child, playing with the child, listening to music together, and hugging the child, although in each case the correlation is small in magnitude. These results suggest that parents who spank are a heterogeneous group. Some parents may spank, hit, and raise their voice, use relatively little time out, taking away, or explaining, and engage in less interaction with their children, including less reading, playing, listening to music, and hugging. But the correlations are of varying and generally low strengths, suggesting that some parents who spank may not fit this pattern. A second way of considering the context of spanking is to try to characterize parents by the predominant tone of their disciplinary tactics.

Child Discipline in the First Three Years

165

table 6.5. Correlation of Disciplinary Practices and Parent-Child Activities with Parent Reports of Ever Spanking Their Childa Practice or activity

Correlationb

Hitting Yelling Time out Taking away Explaining Reading Listening to music Playing Hugging

.39 .43 .16 .23 .11 −.07 −.065 −.09 −.08

Significance p < .0001 p < .0001 p < .0001 p < .0001 p < .01 p < .0001 p < .0001 p < .01 p < .05

a Correlations

between spanking and explaining, time out, and taking away pertain only to children 18–36 months of age; other correlations apply to children of all ages. b The coeffecients reported were derived from linear regression with spanking “ever” as the dependent variable. The regression procedure took into account the survey design.

Although the categories used in the survey do not allow us to calculate precise ratios of neutral/positive to negative approaches, we rated the responses (never, rarely, sometimes, or often) according to a point system ranging from 0 to 3 for each of the six discipline types, and divided the total for negative discipline types into the total for both neutral/positive and negative types. Therefore, a parent who scored 0 would have used only neutral/positive techniques, while a parent who scored 1 would have used only negative techniques. A score of 0.5 suggests that neutral/positive and negative approaches were used with equal frequency. An estimated 5 percent of parents scored 0, suggesting that they never used any of the negative approaches. An estimated 10 percent of parents scored 0.5 or above, suggesting that they used negative means of discipline half or more of the time. Looked at this way, an estimated 85 percent of parents report that the overall tone of their discipline is neutral or positive. Third, we performed a “cluster” analysis using both the six discipline items and the four parent-child activities (playing, listening to music, hugging, and reading with their child). The goal of the cluster analysis was to define groups of parents that had clearly different balances of negative and neutral/positive approaches, or different balances of disciplinary interactions compared with nurturing interactions. Cluster analysis is a technique for exploring whether in a heterogeneous group of people there are subgroups within which people tend to be more like each other (Aldenderfer and Blashfield 1984). The first step in cluster analysis involves deciding what characteristics will be used

166

Wissow

to describe individuals. For our purposes, the key characteristic is the frequency with which parents say they use the six disciplinary practices (never, rarely, sometimes, or often) and engage in the four parent-child activities (from “not at all” to “more than once a day”). The next step is to select a mathematical method for determining the degree of similarity between any two individuals. We used a method known as the “squared Euclidian distance,” a common means for determining similarity and one particularly well suited to our data because the ten characteristics being analyzed are measured on similar scales. We then decided to assemble subgroups, or clusters, of similar individuals using a mathematical approach known as the Ward method (Aldenderfer and Blashfield 1984). This method has a tendency to produce clusters of similar size, which facilitates analysis. Another choice in performing cluster analysis involves deciding how many subpopulations to recognize within the overall study group. This decision ends up being a compromise between homogeneity within each subpopulation (which favors more, smaller subpopulations) and making each subpopulation large enough to allow it to be meaningfully characterized. After inspecting some preliminary cluster analyses of the Commonwealth Survey data, we decided to divide the parents into four groups, which seemed to yield the best compromise between homogeneity and group size. Figure 6.1 shows the composition of the four clusters in graphic form. Each bar represents the estimated percentage of parents within each cluster who say they have done the corresponding activity with their child. Again, each cluster represents a group of parents who have a similar range of styles of discipline and nurturing interactions with their children. Our goal in creating the clusters is to see the various contexts in which spanking takes place. Cluster 1 makes up an estimated 22 percent of parents surveyed who had children 18–36 months of age. It has the smallest proportion of parents who say they have ever spanked their child (36 percent). Although over 80 percent of cluster 1 parents report playing with and hugging their child, less than half report engaging in any of the other disciplinary or nurturing interactions (Table 6.6). About 60 percent of parents in cluster 1 are male, and they are the least likely to report depressive symptoms or frustration with their children. This is not a particularly affluent group – 56 percent make more than $30,000 per year, but 29 percent say their annual income is less than $20,000 – but it is not the poorest of the four clusters. This is also an ethnically diverse group: 20 percent are Hispanic, 11 percent are African-American, and 65 percent are white. We might give this group the descriptive label of “low-interacters.” Its overall pattern may be driven by men who show affection toward and play with their children, but who do not engage in much other activity with them.

Child Discipline in the First Three Years

167

In contrast, cluster 2 might be called “high-interacters.” Cluster 2 includes nearly half (47 percent) of the survey participants with children ages 18–36 months. Two-thirds (67 percent) of cluster 2 parents say they have ever spanked their child – meaning that they are almost twice as likely to spank as cluster 1 parents – yet a large share of them report engaging in all the disciplinary and nurturing interactions, with the exception of hitting. Two-thirds (68 percent) of cluster 2 parents are women. Only a third report one or more depressive symptoms, but 89 percent report daily episodes of frustration with their children. This group is largely white (81 percent), living with a partner (87 percent), and of moderate to upper income (63 percent with annual income of $30,000 or greater, 21 percent with $20,000 or less). This group’s characteristics may be driven by women from relatively affluent homes who spend a lot of time with their children. Cluster 3 is the smallest, making up only 7 percent of survey parents with children 18–36 months old. With the highest proportion of parents reporting that they spank their child (93 percent), it might be characterized as “high negative discipline.” Only 3 percent of cluster 3 parents say they read to their child, and only about half say they listen to music with their child. About half of cluster 3 parents are women (53 percent). What sets cluster 3 apart from the others are the relatively high proportion of families living in poverty (37 percent with incomes below $20,000 per year) and the high proportion (60 percent) of parents reporting one or more symptoms of depression. Parents in cluster 3 are the most likely (25 percent) to report that they are single. The ethnic composition of cluster 3 is very similar to that of cluster 1, with a diversity of ethnic groups represented. Cluster 4, with 23 percent of the survey parents with children 18– 36 months of age, is another group in which the large majority (78 percent) report having spanked their children. Members of this group, characterized as engaging in “high overall discipline,” are most likely to use time out (94 percent), and report relatively small amounts of nurturing activities such as reading (31 percent) and listening to music (3 percent). Cluster 4 is 61 percent male, 79 percent white, and the most affluent of the clusters, with only 15 percent having incomes of $20,000 a year or less. The characteristics of cluster 4 seem to be driven by men who are the family’s disciplinarians but who have relatively little other interaction with their children. To summarize, the clusters suggest that in the Commonwealth population, spanking occurs in varying contexts. Relatively low amounts of spanking occurred in the context of low overall interactions. Relatively high amounts of spanking can occur in several contexts: those characterized by high amounts of parent-child interaction overall, including much positive, nurturing interaction; those where the prevalence of parental depressive symptoms is high and families are burdened by poverty; and those characterized by relative family affluence and in which a parent – usually male – has a role as the disciplinarian.

100

figure 6.1 Child Discipline and Parent-Child Interactions in Clusters.

estimated percent of parents

hug

hug

80

listen to music

listen to music

60

play

play

40

read

read

20

explain

explain

0

take away

time out

ever hit

ever spank

take away

time out

ever hit

ever spank

Cluster 1: Low interaction

20

estimated percent of parents

0

40

Cluster 2: High interaction

60

80

100

20

40

60

80

100

figure 6.1 (Continued)

estimated percent of parents

hug

hug

0

listen to music

play

read

explain

take away

time out

ever hit

ever spank

listen to music

play

read

explain

take away

time out

ever hit

ever spank

Cluster 3: High negative discipline

20

estimated percent of parents

0

40

60

Cluster 4: High overall discipline

80

100

Wissow

170

table 6.6. Characteristics of Parents and Children in Four Clusters, Defined by Discipline and Parent-Child Interactions

Characteristic

Estimated percentage of survey population represented by cluster Parent gender (% female) Parent depressive symptoms (% score of 1 or more) Parent frustration (% one or more episodes a day) % income < $20k/year Parent employment (% not employed) Parent education (% < high school) Marital status (% single) Parent abused as child (%) Child gender (% female) Ethnicity (% white) a

Cluster 1: Cluster 2: low high interaction interaction

Cluster 3: Cluster 4: high negative high overall discipline discipline Significancea

22

47

7

23

—

42

68

53

39

p < .0001

34

35

60

93

p = .005

63

89

92

83

p < .0001

29 27

21 28

37 34

15 11

p = .02 p < .0001

17

9

20

8

p < .05

12

13

25

9

p = .01

25

28

35

26

n.s.

45

48

47

45

n.s.

65

81

62

79

p < .001

Significance value is from survey-weighted chi-squared procedure testing difference in percentage with a given characteristic among the four clusters.

Parents’ Need for Information about Discipline The survey asked parents if they had “all the information” they need, or if they “could use more information” about how to discipline their child. Just over half of parents (about 58 percent) say they have enough information about discipline. Use of particular disciplinary practices is not associated with parents feeling they could use more information. Parents who report more than one episode of frustration each day are more likely to say they could use more information (54 percent) than are parents who report only one episode a day (40 percent) or no episodes (36 percent) (p < .0001). Similarly, parents reporting depressive symptoms are more likely to say they could use information about discipline: 45 percent of those with a symptom score of 1 and 56 percent with a symptom score of 2, compared

Child Discipline in the First Three Years

171

with 40 percent of parents reporting no depressive symptoms (p < .01). Parents of firstborn children are less likely than parents of later-born children to say they could use more information (38 percent and 46 percent, respectively) (p = .0033). A lower percentage of male respondents (37 percent) than female respondents (46 percent) feel a need for information on discipline (p = .001). Health professionals (doctors or nurses) are not the only sources from whom a parent might get information about discipline, but they are potentially among the most readily available, given the frequency with which young children receive medical care. For survey respondents, however, health professionals do not appear to be a very common source of information about discipline. Overall, only 23 percent of parents say they have ever talked about discipline with a health professional. As with the perception of being able to use more information, a history of talking with a health professional about discipline is related to none of the various child disciplinary practices included in the survey, nor to parent ethnicity or marital status. Female respondents are more likely than males (28 percent versus 18 percent) (p < .0001) to have talked with a health professional about discipline. A trend of borderline statistical significance (p < .05) shows the share of parents who report talking with a health professional increasing as a function of increasing income. Parents who report more than one episode of frustration or aggravation with their child are more likely to have talked with a health professional (25 percent) than those reporting only one episode a day (19 percent) (p < .05). Depressive symptoms, however, are not related to having talked with a health professional. Parents’ ratings of their health professionals’ skills are related to whether or not they have talked about discipline, although it is not clear whether the ratings led to or were the result of the parent-provider conversation. Thirty percent of parents who rate their health professional as excellent at helping them understand their child’s growth and development say they have talked to that professional about discipline, compared with 12 percent of parents who rate the professional as poor (p < .0001). There is a similar trend, though of borderline significance, for parent ratings of professionals’ listening and answering skills. Twenty-eight percent of parents who rate the professional as excellent say they have discussed discipline, compared with 17 percent who rate the professional as poor in this area (p = .05). Parents’ Use of Time Out Time out is a disciplinary strategy frequently recommended to parents as an alternative to physical punishment (Howard 1996). Prior studies with small populations have found that mothers vary in their receptivity to using different child behavioral interventions (Heffer and Kelley 1987). In particular, although mothers overall rated positive reinforcement,

172

Wissow

non-physical consequences, and time out as more acceptable than spanking, low-income mothers were substantially less enthusiastic about time out than were middle- or upper-income mothers. These differences could be very important in planning educational campaigns aimed at providing alternatives for physical punishment. Among respondents to the Commonwealth Survey, education and ethnicity, but not income, are related to reports of using time out. The estimated proportion of parents using time out sometimes or frequently is 57 percent among those with less than a high school education, compared with 72 percent among those with a college education (p < .01). White and African-American parents are more likely to report using time out (72 percent and 63 percent, respectively) than are Asians and Hispanics (47 percent and 57 percent, respectively) (p < .01). There are no statistically significant differences by parent gender, marital status, depression, experience of abuse, or employment, nor by child gender.

conclusions The first goal of this analysis has been to learn more about the physical punishment of very young children. Like other studies, our work suggests that parents’ total disciplinary “output” increases as children get closer to early school age. Spanking is relatively uncommon among children under 6 months of age, with only 2 percent of parents saying they have ever used it, but its prevalence increases rapidly. By 6–11 months, about one child in ten has been spanked; by 18–23 months of age, 59 percent have been spanked at least once, according to their parents. Children under 1 year of age may be particularly vulnerable to injury in the course of physical discipline. The first year of life is a peak age for fatalities attributable to physical child abuse: in 1992, the National Committee to Prevent Child Abuse estimated that 43 percent of child abuse fatalities involved children less than 1 year of age (NCPCA 1993). Among the factors we studied, parent frustration is the strongest independent predictor of spanking children less than 1 year old. Parents reporting more than one episode of frustration or aggravation with their child daily have an eight-fold increase in the odds of spanking. Factors contributing to spanking appear to shift among children 12– 36 months of age. In addition to frustration, parental depressive symptoms also predict spanking. Importantly, after accounting for parental frustration and depression, we did not find significant influences by parent age, income, or education, with the exception of reduced odds of spanking children under 12 months in one higher income group (not the highest). These findings combine to suggest several conclusions. For the majority of parents, counseling about disciplinary practices and alternatives to

Child Discipline in the First Three Years

173

spanking needs to take place during the first year of the child’s life, and the earlier the better. By the time children are 18 months old, over a third of them are likely to have been spanked. Particular attempts need to be made to identify parents who experience more than average levels of frustration with parenting, or who are experiencing depressive symptoms, since those parents are over-represented among those who initiate spanking early in a child’s life. National efforts (such as the National Institute of Mental Health’s Depression Awareness, Recognition, and Treatment Program) are in place to detect and treat depression, and scholars have advocated screening for parental depression in the course of pediatric care (Kemper and Kelleher 1996). Our results suggest that asking about the day-to-day stress of parenting may also prove fruitful. The second goal of our analysis has been to identify the contexts in which spanking takes place. As scholars have predicted, spanking takes place against many diverse backgrounds of parent-child relationships. We found that spanking is most common among a group of parents who tend to be single, have lower incomes, and experience more depressive symptoms. These parents also engage in relatively little reading with their children, and they use relatively low amounts of non-physical punishments, such as time out or taking things away. Children of parents in this group may be particularly vulnerable for developing social and emotional problems, both as a function of physical punishment but also from relative lack of positive interactions. Spanking is also very prevalent (practiced by 70–80 percent of parents) as part of two other, very different patterns of interaction. One, more typical of women, includes spanking along with high levels of all the other measured forms of interaction, except hitting. The other, more typical of men, includes spanking with a high prevalence of non-physical punishments (time out and taking things away), but little reading or music and relatively little playing. The data available in the Commonwealth Survey do not allow us to know if the impact of spanking – positive or negative – differs depending on its use in these three contexts. For the purpose of education, however, we suspect that parents in the three groups should be approached very differently. Finally, at least as revealed by our cluster analysis, we might say that a low level of spanking is not necessarily on its own a marker for desirable parent-child interaction. In this particular analysis, a low prevalence of spanking was associated with relatively low levels of all forms of interaction, including those known to foster healthy child development. The third goal of our analysis was to identify parents who feel they need more information about child discipline. Just over half the parents (58 percent) say they have enough information. Parents who are frustrated or who experience depressive symptoms are more likely to say they could use more information – an encouraging result since their prevalence of

174

Wissow

spanking indicates that they would be targets for an educational campaign. Women are more likely than men to say they could use more information: indeed, one challenge for the future will be finding ways to motivate men to seek more education on child discipline and development. Parents of firstborn children are less likely than parents of later-born children to say they could use more information about discipline. It may be that having one child is less stressful, or that by the time parents have more than one child they are more open to parenting advice. Our results are also consistent with prior findings (Heffer and Kelley 1987) in suggesting that parents’ cultural background may play a role in their receptivity to the use of particular disciplinary approaches. Although we did not find significant ethnic differences in the use of spanking, we did find relatively large differences in the use of time out, the most commonly used disciplinary tactic among those covered by the survey and one widely advocated by authorities in child development (Howard 1996). The Commonwealth Survey results do not tell us whether these differences result from beliefs about the utility or appropriateness of time out, or from differences in parent knowledge. Either explanation would be important to clinicians and health educators planning programs for parents. Finally, we found that health professionals are not a common source of information about child discipline. Again, women are more likely than men to have sought information from a health professional, perhaps because they are more likely to say they could use more information or because they are more likely than men to bring children for health visits. Although we cannot prove the direction of the association, parent ratings of provider skills are associated with having sought information about discipline. Parents who think their providers are better at providing help on child growth and development and parents who give higher ratings to their provider’s listening and answering skills are more likely to have gotten advice from the provider about discipline. These results are consistent with findings that a physician’s interviewing style – including such characteristics as asking questions about psychosocial issues, making statements of support and reassurance, and listening in a sympathetic and attentive way – are related to the frequency with which parents discuss physical punishment and other potentially sensitive topics (Wissow et al. 1994). We do not know from the Commonwealth Survey why more parents do not seek information about discipline from medical professionals, especially at a time of the child’s life when medical visits are especially frequent (see Schuster, this volume). Past studies have suggested that parents do not ask about child behavior and related topics for several reasons: They think the doctor is uninterested in the area, they worry that the topics are not appropriate for medical visits, they fear being reported for child abuse, or they do not want to talk about the topics in front of their children (Hickson et al. 1983; Wissow and Roter 1994).

Child Discipline in the First Three Years

175

In summary, the Commonwealth Survey of Parents with Young Children has helped to expand our understanding of disciplinary practices early in life. First, it reinforces our awareness that parent education needs to begin early, in anticipation of children’s beginning to walk and talk. By that time, parent disciplinary output is already high. Second, it confirms that the use of physical punishment happens in a variety of contexts. Longitudinal studies could help define these contexts and identify which, based on outcomes, should be the focus of efforts at parent education. Third, the survey directs our attention toward assessing parents’ emotional state and the extent to which they feel stress about parenting as risk factors for the use of physical punishment, and as markers for parents who may be most receptive to parenting advice. Finally, the survey can serve as a benchmark for efforts that aim to increase the psychosocial content of pediatric care.

references Aldenderfer, M.S., & Blashfield, R.K. (1984). Cluster analysis. Sage University Paper Series on Quantitative Applications in the Social Sciences, 07–44. Beverly Hills, CA: Sage Publications. American Academy of Pediatrics, Conference Participants. (1996). Consensus statements. The short- and long-term consequences of corporal punishment. Pediatrics, 98, 853. Baumrind, D. (1967). Child care practices anteceding three patterns of preschool behavior. Genetic Psychology Monographs, 75, 43–88. Baumrind, D. (1989). Rearing competent children. In Damon, W. (ed.), Child Development Today and Tomorrow. San Francisco: Jossey-Bass. Baumrind, D. (1994). The social context of child maltreatment. Family Relations, 43, 360–68. Center on Child Abuse Prevention Research. (1998). Public Opinion and Behaviors Regarding Child Abuse Prevention: 1998 Survey. Working paper 840. Chicago: National Committee to Prevent Child Abuse. Chamberlain, P., & Patterson, G.R. (1995). Discipline and child compliance in parenting. In Bornstein, M.H. (ed.), Handbook of parenting, Vol. 4. Applied and Practical Parenting. Mahwah, NJ: Lawrence Erlbaum. Cherlin, A.J. (1996). Public and Private Families: An Introduction. New York, NY: McGraw-Hill. Committee on Psychosocial Aspects of Child and Family Health, American Academy of Pediatrics. (1998). Guidance for effective discipline. Pediatrics, 101, 723–28. Day, R.D., Peterson, G.W., & McCracken, C. (1998). Predicting spanking of younger and older children by mothers and fathers. Journal of Marriage and the Family, 60, 79–94. Epstein, M. (1995). Thoughts without a Thinker. New York, Basic Books. Gallup Organization. (1995). Disciplining Children in America. Princeton: Gallup Organization.

176

Wissow

Gil, D.G. (1970). Violence against children. Cambridge: Harvard University Press and the Commonwealth Fund. Grossman, D.C, Rauh, M.J., & Rivara, F.P. (1995). Prevalence of corporal punishment among students in Washington State schools. Archives of Pediatric Adolescent Medicine, 149, 529–32. Gunnoe, M.L., & Mariner, C.L. (1997). Toward a developmental–contextual model of the effects of parental spanking on children’s aggression. Archives of Pediatric Adolescent Medicine, 151, 768–75. Heffer, R.W., & Kelley, M.L. (1987). Mothers’ acceptance of behavioral interventions for children: the influence of parent race and income. Behavior Therapy, 2, 153–63. Hickson, G.B., Altemeier, W.A., & O’Conner, S. (1983). Concerns of mothers seeking care in private pediatric offices: opportunities for expanding services. Pediatrics, 72, 619–24. Holmes, S.J., & Robins, L.N. (1988). The role of parental disciplinary practices in the development of depression and alcoholism. Psychiatry, 51, 24–36. Howard, B.J. (1996). Advising parents on discipline: what works. Pediatrics, 98, 809–17. Kagan, J. (1983). The psychological requirements for human development. In Skolnick, A.S. & J.H. Skolnick (eds.), Family in Transition, 4th edition. Boston, MA: Little, Brown. Kemper, K., & Kelleher, J. (1996). Rationale for family psychosocial screening. Ambulatory Child Health, 1, 311–24. Lakoff, G. (1996). Moral Politics. Chicago: University of Chicago Press. Larzelere, R.E. (1996). A review of the outcomes of parental use of nonabusive or customary physical punishment. Pediatrics, 98, 824–28. LeDoux, J. (1996). The Emotional Brain. New York: Simon & Shuster. Maurer, A. (1979). It Does Happen Here. In Hyman, I.A., & J.H. Wise (eds.), Corporal Punishment in American Education. Philadelphia: Temple University Press. McLoyd, V.C. (1990). The impact of economic hardship on black families and children: psychological distress, parenting, and socioemotional development. Child Development, 61, 311–46. National Committee to Prevent Child Abuse (NCPCA). (1993). Current trends in child abuse reporting and fatalities: the results of the 1992 Annual Fifty State Survey. Working paper #808. Chicago: National Committee to Prevent Child Abuse. Sameroff, A.J. (1997). Understanding the social context of early psychopathology. In Greenspan, S., S. Wieder, & J. Osofsky. Handbook of Child and Adolescent Psychiatry, vol. 1. New York, NY: John Wiley. Seagull, E.A, & Weinshank, A.B. (1984). Childhood depression in a selected group of low-achieving seventh-graders. Journal of Clinical Child Psychology, 13, 134–40. Skolnick, A.S., & Skolnick, J.H. (1983). Family in Transition, 4th edition. Boston, MA: Little, Brown. Smith, J.R., & Brooks-Gunn, J. (1997). Correlates and consequences of harsh discipline for young children. Archives of Pediatric Adolescent Medicine, 151, 777–86. Socolar, R.R.S., Amaya–Jackson, L., Eron, L.D., Howard, B., Landsverk, J., & Evans, J. (1997). Research on discipline: the state of the art, deficits, and implications. Archives of Pediatric Adolescent Medicine, 151, 758–60.

Child Discipline in the First Three Years

177

Stern, D.N. (1985). The Interpersonal World of the Infant. New York: Basic Books. Strassberg, Z., Dodge, K.A., Pettit, G.S., & Bates, J.E. (1994). Spanking in the home and children’s subsequent aggression toward kindergarten peers. Developmental Psychopathology, 445–61. Straus, M.A. (1994). Beating the Devil Out of Them: Corporal Punishment in American Families. New York: Lexington Books. Suransky, V.P. (1982). The Erosion of Childhood. Chicago: University of Chicago Press. Trickett, P.K., & Kuczynski, L. (1986). Children’s misbehaviors and parental discipline strategies in abusive and non-abusive families. Developmental Psychology, 22, 115–23. Wissow, L.S., & Roter, D. (1994). Toward effective discussion of discipline and corporal punishment during primary care visits: findings from studies of doctorpatient interaction. Pediatrics, 94, 587–93. Wissow, L.S., Roter, D., & Wilson, M.E.H. (1994). Pediatrician interview style and mothers’ disclosure of psychosocial issues. Pediatrics, 93, 289–95. Yazzie, R. (1994). Life comes from it: Navajo justice concepts. New Mexico Law Review, 24 (spring), 175–90.

7 Breastfeeding in the United States Today: Are Families Prepared? Wendelin M. Slusser and Linda Lange

Breastfeeding provides an important indication of how a family is organized and functions and it determines the initial nutrition and feeding of an infant child. Although breastfeeding is an age-old and essential behavior for species survival, the practice has declined over this century because of the availability of cow’s milk formula. Today, rather than being an automatic behavior for child survival, breastfeeding is a choice that depends on social, family, and health system factors. We know that more families are making the decision to breastfeed. That choice – one of the first parenting decisions reached by new parents – is a product of a series of complex decisions and adaptations. It is influenced not only by desire and inclination but also by other family, social, and work-place issues. Success is not assured for all families. Indeed, some families are more successful than others in providing this nutritional and nurturing resource to their children. The Commonwealth Survey provides a unique and important source of independent, nonproprietary data about breastfeeding and its determinants in the United States today, including information about the decision to breastfeed and some of the challenges families face. Much of the available data on breastfeeding in the United States comes from surveys conducted by Ross Laboratories, manufacturers of breastmilk substitutes. These data are used by the Federal government in Department of Health and Human Services annual reports (U.S. Department of Health and Human Services 1998) and by authors who publish in professional and research journals (Ryan 1997). Analysis of the Commonwealth data enables us to look at the challenges families face in deciding to breastfeed and the level of assistance they receive from health care providers and others in the health care system. It also allows us to consider what policies would support and promote breastfeeding for families. The results could be important for health care providers, policymakers, the public health community, and 178

Breastfeeding in the United States

179

others interested in understanding the current determinants of breastfeeding initiation and duration for women and their families.

the benefits of breastfeeding How to feed a new baby is one of the very first choices new parents make. Today, the majority of parents (62 percent) in the United States choose to breastfeed their infants, with more than half of them stopping by the end of six months (U.S. Department of Health and Human Services 1998). The decision to breastfeed is the product of a complex set of cultural, familial, educational, and practical considerations. For the family, negotiating the breastfeeding experience depends on a variety of factors but can be influenced by interactions with health care providers, employers, and others in the public sector. These interactions include education and support of the family, starting with prenatal care providers, continuing in the hospital or birthing center at delivery and in the immediate postpartum period, and during visits with the infant’s well-child care provider. For the mother who intends to return to school or the work force, the need for continuing sources of education and support services can be extensive, involving key players in the work or school setting and child care providers. The Early Connection between Parent and Infant The moments surrounding the birth of a baby, up to a few weeks thereafter, have been described by Klaus (1999) as a “sensitive period” for parents. This is a time when mothers, and perhaps fathers, may be especially susceptible to experiencing the connection with their new baby and their new role as parents. As the mother learns her baby’s unique cues and responds to them, the mother and her baby develop a relationship that enables her to care for the baby and meet the baby’s individual needs. Early breastfeeding contributes to the initial connection between mother and infant. This initial period and the bonding that results from it influence later parenting behavior. A great deal of research has been devoted to understanding this complex new relationship. Some studies have focused on the time mother and infant spend together immediately after birth. Investigators have noted that if the lips of the infant touch the mother’s nipple in the first hour of life, the mother is likely to keep the baby in her room for longer periods during the hospital stay than if she does not have this early contact with her infant (Klaus 1999). Another study indicates that mothers who have early contact, room-in, and suckle their infants in the first hour of life breastfeed for a longer duration than mothers who do not have this early and close contact (DeChateau and Wiberg 1977; Salariya et al. 1978). Some researchers believe that this maternal behavior may be explained in part physiologically by

180

Slusser, Lange

the release of the hormone oxytocin. The mother’s plasma level of oxytocin rises dramatically immediately after delivery if the infant is placed skinto-skin on her chest immediately after birth. Oxytocin is known to play a role in maternal bonding in animals and may also play such a role in human bonding (Nissen et al. 1995). Along with other maternal-infant contact in the newborn period, breastfeeding contributes to a complex set of behaviors that enhance the connection of parents to their newborns (Klaus 1999). Cognitive Development Breastfeeding has also been suggested to have a positive influence on infant cognitive development, with lifelong consequences. Recent work in brain development has acknowledged the unique properties of human milk for brain development. Recognition of the specific function of the long chain polyunsaturated fatty acids, present in human breast milk, as components of neural membrane development necessary for optimal brain development has provided a biological basis for this interest. Breastfeeding and the critical stages in brain development, connections between brain development and cognitive and emotional development, and the influence of early life experiences, including stress and parent relationships with infants, have become inseparable. These issues are now a focus of national attention (Hamosh et al. 1991; Lucas et al. 1992; Rogan and Gladden 1993; Hamosh 1998a). Studies over the past 70 years have found small but consistently higher cognitive ability among children who were breastfed. Groundbreaking work by Lucas and coworkers showed significantly higher cognitive benefits among premature infants fed human milk compared with premature infants fed formula (Lucas et al. 1992). Premature infants (weighing <1850 grams) who were gavage fed breastmilk for the first four weeks of life showed a 7–8 point advantage in IQ at 7.5–8 years of age. The authors were able to control for certain confounding factors, including maternal behavior, maternal intelligence, and socioeconomic factors, but not for genetics. Lucas and colleagues hypothesized that long-chain polyunsaturated fatty acids in breastmilk contributed to the difference in IQ. Hamosh (1998a, 1998b) agrees with this hypothesis, especially for the premature infant who is deficient in long-chain polyunsaturated fatty acids as a result of missing maximum transfer of these nutrients from mother to fetus during the last part of pregnancy. Other research has found that the full-term infant may benefit developmentally from breastfeeding, as seen in a long-term study in New Zealand (Horwood and Fergusson 1998). This 18-year longitudinal study showed a small increase in cognition and educational achievement using standardized tests, ratings by teachers, and other academic testing

Breastfeeding in the United States

181

among 1,000 children who were breastfed as infants. Other researchers, however, suggest that the observed advantages of breastfeeding on IQ may be related to genetic, social or environmental factors, or other behaviors associated with the propensity to breastfeed, rather than to the nutritional benefits of breastfeeding on neurodevelopment (Jacobson et al. 1999). There is a real need for well-designed prospective studies to determine whether the effects of breastfeeding are causal rather the casual (Hamosh 1998b). As one research team has stated, “Although all of the modern studies attempted to control for social factors, it is almost impossible to separate completely the mother from the milk” (Rogan and Gladden 1993). Child and Maternal Health Breastfeeding initiates a complex interrelationship involving infant, maternal, and family health as infants, mothers, and families reap health benefits from breastfeeding in a number of ways. Breastmilk provides complete nutrition for the infant, sustaining optimal growth and hydration in the first six months of life. For infants, health benefits lead to reductions in infectious disease, including acute otitis media, respiratory infections, bacteremia, meningitis, urinary tract infections, necrotizing enterocolitis, diarrhea, and other acute gastrointestinal illness, where the risk is six times greater for formula fed than for breastfed infants in the United States (World Health Organization 1981; Beaudry et al. 1995; Teele 1989; Koopman 1985). Some evidence also suggests a reduced risk for chronic disease and allergies, including type I insulin dependent diabetes, Crohn’s disease, lymphoma, eczema, asthma, and some food allergies (Chandra 1986; Zeiger et al. 1986; Black et al. 1989; Brown et al. 1989; Victoria et al. 1989; Howie et al. 1990; Teele 1991; Cunningham et al. 1991; Duncan et al. 1993; Goldman 1993; Lanting et al. 1994; Wright et al. 1995; Slusser and Powers 1997). Health benefits of breastfeeding are especially important for children placed in out-of-home childcare. In out-of-home group childcare settings, rates of infectious disease – including diarrheal disease, upper respiratory disease, and meningitis – are significantly lower among breastfed infants than among formula-fed infants (Dahl et al. 1991; Ponka et al. 1991; Paulley and Gaines 1993; Super et al. 1994; Sempertegui et al. 1995). Three additional studies have looked at feeding mode and its impact on infectious disease rates in children cared for out of the home. These studies suggest a significant reduction in infectious disease rates among infants who continued to be fed breastmilk after their mothers returned to work. Collectively, these findings are important, given the trend toward center-based childcare and the much higher rates of diarrheal disease, upper respiratory disease, and meningitis for infants and toddlers who attend (Arnold et al. 1993; Jones and Matheny 1993; Cohen and Mrtek 1994).

182

Slusser, Lange

For women, as well, there are also health benefits from breastfeeding. Some studies suggest a preventive effect against reproductive cancers, including pre-menopausal breast cancer (Newcomb et al. 1994) and ovarian cancer (Whittemore et al. 1992). Regarding osteoporosis, there may be an increase in bone mineral density after breastfeeding for six to nine months or longer (Kritz-Silverstein et al. 1992). Maternal weight loss after pregnancy is enhanced by exclusive breastfeeding for periods of six months or longer (Dewey et al. 1993). Exclusive breastfeeding yields a family planning benefit, with improved child spacing equivalent to hormonal methods if the “lactation amenorrhea” criteria are observed (Gray et al. 1990). Mothers breastfeeding longer than six months have reported emotional benefits, including happiness and emotional security (Reamer and Sugarman 1987). Working mothers also may benefit from a reduction in absenteeism from work, improved productivity at the work site, reduced health care expenses, and reduced maternal stress (Auerbach and Guss 1984; Keanrey and Cronenwett 1991; Jones and Matheny 1993; Cohen and Mrtek 1994). Financial Savings Breastfeeding may confer significant economic advantages for families and society (Wong et al. 1994). It has been estimated that if exclusive breastfeeding were practiced in the United States for the first 12 weeks of each infant’s life, $2.16 billion annually would be saved as a result of reductions in illness and disease nationwide. In addition, $3.02 billion would be saved annually from household budgets, owing to reduced expenditures for formula, family planning, and health care (Labbok 1995). According to one estimate, if women participating in the Federal Women, Infants, and Children Supplemental Feeding Program (WIC) were to breastfeed for one month, $30 million would be saved in formula costs by the United States government (Walker 1992). At the family level, a woman who breastfeeds for the first year of an infant’s life saves more than $400 from the family’s budget through savings in the purchase of breastmilk substitutes (Gartner et al. 1997). Data from a California study suggest that if working mothers fed their infants breastmilk for 100 work days, 27.3 percent less absenteeism would occur at the work site and 35.7 percent fewer health care claims would be filed (Cohen and Mrtek 1994).

breastfeeding trends in the united states The Healthy People 2000 objectives for breastfeeding, set by the U.S. Public Health Service, established goals of increasing breastfeeding rates to at least 75 percent in the early postpartum period and to at least 50 percent

Breastfeeding in the United States

183

until infants are six months of age. Although initiation rates have come close to reaching the goal, breastfeeding at six months remains far below the objective, especially for certain populations, including low-income and minority families (Ryan 1997; Healthy Children 2000 1991). Breastfeeding Trends According to data collected by Ross Laboratories, breastfeeding initiation and duration rates have been rising in the United States after an almost 10-year decline through the 1980s. In the last decade, breastfeeding initiation has increased for all women in the United States, from a low in 1990 of 51.5 percent to 62.4 percent in 1997, the highest rate recorded in recent years (U.S. Department of Health and Human Services 1998). There has also been considerable success in closing the breastfeeding gap between white and non-white women. The largest improvements in breastfeeding initiation have occurred among black, young, and low-income women, for whom rates have traditionally been lowest. Since 1990, rates of breastfeeding increased by 80 percent for black women and 33 percent for Hispanic women, changes that have contributed to a substantial decrease in the disparity between white and minority mothers. By 1997, 66.5 percent of white women, 63.6 percent of Hispanic women, and 41.3 percent of black women initiated breastfeeding. Although the increases have been greatest among black and Hispanic women, white women still remain more likely to start to breastfeed than other groups (Ryan 1997; U.S. Department of Health and Human Services 1998). By 5 to 6 months postpartum, breastfeeding duration rates decrease by almost 40 percent for white and Hispanic women and over 25 percent for black women. Along with increases in initiation, however, there has been an increase in duration rates. Between 1989 and 1995, breastfeeding duration for six months or more increased by 19.3 percent, from 18.1 percent to 21.6 percent. The increase continued through 1998, when breastfeeding rates six months postpartum stood at 28.6 percent for white mothers, 24.5 percent for Hispanic mothers, and 14.5 percent for black mothers (U.S. Department of Health and Human Services 1998). These figures are still far below government goals. Racial and ethnic disparities found in breastfeeding initiation and duration rates closely mirror those found in pregnancy and birth outcomes, although movement toward eliminating maternal and child health disparities between white and minority populations called for by the federal government have been more successful in breastfeeding (U.S. Department of Health and Human Services 1998; National Center for Education in Maternal and Child Health 1998). Regardless of the steep decline in breastfeeding at six months, more women across all socioeconomic and regional groups have initiated and continued to breastfeed beyond six months in

184

Slusser, Lange

recent years. The most significant recent increases in breastfeeding rates are seen among women at greatest risk not to breastfeed – black, young (less than 25 years of age), and low income women, as well as those who have a grade school education only, are primiparous, are employed full time outside the home, are feeding low birthweight infants, are participating in the WIC program, and are living in the South Atlantic region of the United States. Women who are older, better educated, living in the West, multiparous, not enrolled in WIC program, not employed outside the home, feeding infants of normal birth weight, and in higher income brackets are still, however, more likely to initiate breastfeeding and to breastfeed for six months or longer (Ryan 1997). Cultural Differences The reasons women give for breastfeeding are complex and varied, and may help explain variations in breastfeeding initiation rates across racial and ethnic groups. For example, social support is influential for every group and age, but support may come from a variety of sources. African American women, for example, are more likely to be positively influenced to breastfeed by close friends. Partners and their own mothers are more likely to affect the breastfeeding decisions of white and Hispanic women. The sociocultural environment also influences breastfeeding. Women who live in the western part of the United States, regardless of race or ethnicity, are more likely to breastfeed than women living in other parts of the country. This finding may be indicative of social support for breastfeeding independent of the individual’s immediate social milieu (Ryan 1997). However, acculturation into American society also plays a role. Newly arrived immigrant Mexican women have been found to be more likely to initiate and continue to breastfeed than more acculturated immigrants (Rassin et al. 1994). Perceptions of convenience, benefits to the infant, and personal comfort level – as well as the balance among these factors – can also vary according to race and ethnicity. Baranowski and colleagues (1986) found that for African American women, the balance between the benefits to the infant and the inconvenience of breastfeeding to themselves may be important, while for Caucasian mothers, the benefit to the infant is seen as the most important factor in deciding to breastfeed. Hispanic mothers appeared to be influenced most by the convenience factor; some described breastfeeding as inconvenient and “annoying.” For adolescent mothers, the picture is different. For adolescent white and Mexican-American mothers, the perceived benefits of breastmilk are the strongest predictor of deciding to breastfeed. Other influential factors for Mexican-American teen mothers include the preference of their partner or mother. For white teens, the influence of the health care provider and

Breastfeeding in the United States

185

having breastfeeding role models are important. For African American teens, however, living with the father of the child is the strongest predictor of breastfeeding, more influential than encouragement from providers, peers, mother, or non-cohabiting partners. The timing of the decision is also significant for African American teens. Those who made their decision to breastfeed later in their pregnancies were more likely to breastfeed than who decided early in pregnancy (Wiemann et al. 1998).

challenges for breastfeeding families A mother who does not start to breastfeed when her child is born is unlikely to be able to start much later, owing to the physiological limitations of lactation (Chaffee and Greisheimer 1969). Further, the health care system cannot help women or children catch up, in terms of emotional satisfaction, cognitive development, or health benefits, on what they missed by not breastfeeding. Thus, for women and children, the opportunity to gain the full range of benefits from breastfeeding is lost rather quickly and definitively. Working Mothers Family life for women and families with very young children has become more complex in the last half century, with increased labor force participation among mothers. While 12 percent of married women with children under six years of age participated in the labor force in 1950, the figure had more than quadrupled by 1988, rising to 57 percent. By 1994, 61 percent of married women in the labor force had children under 6 years old, with the majority of those women returning to work within the first three years of the child’s life (Children’s Defense Fund 1996). By 1997, 57.9 percent of women with children under one year of age were in the labor force (Kurinji et al. 1989). With no national policy or generalized support for working women who continue to breastfeed after they return to work, demands on the family choosing to breastfeed are even more challenging (Spisak and Gross 1991). Working outside the home is one of the most frequently identified barriers to continuation of breastfeeding (Surgeon General’s Workshop 1984; Institute of Medicine 1991; Spisak and Gross 1991; Duckett 1992; Auerbach and Guss 1984; Hight-Laukaran et al. 1996). Maternal participation in the work force, childcare, and breastfeeding are interconnected. The degree of control a woman has over her job and the flexibility she has in determining her hours of employment affect her relationship with her child and the duration of breastfeeding (Moore and Jansa 1987; O’Gara et al. 1994). Obstacles for the working and breastfeeding mother include the structure of the workday and the work environment, lack of awareness and information regarding the benefits of breastfeeding,

186

Slusser, Lange

and lack of childcare at or near the workplace (Auerbach and Guss 1984; Barber-Madden et al. 1987). These obstacles result in lower rates of breastfeeding among employed women. In one study of women employed at a Southern California public utility, 55 percent of both employed and not employed mothers started to breastfed in the hospital, but 24 percent of mothers not employed and 10 percent of employed mothers continued to breastfeed for six months (Cohen and Mrtek 1994). Work site requirements include a safe, private, and clean physical space, time (an hour divided into appropriate increments of 20-30 minutes), and clean, refrigerated storage (Moore and Jansa 1987). With welfare reform and passage of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996, low-income mothers, as a group historically less likely to breastfeed, are also joining the workforce, many for the first time and early in their infants’ lives. Breastfeeding may be compromised if those mothers are unable to continue breastfeeding when they return to work or school due to barriers in child care programs and services (Ryan and Martinez 1989; Hamosh et al. 1991; Newcomb et al. 1994; Cohen et al. 1995; Lindberg 1996). Closely linked to employment issues, child care adds another complexity to a family’s decision to breastfeed. In 1991, 4,558,000 children under five were enrolled in out-of-home childcare in the United States, most of these children entering as early as 11 weeks of age (Children’s Defense Fund 1996). Enrollment for infants in center-based care has tripled in the last thirty years. Yet in 1990, only half of center-based child care provided infant care, compared with almost all family day care sites. This leaves breastfeeding families with relatively few child care alternatives (Hofferth 1996; U.S. Department of Education 1995; Scarr et al. 1993), especially if they would like to continue breastfeeding and prefer center-based child care. One reason cited for the increase in center-based enrollment of older children, ages 3–4, is that preschool programs can help prepare children for success in school. New interest in child development may stimulate a similar trend for 0–3 year olds (Hofferth 1996). At present, however, no data describe whether or not breastfeeding families find support in child care settings of any kind. The Health Care System Professional education of physicians, nurses, hospital staff, and other health providers has been identified as a potential barrier to breastfeeding, since some women will not begin or continue to breastfeed without support from their health care providers. Families interact with the health care system at many points during pregnancy, at the delivery of the baby, and at well-child care visits as they grow into their role as parents. The prenatal period is an important time for influencing a family’s decision to

Breastfeeding in the United States

187

breastfeed. Mothers who receive advice to breastfeed during prenatal care visits are much more likely to breastfeed than mothers who do not receive such advice (Balcazar et al. 1995). Mothers who are married, have higher incomes, and have more education are more likely to receive this information than other mothers (Kogan et al. 1994). Mothers who attend childbirth classes are more likely to breastfeed than mothers who do not (Piper and Parks 1996). Thus, interactions with health care providers present important opportunities to educate and support families to breastfeed. Rather than facilitating breastfeeding, many hospital policies and procedures actually discourage breastfeeding. Separation of mother and infant after delivery, feeding glucose water in the hospital nursery, and distribution of free formula in the hospital are thought to provide powerful barriers to breastfeeding (Spisak and Gross 1991). The practice of distributing formula discharge packs to new parents has been shown to reduce breastfeeding rates, and contributes to breastfeeding failure (Howard et al. 1993; Howard et al. 1994; Gartner et al. 1997). The type of delivery can also be a factor in a mother’s ability to breastfeed. Currently, nearly one-quarter of all births in the United States are delivered by Cesarean section, compared with 4.5 percent in 1965 and 15 percent in 1979. Although women who deliver by Cesarean are less likely to initiate breastfeeding, especially if the Cesarean is unplanned, information from one study indicates that their duration rates are no different from those of other mothers at 3 months and at more than 3 months (DiMatteo et al. 1996). The amount of time a new mother stays in the hospital also affects breastfeeding. Length of hospital stay after childbirth has decreased markedly over the past two decades, with rapid fluctuations in the past ten years. Women who once stayed between 12 and 24 hours after an uncomplicated delivery are now staying longer, up to 48 hours, because of national and state legislation (Udon and Betley 1998). Physiological changes that mothers and newborns experience in the first few days after birth, including beginning lactation and newborn jaundice, make shorter stays problematic for initiation and duration of breastfeeding (Braveman et al. 1995). Hospital rooming-in, guidance and support from nursing staff, and the ability to feed on demand are predictive of successful breastfeeding, and decreased length of stay prevents adequate time to devote to breastfeeding support and teaching (Patton et al. 1996; Perez-Escamilla et al.1994). Changes in the health care delivery system have also affected breastfeeding families. As a source of pregnancy care and well child care, managed care has the potential to provide continuity of care, reduce costs, and link families with community services. Findings on the effects of managed care on breastfeeding are mixed, however. As managed care has become a dominant system of care, some community-based providers may be excluded from managed care organizations. In addition, financial incentives

188

Slusser, Lange

for providers may lead to less care, and the care that is provided may follow the traditional medical model rather than a developmental or behavioral model of care (Brach and Scallet 1996). Yet mothers who do not have private insurance and receive prenatal care at publicly funded sites of care or HMOs are more likely to report receiving prenatal advice on breastfeeding than women who are privately insured and receive care from private physicians (Kogan et al. 1994). The beneficial effects of breastfeeding on infant health could reduce costs for managed care services by reducing the occurrence and duration of otitis media in the first year of life (Dewey et al. 1995; Paradise et al. 1997). In one study of an HMO population, infants exclusively breastfed for four months or more had half the mean number of acute otitis media episodes as did infants not breastfed at all. In addition, the breastfed children had 40 percent fewer acute otitis media episodes than did infants whose diets were supplemented with other foods prior to six months of age. In the same study, for non-atopic (non-allergy prone) children, recurring wheezing at age six was three times more likely if they had not been breastfed as infants (Wright et al. 1995).

analysis of the commonwealth survey Methods Our analysis focuses on responses from the 1,229 non-Hispanic white, African American, and Hispanic biological mothers who participated in the Commonwealth Survey. The demographic indicators selected for analysis include marital status, education, mother’s age at the time the baby was born, race and ethnicity, region of residence, household income, and employment at the time the survey was conducted.1 The survey is unique in its inclusion of a comprehensive variety of health care system access and utilization factors that have been associated with breastfeeding initiation and duration. These indicators include prenatal care, childbirth education classes, type of delivery, breastfeeding encouragement in the hospital, home visits after the baby was born, and participation in government assistance programs. 1 Some of the survey questions selected for this analysis are referenced to the time of the administration of the survey rather than to the time of the child’s birth. These include marital status, educational attainment, and household income. Unlike race and ethnicity, or maternal age, which was computed to the mother’s age at the time of the child’s birth, these maternal characteristics are mutable over time and might have changed between the time the child was born and the time the survey was administered. These characteristics are included in the analysis because they provide an approximation of the characteristics of the mothers reporting in the survey at the time their children were born and reflect the sample characteristics at the time the survey was conducted. Maternal employment at the time of the survey will not be used in the description of the sample or as a determinant.

Breastfeeding in the United States

189

Participation in government assistance programs since the child’s birth is included in the analysis because of the reported success of the WIC program in increasing breastfeeding initiation. The survey does not distinguish, however, among various programs, including AFDC, WIC, Food Stamps, Medicaid, and other state or local public assistance programs. A unique question in the survey asks whether mothers were encouraged or discouraged from breastfeeding by doctors or nurses while still in the hospital. Although most mothers reported that they were either encouraged or discouraged, some volunteered that they had received mixed messages – or both encouragement and discouragement – about breastfeeding. Still other mothers said the topic was not discussed and they had received no information on breastfeeding. For this analysis, mothers who said they were encouraged to breastfeed are compared with those giving other responses. Parents were also asked if they received a home visit from a nurse or other health care professional to help them learn to care for their newborn. Although the content of the visit is not defined or described as specifically including breastfeeding support or education, this type of contact has been associated with success in linking families with young children to other health care and social services. Parents were asked if their children were breastfed more than one month, less than one month, or not at all. For this analysis, only biological mothers and their responses are considered. After describing the infant feeding practices of all families in the sample, we present the first logistical analysis, which compares mothers who initiate breastfeeding with those who never breastfeed to determine what influences this primary decision. The second logistical analysis considers only mothers who breastfed, and permits a look at differences between mothers who start to breastfeed yet stop before the first month of the child’s life and those who continue beyond one month duration. Infant Feeding Practices Family composition and maternal and household demographic indicators are generally consistent with findings from other breastfeeding studies. Among all mothers in the Commonwealth Survey, 66 percent started breastfeeding their babies (Table 7.1), a figure slightly above those found in other national studies. For all mothers, that percentage drops to 53 percent – a decrease of almost 20 percent – after just 1 month to a percentage barely above government expectations for 6 months. As in other studies, race and ethnicity are associated with feeding choices regarding breastfeeding or formula feeding. Almost 60 percent of African American mothers decided not to start to breastfeed, compared with about 30 percent of white and Hispanic mothers. Among mothers who choose to breastfeed, white mothers are somewhat more likely to continue breastfeeding after the first month (58 percent), with similar findings for Hispanic mothers (54 percent).

Slusser, Lange

190

table 7.1. Association between Maternal Characteristics and Health Services Indicators and Infant Feeding Choice for Biological Mothers (%) Breastfed Breastfed >1 month <1 month (53.2%) (12.4%)

Never breastfed (34.4%)

All women (100%)

Race ethnicity∗∗∗ (n = 1220) White African American Hispanic

57.8 27.5 54.0

11.5 13.3 16.2

30.7 59.2 29.8

73.0 13.4 13.6

Maternal Age∗∗∗ (n = 1216) < 25 years 25–29 years 30–34 years 35+ years

36.3 55.6 66.6 67.3

18.2 13.1 6.7 5.4

45.5 31.2 26.7 27.3

32.5 31.1 24.3 12.0

Marital status∗∗∗ (n = 1212) Married Divorced Single

59.3 41.7 23.1

11.3 13.9 17.9

29.4 44.4 59.1

80.5 5.8 13.7

Education∗∗∗ (n = 1216) High school High school graduate Some college College grad/grad school

32.4 42.0 55.3 80.5

15.7 13.1 14.5 6.4

51.9 44.9 30.2 13.1

15.5 32.4 29.5 22.7

Household income∗∗∗ (n = 1114) < $10,000 < $20,000 < $30,000 < $40,000 < $60,000 $60,000+

38.1 42.3 48.8 52.6 57.4 75.6

17.0 14.5 14.1 17.6 12.3 2.5

44.8 43.1 37.1 29.8 30.3 21.9

14.7 13.0 16.6 13.0 22.1 20.7

Region∗∗∗ (n = 1220) North East North Central South West

53.33 53.54 42.81 69.09

11.02 8.74 15.06 13.70

35.66 37.71 42.13 17.21

19.82 24.71 33.94 21.53

Government assistance∗∗∗ (n = 1216) Yes No

41.1 63.0

15.2 10.1

43.6 26.8

45.7 54.3

Prenatal care (n = 1184) First trimester Second trimester Third trimester

53.2 54.1 58.9

12.8 5.9 5.0

33.9 40.0 36.2

93.4 6.0 0.6 (continued)

Breastfeeding in the United States

191

table 7.1. (Continued) Breastfed >1 month (53.2%)

Breastfed <1 month (12.4%)

Never breastfed (34.4%)

All women (100%)

Childbirth class∗∗∗ (n = 1219) Yes No

60.3 35.5

12.0 13.5

27.7 51.0

71.0 29.0

Type of delivery (n = 1219) Vaginal Cesarean

53.6 52.1

12.9 10.8

33.5 37.2

77.4 22.6

Encouraged to breastfeed∗∗∗ (n = 1193) Yes No

60.4 35.2

14.1 8.0

25.4 56.8

73.2 26.8

Home visit (n = 1220) Yes No

49.6 54.3

13.5 12.1

36.9 33.6

22.8 77.2

∗p

< .05. ∗∗ p < .01. ∗∗∗ p < .001.

Only 27 percent of African American mothers continued to breastfeed after the first month, a decrease of 33 percent and well above the decrease for the total sample. The families represented in the survey are young, with the majority of mothers under 30 years of age when their children were born and when their decisions were made regarding how to feed their babies. Older women are more likely to breastfeed than younger mothers. Forty-five percent of the youngest mothers, under 25 years of age, chose not to breastfeed at all, and the young women who did start to breastfeed were the least likely to continue for more than one month (36 percent). There is a sharp difference between two-parent and one-parent families in how they feed their infants. Women who are married are much more likely than single women to begin and to continue to breastfeed their children: 60 percent of single mothers did not breastfeed their babies, while 71 percent of married women did. Among single women who started to breastfeed, less than one-quarter continued to breastfeed for longer than one month, compared with almost 60 percent of married women. Overall, the higher the educational attainment of the mother, the more likely she is to initiate and then continue to breastfeed beyond the first month. College graduates are considerably more likely to initiate breastfeeding than women who attended college but did not graduate (69.8 percent versus 86.9 percent), although college attendees are more likely to breastfeed than mothers who did not attend college at all. In terms of duration, college attendees resemble high school graduates more closely

192

Slusser, Lange

than they do college graduates. Breastfeeding after the first month drops by 25 percent among high school graduates, 21 percent among college attendees, and only 7 percent among college graduates. Women who have not graduated from high school are the least likely to start breastfeeding. As household income rises, the likelihood of breastfeeding and continuing to breastfeed also rises. Mothers reporting annual household income greater than $60,000 are the most likely to breastfeed beyond one month. Conversely, mothers with the lowest household incomes are the least likely to breastfeed. Where families live also makes a difference in their decision to breastfeed. As in other studies, women and their families who live in the western part of the country are most likely to start and continue to breastfeed, while Southern women are the least likely to breastfeed. Family participation in government assistance programs, including the WIC, Food Stamp, Medicaid, or other federal, state, or local government assistance program, is negatively associated with breastfeeding – a finding consistent with the fact that families qualify for these programs because they have low incomes. Almost half of all mothers (46 percent) reported they had received some type of government assistance. Mothers who received government assistance were less likely to breastfeed (41 percent) than mothers who did not receive government assistance (63 percent). The large majority of families represented in the Commonwealth Survey received good prenatal and obstetrical care by standard measures. Ninetythree percent of all mothers started prenatal care in the first trimester, a figure well above the national figure of 80.2 percent (U.S. Department of Health and Human Services 1994). Because of its high prevalence in this study, early prenatal care shows no association with breastfeeding. A substantial majority of mothers (71 percent) also report that they attended childbirth classes during their pregnancies. Mothers who attended childbirth classes were much more likely to start and continue to breastfeed (60 percent) than mothers who did not attend classes (35 percent). Consistent with other studies, the majority of mothers had vaginal deliveries. However, this study shows no association between type of delivery and breastfeeding, and mothers were equally likely to breastfeed regardless of delivery type. Almost three-quarters of all mothers report that they were encouraged to breastfeed by doctors or nurses in the hospital or birthing center after their babies were born and were much more likely to breastfeed if they received encouragement. Starting to Breastfeed Table 7.2 shows the unadjusted odds for breastfeeding initiation, and includes the demographic and health services indicators included in the chi-square analysis. Here, the analysis compares all women who initiated breastfeeding, regardless of duration, with all women who never started

Breastfeeding in the United States

193

table 7.2. Unadjusted Odds of Breastfeeding Initiation Compared with Never Breastfeeding for Biological Mothers Variable

Odds ratio

95% CI

3.27∗∗∗∗ 3.42∗∗∗∗

(2.27, 4.71) (2.19, 5.33)

1.32 2.49∗∗∗∗ 7.16∗∗∗∗

(0.88, 1.54) (1.60, 3.87) (4.15, 12.33)

Marital status (n = 1212) Married Divorced Compared with single

3.46∗∗∗∗ 1.81∗

(2.49, 4.81) (1.09, 2.99)

Household Income (n = 1114) $10–19,999 $20–29,999 $30–39,999 $40–59,999 $60,000+ Compared with < $10,000

1.07 1.38 1.91∗∗ 1.86∗∗ 2.89∗∗∗∗

(0.68, 1.69) (0.87, 2.19) (1.14, 3.19) (1.13, 3.07) (1.68, 4.99)

Maternal age (n = 1216) 25–29 years 30–34 years 35+ years Compared with <25 years

1.84∗∗ 2.92∗∗∗∗ 2.23∗∗

(1.29, 2.61) (1.53, 3.42) (1.35, 3.67)

2.11∗∗∗∗

(1.57, 2.83)

1.31 1.20 3.50∗∗∗∗

(0.83, 1.98) (0.83, 1.74) (2.27, 5.41)

2.71∗∗∗∗ 3.86∗∗∗∗

(2.02, 3.65) (2.76, 5.15)

Race/ethnicity (n = 1220) White Hispanic Compared with African American Education (n = 1216) High school graduate Some college College grad/grad school Compared with < high school grad

No government assistance (n = 1216) Compared with government assistance Region (n = 1220) Northeast North Central West Compared with South Childbirth class (n = 1219) Encouraged to breastfeed (n = 1193) ∗ p < .05.

∗∗ p < .01.

∗∗∗ p < .001.

∗∗∗∗ p < .0001.

194

Slusser, Lange

to breastfeed. Among the demographic predictors, women who are white or Hispanic or who are married are more than three times as likely to have started to breastfeed as African American or unmarried mothers. The strongest effect is seen in education, where women who are college graduates are more than seven times as likely to breastfeed as mothers who have not completed high school. Families in the higher household income groups are increasingly more likely to breastfeed. Women between 30 and 34 years of age are more likely to breastfeed than either older or younger women. Consistent with household income, families who do not qualify for government assistance programs are twice as likely to breastfeed as low-income families. Western families are more than three times as likely to breastfeed as Southern families. Both health system indicators play a large role getting mothers started with breastfeeding. Attendance at childbirth classes makes mothers almost three times as likely to breastfeed as mothers who do not attend. Families who are encouraged to breastfeed while in the hospital are almost four times as likely to start to breastfeed as families who are not encouraged, a factor second only to education in importance as a predictor. The multivariate model in Table 7.3 presents the factors that predict the likelihood of breastfeeding initiation, controlling for all other factors in the model. Education emerges as the strongest demographic predictor, although race/ethnicity remains a significant predictor: White and Hispanic mothers are more than twice as likely as African American mothers to start to breastfeed. Marriage and maternal age are less important factors in initiation than race/ethnicity, education, or region of the country where mothers live. Among health services indicators, both childbirth education and encouragement to breastfeed in the hospital remain significant predictors of breastfeeding initiation, with encouragement to breastfeed having the larger impact of the two predictors, controlling for all other factors. This finding is important because both factors are subject to changes in the health care delivery system. Household income and participation in government assistance programs were dropped from the model because they did not contribute to the fit of the final model.2 Continuing to Breastfeed For this analysis, the sample includes only mothers who breastfed (n = 661), and compares mothers who breastfed for more than 1 month with those who stopped at 1 month or less. Table 7.4 presents the unadjusted odds for breastfeeding duration. As in the demographic predictors for

2 The household income variable contains a great deal of missing data and destabilized the model. The inclusion of maternal age and marriage, even though the confidence intervals were below one at the lower bound, contributed to the overall fit of the model.

Breastfeeding in the United States

195

table 7.3. Likelihood of Breastfeeding Initiation Compared with Never Breastfeeding, Controlling for Significant Demographic and Health Services Factors (n = 1178) Variable Race / ethnicity White Hispanic Compared with African American Education High school graduate Some college College grad/grad school Compared with < high school grad Marital status Married Divorced Compared with single Maternal age (at time of infant’s birth) 25–29 years 30–34 years 35+ years Compared with <25 years Region Northeast North Central West Compared with South Childbirth class Encouraged to breastfeed −2 log likelihood Model chi-squared p-value

Odds ratio

95% CI

2.12∗∗∗∗ 2.88∗∗∗∗

(1.30, 3.46) (1.54, 5.36)

1.27 2.33∗∗∗ 6.02∗∗∗∗

(0.75, 2.15) (1.29, 4.23) (3.01, 12.04)

1.58 1.28

(0.99, 2.51) (0.71, 2.30)

1.15 1.35 1.05

(0.74, 1.80) (0.81, 2.26) (0.56, 1.96)

0.99 0.98 2.35∗∗∗∗

(0.60, 1.63) (0.65, 1.50) (1.38, 4.01)

1.67∗∗ 4.35∗∗∗∗ 1045.32 258.234 0.0001

(1.10, 2.52) (2.96, 6.38)

∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001. ∗∗∗∗ p < .0001.

starting to breastfeed, mothers who breastfeed for longer than 1 month are more likely to be white or Hispanic, well-educated, married, older, wealthier, receiving no government assistance, and to have attended childbirth classes. Region of residence remains important. However, encouragement to breastfeed in the hospital, while predictive of initiation, is not predictive of breastfeeding duration. Women who received encouragement are no more likely to breastfeed beyond the first month than women who did not receive encouragement. The effect of encouragement in the

Slusser, Lange

196

table 7.4. Unadjusted Odds of Breastfeeding Duration of Greater than 1 Month, Compared with Breastfeeding Duration of Less than 1 Month, for Biological Mothers Variable Race/ethnicity (n = 752) White Hispanic Compared with African American Education (n = 722) High school graduate Some college College grad/grad school Compared with < high school grad Marital status (n = 723) Married Divorced Compared with single Maternal age (n = 724) 25–29 years 30–34 years 35+ years Compared with <25 years Household income (n = 665) $10-19,999 $20-29,999 $30-39,999 $40-59,999 $60,000+ Compared with <$10,000 Region (n = 725) Northeast North Central West Compared with South No government assistance (n = 722) Childbirth class (n = 724) Encouraged to breastfeed (n = 715) χ 2 : ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001. ∗∗∗∗ p < .0001.

Odds ratio

95% CI

2.44∗∗ 1.62

(1.34, 4.42) (0.81, 3.24)

1.55 1.86 6.12∗∗∗∗

(0.84, 2.89) (0.98, 3.51) (2.80, 13.36)

4.07∗∗∗∗ 2.34∗∗

(2.49, 6.66) (1.15, 4.72)

2.12 5.02 6.25

(1.28, 3.52) (2.51, 10.04) (2.85, 14.51)

1.30 1.54 1.33 2.09∗∗ 13.3∗∗∗∗

(0.66, 2.55) (0.76, 3.11) (0.66, 2.70) (1.02, 4.29) (4.22, 42.18)

1.70 2.15∗∗ 1.77∗∗

(0.90, 3.22) (1.19, 3.90) (1.08, 2.91)

2.30∗∗∗

(1.48, 3.55)

1.91∗∗ 0.97

(1.21, 2.99) (0.53, 1.78)

Breastfeeding in the United States

197

immediate postpartum period apparently does not persist beyond the first month of breastfeeding. In Table 7.5, the model for predicting breastfeeding duration indicates that the influence of education, maternal age, marriage, and where families live (in the Western or North Central parts of the country) determines duration. Race/ethnicity is not a factor that predicts duration of breastfeeding. Controlling for other factors, once white, Hispanic, and African American women start to breastfeed, they are equally likely to continue beyond the first month. College or graduate education, followed closely by age, is the strongest predictor of whether or not mothers will continue to breastfeed their children in this model. The effect of prenatal childbirth education – the only remaining contribution from the health care system – drops out, superceded by the demographic characteristics and perhaps other factors not measured in this survey. The absence of health services or other untested factors that might mitigate the influence of the demographic characteristics alone is important. These results point to the need for consideration of other indicators to determine what factors other than demographic characteristics, such as support after hospital discharge or employment, affect breastfeeding duration.

discussion and policy implications Results from the Commonwealth Survey provide insight into the experience of mothers and their families and the factors that influence their decisions regarding how they feed their babies. The survey confirms and elaborates on prior research in identifying mothers who traditionally have low breastfeeding initiation and duration rates, including single, less educated, low income, and African American women and those receiving government assistance. It is not clear from the survey indicators alone why mothers who are better educated, older, and wealthier are more likely to breastfeed and to continue for longer periods, although information on services available through the health care system sheds some light on these differences. Understanding the factors that affect decisions to initiate and sustain breastfeeding could help in the development and implementation of initiatives to help families become informed about their choices as new parents, and to act on them. Prenatal Services Childbirth Education. Most families who participated in the Commonwealth Study recognized the benefit of receiving prenatal care and of starting it early in the pregnancy. Facilitated by a high level of prenatal insurance coverage (97 percent), most mothers started prenatal care in the first trimester. Fewer – just over 70 percent – participated in childbirth

Slusser, Lange

198

table 7.5. Likelihood of Breastfeeding Duration Greater than 1 Month Compared with Breastfeeding Duration of Less than 1 Month, by Demographic and Health Services Factors (n = 661) Variable Race/ethnicity White Hispanic Compared with African American Education High school graduate Some college education College grad/grad school Compared with < high school grad Marital status Married Divorced Compared with single Maternal age (at time of infant’s birth) 25–29 years 30–34 years 35+ years Compared with <25 years Household income $10-19,999 $20-29,999 $30-39,999 $40-59,999 $60,000+ Compared with <$10,000 Region Northeast North Central West Compared with South −2 log likelihood Model chi-squared p-value ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001. ∗∗∗∗ p < .0001.

Odds ratio

95% CI

1.09 1.20

(0.51, 2.35) (0.48, 3.00)

1.61 1.61 3.10∗

(0.77, 3.36) (0.75, 3.46) (1.09, 8.82)

2.01∗ 1.56

(1.06, 3.82) (0.67, 3.64)

1.64 2.96∗∗∗∗ 2.77

(0.91, 2.97) (1.41, 6.22) (0.95, 8.03)

1.07 1.02 0.69 0.75 3.97

(0.52, 2.20) (0.45, 2.10) (0.30, 1.58) (0.30, 1.85) (0.95, 16.54)

1.21 2.26∗ 1.97∗

(0.59, 2.50) (1.14, 4.28) (1.04, 3.74)

525.488 84.813 0.0001

Breastfeeding in the United States

199

education classes, perhaps for financial as well as cultural reasons. Childbirth education, generally provided through community-level organizations rather than within the traditional health care system, may not be covered by prenatal insurance and may therefore be paid for as an outof-pocket expense. These services, which typically provide breastfeeding information, have traditionally attracted self-selected middle-income families, although some studies have shown that culturally sensitive prenatal education can attract non-traditional participants (Thompson et al. 1990). The Commonwealth data are interesting in indicating that attendance at these classes had no durable effect in helping families continue to breastfeed. The Role of the Prenatal Services Provider. The role of the health care provider in providing direct support to the breastfeeding family is especially important in the prenatal period, as well as the immediate postpartum period, during the first month of the baby’s life. Research has demonstrated that the primary care provider can have a strong influence on breastfeeding success (Essex et al. 1995). The primary health care provider has a strong and far-reaching role in helping to overcome barriers to breastfeeding through support and encouragement, as defined by professional organizations. Indeed, the opportunities to support the breastfeeding family are abundant for the primary health care provider, and the need to recognize this role is critical to breastfeeding promotion. Policy statements by professional provider organizations support breastfeeding as the optimal mode of infant feeding. In December 1997, the American Academy of Pediatrics issued and published new recommendations by its Breastfeeding Work Group. The recommendations specify that the infant should be breastfed within the first hour of life, breastfed on demand, and be fed no supplementation. Further, the recommendations suggest appropriate follow-up after birth, continuation of breastmilk feeding if mother and infant are separated, and exclusive breastfeeding for the first 6 months of life and continuing for one or more years (Gartner et al. 1997). Primary health care professionals receive only minimal lactation management education in their training. In one national survey of pediatric physicians and residents, 90 percent of respondents agreed that they should be involved in breastfeeding promotion, but their clinical knowledge and experience suggested a very low degree of competency (Freed et al. 1995). A smaller study of pediatric residents found that only 14 percent of the total sample described themselves as confident or very confident to manage common breastfeeding problems (Williams and Hammer 1995). Obstetricians and family practitioners also lack the necessary skills to counsel mothers adequately (Donnelly 1994). Studies demonstrate the paucity of information on lactation management in the preservice

200

Slusser, Lange

curriculum (Newton 1992). The Department of Health and Human Services has addressed this void by developing a template for introducing the subject of lactation into the curriculums of nursing, medical, and nutrition students in the United States. In order to decrease the breastfeeding dropout rate, health professionals will need to play a much more active role in supporting breastfeeding mothers. First, they should learn to recognize every encounter with the pregnant woman, new mother, and her family as an opportunity to promote breastfeeding and to view the mother and infant as a dyad, not individually. They should promote early follow-up, so that mothers and infants discharged from the hospital soon after delivery can be seen by a health professional within four days. Health professionals should also be educated in lactation management, in professional and governmental policies relating to breastfeeding, and in communicating information about breastfeeding to their patients. Educating health professionals about women at high risk not to breastfeed could help providers follow those families more closely and increase breastfeeding rates. In addition, providers should be able to refer women with lactation problems in a timely manner. Primary care providers could also coordinate breastfeeding support efforts with the Women, Infant and Children Supplementary Food Program (WIC) by providing the same consistent messages and, for some mothers, utilizing the WIC monthly encounter visits for follow-up for breastfeeding families. Providing support and encouragement to the working breastfeeding family is another role that primary health care providers can fill. Medicaid Services. By 1994, Medicaid programs in every state provided eligible families (those with incomes below 133 percent of the federal poverty level) with prenatal care for pregnant women and health care for infants during the first year of life. Two-thirds of the states exercised their option to provide benefits to women and children with incomes up to 185–200 percent of the poverty line. As a result, Medicaid now covers up to half of all births in many states, a tremendous increase from the 17 percent of U.S. births covered by Medicaid in 1985 (Howell et al. 1991; Long and Marquis 1998). In California, for example, 48 percent of all 1994 births were reimbursed under the state’s Medicaid program (known as Medi-Cal), while only 20 percent were covered by private insurance and 26 percent by HMOs (California Department of Health Services 1997). By the first half of the 1990s, enhanced prenatal care services to Medicaid-enrolled women were provided by more than 80 percent of the states (Buescher et al. 1991; Baldwin et al. 1998). California’s enhanced Medi-Cal Comprehensive Perinatal Services Program (CPSP) also includes lactation support for mothers who elect to breastfeed. As health care for both Medicaid consumers and privately insured citizens has moved toward managed care, lactation support services to breastfeeding mothers

Breastfeeding in the United States

201

and their infants provided under Medicaid programs could serve as models for lactation benefits for privately insured families in managed care as well. In California, CPSP provides breastfeeding mothers with nutritional counseling and lactation support from licensed professionals, rental or purchase of breast pumps, and banked human milk for infants (California Department of Health Services 1998). These benefits may be especially important to groups less likely to initiate and continue to breastfeed but who may be disproportionately represented in Medicaid enrollments. In California, births to Hispanic women (68 percent) and African American women (57 percent) were much more likely to be covered by Medi-Cal than births to white women (25 percent). Policy Implications r One strategy to improve breastfeeding initiation rates among families may be to increase attendance at childbirth education classes, especially for some groups at risk for not starting to breastfeed, including minority and low-income women. Private insurers, managed care organizations (especially Medicaid managed care), and the government should be encouraged to recognize childbirth education classes as a necessary service for the prenatal period, indicative of quality of care, and provide and reimburse for this service. r Further research should be considered to better understand participation in childbirth education among minority and low-income women and its effectiveness. Other educational strategies may be more successful for promoting breastfeeding among these women. r A breastfeeding risk assessment tool for pregnant women could be developed for use in clinical practice to identify women at highest risk not to breastfeed. Such a tool could increase breastfeeding initiation rates and reduce failure rates. r Advocacy training could be provided for health professionals at all levels to support breastfeeding women and infants. r Medicaid lactation support models could be considered for broader use under managed care. The Medi-Cal lactation support model, for example, could be used directly to support breastfeeding initiation and duration by reimbursing providers for lactation services. Indirectly, measuring breastfeeding support programs and rates of initiation and duration could be incorporated into quality-of-care indicators for all managed care providers and systems. Hospitals and Birthing Centers The Commonwealth Survey families were prepared for the births of their babies: 99 percent were delivered in hospitals or birthing centers, and

202

Slusser, Lange

97 percent of families were covered by obstetrical insurance. Unlike mothers who participate in childbirth education, which is self-selecting, mothers who deliver in hospitals or birthing centers are all potential recipients of hospital-based interventions intended to promote breastfeeding. The survey results indicate that almost three-quarters of families were encouraged to breastfeed by a doctor or nurse in the hospital. This intervention was, of necessity, brief, as the majority of mothers (77 percent) had a vaginal delivery, and half of those mothers were discharged in less than 24 hours. Even so, encouragement to breastfeed in the hospital is the strongest factor in starting women on breastfeeding. Also of great importance is the finding that although hospital encouragement has a strong effect on initiation, it does not predict duration of breastfeeding beyond the first month. This evidence, along with that regarding childbirth education, implies that mothers may require additional, or different, encouragement and support for a period of time beyond the initial hospital stay in order to continue breastfeeding after they leave the hospital. The fact that mothers do not stay in the hospital for very long may contribute to the lack of persistence for this intervention. Socioeconomic factors, including the ability of the family to know about and purchase additional help for breastfeeding, may also be relevant. The women most likely to continue breastfeeding were highly educated, married, older, and affluent. Because our data suggest that doctors and nurses in the hospital play a critical role in getting families started on breastfeeding, the more than onequarter of mothers who say they did not receive encouragement to breastfeed are also of interest. In fact, the experience of many of these mothers was more complicated than reflected in the survey question, which asked only if they had been encouraged or discouraged to breastfeed. Some mothers reported independently that they had received mixed messages of encouragement and discouragement or, for one-fifth of the sample, no information at all about breastfeeding. Some of these mothers started to breastfeed (43 percent) but most did not (57 percent). The strong effect of encouragement to breastfeed right after the infant’s birth may be understood better in light of Klaus’ description (1999) of the exquisitely sensitive period around the time of birth. It is not surprising that ambiguous messages or lack of encouragement may deliver strong, negative associations. The Role of Hospitals. Reliance on modern health care facilities may become less of a barrier to breastfeeding families because of the U.S. UNICEF Baby Friendly Hospital Initiative (USBFHI). The U.S. initiative follows a worldwide effort started in 1991 by UNICEF and the World Health Organization (WHO) in collaboration with world leaders, health experts, and non-governmental organizations in response to the declining rates of breastfeeding worldwide. As of August 1996, over 8,000 hospitals in

Breastfeeding in the United States

203

159 countries had been designated as Baby Friendly. As of August 1996, 309 hospitals (including 18 birthing centers) in 43 states have received USBFHI certificates of intent. In 1999, only 17 American hospitals and birth centers had been formally assessed and received official UNICEF/WHO designation as Baby Friendly.TM The initiative originated in order to address hospital policies and marketing of breastmilk substitute that were identified by global experts in a 1989 UNICEF/WHO publication as major barriers to families who choose to breastfeed (World Health Organization 1989). The extent and effect of direct marketing of breastmilk substitutes on initiation or duration of breastfeeding through the receipt of free formula in the hospital setting cannot be determined from the Commonwealth data. Although it is not clear how mixed or negative information was transmitted to the mothers who reported it, providing mothers with hospital discharge formula packs may have contributed to the mixed message some received. Policy Implications r Research could determine why Baby Friendly Hospital Initiative protocols or other curricula are not being utilized in hospitals to encourage breastfeeding. The influence of formula gift packs at discharge should also be studied. r Further efforts to clarify hospital-based support and education are warranted. The survey does not illuminate what actions constituted “encouragement,” and mothers were free to define the concept themselves. Further hospital-based research should be more specific as to methods and outcomes of breastfeeding support. Community Services Community-level barriers for breastfeeding families have been identified as lack of education about breastfeeding among mothers; lack of support from traditional networks, including family and friends; and lack of support services in the community (Spisak and Gross 1991). The survey responses indicate, however, that most families (74 percent) actively prepared themselves to become parents by reading books, magazines, and newspapers or watching television or videos about parenting issues. Clearly, many parents look to the media, a potentially powerful tool in the community, to learn about raising their children. What the survey does not show is how information from specific media campaigns or promotions may influence mothers and families regarding breastfeeding. This may be especially important for minority groups, who may have been targeted, perhaps regionally, for media campaigns by commercial interests. The region in which a family lives makes a difference both in starting to breastfeed and in continuing. Families in the western part of the

204

Slusser, Lange

United States are more than twice as likely to start to breastfeed as families who live elsewhere in the country. This may point to community and cultural factors present in the West, beyond those measured by the survey, that support a family’s decision to start breastfeeding. Families who continue to breastfeed, however, are most likely to live in the North Central region. The availability of best practices in breastfeeding programs and services or public support for breastfeeding may be more prevalent in that area. It is not clear from the survey data what other community factors over the first month postpartum may have contributed to a family’s ability to breastfeed beyond one month. Very few families received a home visit from a nurse or other type of provider, a resource that could be promoted for all new parents as well as families identified as being at risk not to breastfeed. Here, the visit did not have an effect on initiation or duration of breastfeeding. The content or purpose of the home visit, the credentials of the home visitor, and possible visits to lactation consultants, peer counseling, or mother-to mother support were not measured. In other countries, home visiting is a routine event for all new parents, as well as for high-risk families. They are generally carried out by registered nurses to provide health education, preventive care, and social services to young families (Ross, Loening and Mbele 1987; Kammerman and Kahn 1995). An English study that found that a 24 percent increase in breastfeeding among high-risk home-visited families contributed to a reduction in post-perinatal mortality (Carpenter et al. 1983). Unfortunately, there has been no scientific investigation into how many visits are needed, how frequently visits are needed, and what the content of visits should be (Braveman et al. 1995). Women, Infants and Children Supplemental Food Program (WIC). Families reporting that they received public services, which may have included WIC, were less likely to initiate or continue to breastfeed. It was not possible to isolate WIC participation in the data as an independent factor. This may obscure any effects the WIC program may have had on breastfeeding initiation or duration, an important point because WIC is a community-level program for low-income women that promotes breastfeeding. However, the data do substantiate the same large percentage of children supported by government programs found in other sources. Between 1989 and 1995, the WIC program, administered through the U.S. Department of Agriculture, has been highly successful in increasing breastfeeding initiation and six-month duration rates among WIC participants through financial and technical support for community-based breastfeeding promotion programs (Ryan 1997). The WIC program, targeted in geographic areas with high rates of infant mortality, low birthweight, and low income, is directed toward improving pregnancy outcomes and health

Breastfeeding in the United States

205

status for low-income pregnant women, lactating mothers, and young children at nutritional risk (Rush et al. 1988). Almost half of all infants in the United States (47 percent, or 1,819,091 children) participated in the WIC program in 1995. The program has a unique opportunity to affect breastfeeding practice in the United States by providing a critical link to promoting breastfeeding among groups that have traditionally tended not to breastfeed (Ryan 1997). The Role of the Provider in the Community. The American Academy of Pediatrics (AAP) Committee on Practice and Ambulatory Medicine declared in a policy statement in 1997 that the AAP would continue to promote breastfeeding as the best form of infant nutrition and to recommend against direct-to-consumer advertising of infant formula (American Academy of Pediatrics 1997). In 1996, the AAP adopted Resolution 63, “Third-party payers should compensate for breastfeeding management visits by certified lactation consultants,” recommending that all thirdparty payers provide or reimburse for ongoing lactation services to nursing mothers as a cost-effective measure. The resolution recommends compensation for any necessary physician-referred visits to International Board of Lactation consultant examiners, certified lactation consultants, or other appropriately trained providers. The American College of Obstetricians and Gynecologists, the American Dietetic Association, and the Midwifery Association also have policies supporting breastfeeding as the optimal form of infant nutrition (Report of the Second Surgeon General’s Workshop on Breastfeeding and Human Lactation, 1991). The Role of Public Policy in the Community. In 1981, breastfeeding was recognized as a protected constitutional right by the U.S. Court of Appeals in Dike v. Orange County School Board, 650 F.2d783 ( 5th Circuit, 1981) (Baldwin 1997). Since then, thirty-two states have enacted legislation to amend criminal laws (such as indecent exposure), protect breastfeeding as a civil right, support WIC and breastfeeding promotion, and protect breastfeeding in labor laws. The Federal government has taken steps to reduce the formula industry’s influence on breastfeeding families. In 1994, President Clinton signed World Health Assembly Resolution 47.5, which reaffirmed the World Health Organization’s International Code of Marketing (World Health Organization 1981). The code, ratified by all other WHO member countries in 1981, is intended to contribute to the provision of safe and adequate nutrition for infants by protecting and promoting breastfeeding, curbing aggressive marketing by producers of breastmilk substitutes, and ensuring the proper use of breastmilk substitutes when they are necessary. This code and subsequent resolutions include the right of the baby to its mother’s milk, the right of health workers to counsel mothers, and the

206

Slusser, Lange

right of mothers to feed their babies free from all commercial pressures. An underlying intent of the code is to end all free and subsidized breastmilk substitutes in all parts of the health care system. Professional organizations have addressed these issues by recommending certain conduct from formula companies, including no direct advertising to the public (American Academy of Pediatrics 1997). Direct marketing of breastmilk substitutes to the general public started in the United States in 1989, raising the possibility that the media may be establishing new barriers to breastfeeding, including negative attitudes such as embarrassment and perceived lifestyle limitations, among women and their families (Spisak and Gross 1991). Although the Commonwealth data provide no direct information about whether or not public education, marketing, and public sentiment regarding the perceived propriety of breastfeeding has any effect on decisions mothers make about breastfeeding, women who are socially and economically privileged, regardless of race or ethnicity, breastfeed longer. Indirectly, this may indicate that some women may be in a position to be better informed about the benefits of breastfeeding or are better able to withstand public scrutiny or marketing pressures in opposition to breastfeeding. Policy Implications r WIC’s ability to support breastfeeding for low-income women could be improved through better data collection and promotion strategies. Data collection could examine the influence of WIC programs on lowincome families nationally, regionally, and locally, and promotion strategies could be made more culturally sensitive to diverse populations. r Reimbursement for lactation support could be improved to provide better access, especially for minority middle- and low-income women, to this component of the comprehensive health care system. As promoted by AAP Resolution 63, reimbursement for lactation consultants could be included in future surveys as a measure of health care benefits and quality assurance. r Surveys of mothers and families could examine their experiences with breastfeeding in private or public settings, including work sites and school, to determine whether they have experienced discrimination while breastfeeding, the influence of commercial and social marketing in their breastfeeding decision-making process, and the impact of stigmatization of breastfeeding in public places and the work place. Returning to Work or School Data from the Commonwealth Survey are consistent with national employment figures for women with young children. More than half of the

Breastfeeding in the United States

207

survey families, each with at least one child under the age of three, report that the mother in the household is currently employed. The issue of continuing to breastfeed after returning to work or school cannot be directly addressed using the Commonwealth information because there are no data linking breastfeeding duration and labor force participation beyond the first month of breastfeeding, a period when many women may not have returned to work. Indirectly, however, one of the most interesting findings concerns the relationship between returning to work and the extent to which prenatal, intrapartum, and immediate postpartum interventions for the breastfeeding family are insufficient to support their decision beyond the first month of the child’s life. This suggests that isolated employer-based support for breastfeeding in the workplace after maternity leave ends would come too late for mothers who stopped breastfeeding prior to returning to employment or school activities outside the home. These findings indicate that some mothers, and perhaps all, might require periodic and ongoing encouragement measures during the period after hospital discharge but prior to returning to work to continue breastfeeding. Indirectly, therefore, the prenatal and intrapartum measures of provider and hospital support may provide some indication as to the effort required to support a mother and her family in their choice to continue breastfeeding. That the effort or support might be required throughout maternity leave and after return to work, both areas within the purview of employers and insurers as well as health care providers, is an important finding. There is little evidence regarding the effectiveness of workplace lactation support programs to date (Moore and Jansa 1987; Cohen and Mrtek 1994). Demographic predictors for breastfeeding duration are another area where the survey provides important data regarding breastfeeding and the work place, especially for women who are less likely to continue breastfeeding and who may be in employment situations unfavorable to maintaining lactation. Low-income or less educated women, regardless of race or ethnicity, are already at risk not to continue to breastfeed beyond the first month. These mothers may also be disproportionately more likely to work in low-wage jobs where job flexibility, the structure of the workplace, or benefits may be less supportive to continuing breastfeeding. Their children may be less likely to get the health and developmental benefits available to children in families with greater socioeconomic resources. Given the trends in employment, the essential role that the employer plays in supporting breastfeeding among working mothers will be more critical in the next decade as more and more women separate from their infants for extended periods of time during the day. This can include supporting mothers in the notion that breastmilk feeding, childcare, and employment are possible; integrating and coordinating support for the working breastfeeding family; and providing counseling in strategies on how

208

Slusser, Lange

to maintain a breastmilk supply and how to store breastmilk (Scrimshaw et al. 1987; Romero-Gwynn and Carias 1989). Policy Implications r Strategies to facilitate breastfeeding for working women include education of health care providers, employers, families, and society about supporting breastfeeding women who work outside of the home, the economic benefits to the employer and society as a whole, and breastmilk storage (Barber-Madden et al. 1987). Employers and providers should be informed and involved in policy decisions and strategies regarding postpartum and newborn support for mothers and infants prior to their return to work or school. Employer-based programs could include lactation support in the prenatal, intrapartum, and postpartum periods and throughout the breastfeeding period. Such programs are seen by some companies as inducements for employee recruitment and retention (Dickson 1997). r Policies and strategies to promote breastfeeding for employed mothers should take into consideration the socioeconomic backgrounds and work site environments of all mothers. r Employers, especially larger companies, could shape benefit packages to include lactation support for wellness promotion, employee recruitment, and retention.

conclusions Based on findings from the Commonwealth Survey, we know that families appear to be well prepared to become new parents according to present health services indicators such as health insurance and early prenatal care. In fact, families are not as well prepared as they could be based on other measures of family health, including breastfeeding initiation and duration. Many families are not receiving appropriate and effective comprehensive health services, including childbirth education, hospital-based support, and home visiting, that could provide a link between maternal and child health, maternity care, and well-child services that lead to better outcomes for mothers, children, and families. Surveillance of breastfeeding rates as a sentinel indicator of appropriate and effective health care for women, children, and families could provide this link. National surveys and community report cards that include breastfeeding measures, such as the ones included in the Commonwealth Survey and additional measures recommended in this chapter, may be more sensitive and comprehensive indicators of health care for mothers and families than current measures. Further, as a more sensitive indicator, breastfeeding may prove to be effective in measuring quality of care in addition to

Breastfeeding in the United States

209

adequacy of care. Prenatal care, low birthweight, and infant morbidity and mortality have long been sentinel indicators of the nation’s health and reflective of the effectiveness of public health measures to improve the health of women and infants. After decades of improvement, however, the trend has slowed, especially for minority women and their children, while rates of breastfeeding initiation and duration for all women, including minorities, are on the rise. This suggests that some populations are receiving high-quality care, not merely adequate care.3 For breastfeeding rates to change, change must occur in many health care domains. As a sentinel indicator, breastfeeding, measured at the national or community level, could provide a unique link among several family health domains, including women’s health, pregnancy care, and well-child care. Preparation and support for breastfeeding reaches across many professional disciplines and components of both traditional medical and comprehensive public health care delivery system models for family health care. These include direct care providers and hospitals, insurers, and health plans as well as public agencies and programs such as WIC, Title V initiatives, and nontraditional providers, such as lactation consultants and home visitors. This matrix of health domains and cross-disciplinary services could provide a powerful health services model for the quality of family health. The 1998 National Breastfeeding Policy Conference, co-sponsored by the Centers for Disease Control and Maternal and Child Health Bureau, recommended that breastfeeding be placed at the center of the public health effort to improve the health and well-being of women, children, and families in this country (Slusser et al. 1999). To accomplish this, conference participants recommended forming alliances with existing public health 3 Initiation of early prenatal care has increased modestly from 81.3 percent in 1995 to 81.9 percent in 1996. Still, only 71 percent of black women, compared with 84 percent of white women, started care in the first trimester. Although the percentage of women who receive no prenatal care is decreasing, from 6 percent from 1983–91 to 4 percent in 1996, black women remain more likely to receive no prenatal care than white women. The percentage of low birthweight births rose for all births from 6.8 percent in 1985 to 7.4 percent in 1996. And, although the rate of low birthweight babies to black women has been declining since 1992, it still is more than double that among white women, and is closely tied to greater risk of neonatal mortality. Infant mortality, an important indicator of the well-being of infants, children and pregnant women, is associated with maternal health, quality of and access to medical care, socioeconomic conditions, and public health practices. It is one of six priority areas designated by the DHHS Initiative to Eliminate Racial and Ethnic Disparities in Health. In 1996, while infant mortality rates declined by 4 percent for all infants from the previous year, the rapid rate of decline that started 30 years ago has been slowing since the 1980’s. The racial disparity, where black infants are 2.4 times more likely to die in the first year of life than white infants, has remained the same. While neonatal deaths are also decreasing, the disparity between black and white infants remains, especially for low birthweight and premature infants and, in the postnatal period, for SIDS, pneumonia and influenza. These areas are especially susceptible to the benefits of breastfeeding.

210

Slusser, Lange

initiatives, programs, and activities wherever possible. Doing so would demonstrate that breastfeeding should not be perceived as separate from other national public health concerns, such as prenatal care or well-child care, but as an integral part of maternal and child health. Surgeon General Satcher has called for targeted programs to improve the health of woman and children who fall outside of the nation’s goals for health. Establishing breastfeeding indicators as sentinel events for family health is one step toward accomplishing that goal.

references American Academy of Pediatrics Committee on Practice and Ambulatory Medicine. (1997). Pediatrician’s Responsibility for Infant Nutrition. Pediatrics, 99, 749–750. Arnold, C., Makintube, S., & Istre, G.R. (1993). Day care attendance and other risk factors for invasive Hemophilus influenzae type B disease. American Journal of Epidemiology, 138, 333–40. Auerbach, K.G., & Guss, E. (1984). Maternal employment and breastfeeding: a study of 567 women’s experience. American Journal Diseases of Children, 138, 958–60 Aueberbach, K.G. (1992). One result of marketing: breastfeeding is the exception in infant feeding. Journal of Tropical Pediatrics, 38, 210–213. Balcazar, H., Trier, C., & Cobas, J. (1995). What predicts breastfeeding intention in Mexican-American and Non-Hispanic white women? Evidence from a national survey. Birth, 22, 74–80. Baldwin, E.N. (1997). A brief summary of breastfeeding and the law. http://www.lalecheleague.org/LawBills.html. Baldwin, L., Larson, E., Connell, F., Nordlund, D., Cain, K., Cawthorn, M., Byrns, P., & Rosenblatt, R. (1998). The Effect of Expanding Medicaid Prenatal Services on Birth Outcomes. American Journal of Public Health, 88(11), 1623–28. Baranowski, T., Rassin, D., Richardson, C., Brown, J., & Bee, D. 1986. Attitudes towards breastfeeding. Journal of Developmental and Behavioral Pediatrics, 7(6), 367–77. Barber-Madden, R., Petschek, M.A., & Pakter, J. (1987). Breastfeeding and the working mother: barriers and intervention strategies. Journal of Public Health Policy, 8, 531–41. Beaudry, M., Dufour, R., & Marcoux, S. (1995). Relation between infant feeding and infections during the first six months of life. Journal of Pediatrics, 126(2), 191–97. Black, RE., Lopez de Romana, G., Brown, R., Bravo, N., Bazalar, O., & Kanashiro, H. (1989). Incidence and etiology of infantile diarrhea and major routes of transmission in Huascar, Peru. American Journal of Epidemiology, 129, 785–99. Brach, C., & Scallet, L.J. (1996). Managed care’s implications for children and families: an overview of trends and issues. The Policy Resource Center, Annie E. Casey Foundation. Braveman, P., Egerter, S., Pearl, M., Marchi, K., & Miller, C. (1995). Problems associated with early discharge of newborn infants: a critical review of the literature. Pediatrics, 96(4), 716–26.

Breastfeeding in the United States

211

Brown, K., Black, R., Lopez De Ronana, G., & Creed de Kanashiro, H. (1989). Infantfeeding practices and their relationship with diarrheal and other diseases in Huascar (Lima), Peru. Pediatrics, 83(1), 31–40. Buescher, P., Roth, M.S., Williams, D., & Goforth, C.M. (1991). An evaluation of the impact of maternity care coordination on Medicaid birth outcomes in North Carolina. American Journal of Public Health, 81(12), 1625–69. California Department of Health Services. (1998). Medi-Cal Program Breastfeeding Benefits. California Department of Health Services, Office of Women’s Health. (1997). Profile of Women’s Health Status in California, 1984–1994. Carpenter R., Gardner, A., Jepson, M., Taylor, E., Salvin , A., Sunderland, R., Emery, J., Pursall, E., & Roe, J. (1983). Prevention of unexpected infant death. Evaluation of the first seven years of the Sheffield Intervention Programme. The Lancet, 8327, 723–27. Chaffee, E.E., & Greisheimer, E.M. (1969). Basic Anatomy and Physiology. 2nd ed. Philadelphia and Toronto: Lippincott. Chandra, R.K. (1986). Nutritional regulation of immunity and infection: from epidemiology to phenomenology to clinical practice. Journal of Pediatric Gastroenterology and Nutrition, 5, 844–52. Children’s Defense Fund. (1996). The State of America’s Children Yearbook 1996. Washington, DC: Children’s Defense Fund. Cohen, R., Mrtek, M.B., & Mrtek, R.G. (1995). Comparison of maternal absenteeism and infant illness rates among breast-feeding: U.S. women in two corporations. American Journal of Health Promotion, 10(2), 148–53. Cohen, R., & Mrtek, M.B. (1994). The Impact of Two Corporate Lactation Programs on Incidence and Duration of Breastfeeding by Employed Mothers. American Journal of Health, 8(6), 436–41. Cunningham, A.S., Jeliffe, D.B., & Jeliffe, E.F.P. (1991). Breastfeeding and health in the 1980’s: a global epidemiologic review. Journal of Pediatrics, 118 (5), 659–65. Dahl, I.L., Frufman, M., Hellberg, C., & Krabbe, M. (1991). Absenteeism because of illness at daycare centers and in three-family systems. Acta Paediat Scand, 80, 436–45. DeChateau, P. & Wiberg, B. (1977). Long term effect on mother-infant behaviour of extra contact during the first hour postpartum. Acta Paediatr, 66, 145–151. Dewey, K.G., Heinig, M.J., & Nommsen, L.A. (1993). Maternal weight loss patterns during prolonged lactation. American Journal of Clinical Nutrition, 59, 162–166. Dewey, K.G., Heinig, M.J., & Nommsen-Rivers, L.A. (1995). Differences in morbidity between breast-fed and formula fed infants. Journal of Pediatrics, 26(5, pt 1), 696–702. Dickson, V. (1997). CIGNA Health Care Corporation Working Well Mom’s Program. Personal communication. DiMatteo, M.R., Morton, S.C., Lepper, H.S., Dasmush, T.M., Carney, M.F., Pearson, M., & Katin, K.L. (1996). Cesarean childbirth and psychosocial outcomes: a metaanalysis. Health Psychology, 15(4), 303–314. Donnelly, B. (1994). Breastfeeding: ways to promote this natural resource. Contemporary OB/GYN, November, 107–118. Duckett, L. (1992). Maternal employment and breastfeeding. NAACOG’s Clinical issues in Perinatal and Women’s Health Nursing, 3(4), 701–12.

212

Slusser, Lange

Duncan, B., Ey, J., Hober, C., Wright, A., Martiniez, F., & Taussig, L. (1993). Exclusive Breastfeeding for at Least 4 Months Protect Against Otitis Media. Pediatrics, 91(5), 867–72. Essex, C., Smale P., & Geddis, D. (1995). Breastfeeding rates in New Zealand in the first 6 months and the reasons for stopping. New Zealand Medical Journal, 108, 355–57. Freed, G.L., Clark, S.J., Lohr, J.A., & Sorenson, J.R. (1995). Pediatrician involvement in breast-feeding promotion: a national study of residents and practitioners. Pediatrics, 96(3), 490–94. Gartner, L., Black, L.S., & Eaton, A.P. (1997). Breastfeeding and the use of human milk. Pediatrics, 100(6), 1035–1039. Goldman, A.S. (1993). The immune system of human milk: antimicrobial antiinflammatory and immunomodulating properties. Pediatric Infectious Disease Journal, 12, 664–71. Gray, R.H., Campbell, O., Apelo, R., Eslami, S.S., Zacur, H., Ramos, R.M., Gehret, J.C. & Labbok, M.H. (1990). Risk of ovulation during lactation. The Lancet, 335, 25–29. Hamosh, M., Dewey, K., Garza, C., Goldman, A., Lawrence, R., Picciano, M., Quandt, S., Rasmussen, K., & Rush, D. (1991). Nutrition During Lactation. Washington, DC: National Academy Press. pp 28–49 and 153–196. Hamosh, M. (1998a). Long-chain polyunsaturated Fatty Acids. Biology of the Neonate, 74(2), 106–120. Hamosh, M. (1998b). Early and late effects of breast-feeding: does breastfeeding really matter? Biology of the Neonate, 74(2), 177–189. Healthy Children 2000: National health promotion and disease prevention objectives related to mothers, infants, children, adolescents, and youth (1991). Washington, DC: U.S. Department of Health and Human Services. Hight-Laukaran, V., Rutsein, S., Peterson, A., & Labbock, M. (1996). The use of breastmilk substitutes in developing countries: the impact of women employment. American Journal of Public Health, 86(9), 1235–40. Hofferth, S. (1996). Child care in the United States today. The Future of Children. 6(2), 41–61. Horwood, L.J. & Fergusson, D.M. (1998). Breastfeeding and later cognitive and academic outcomes. Pediatrics, 101, e9. Howard, C.R., Howard, F.M., & Weitzman, M. (1993). The physician as advertiser: the unintentional discouragement of breastfeeding. Obstetrics and Gynecology, 81(6), 1048–51. Howard, C.R., Howard, F.M., Weitzman, M., & Lawrence, R. (1994). Antenatal formula advertising: another potential threat to breastfeeding. Pediatrics, 94(1), 102–04. Howell, E., & Ellwood, M. (1991). Medicaid and Pregnancy: Issues in Expanding Eligibility. Family Planning Perspectives, 32(2), 123–28. Howie, P., Forsyth, J., Ogston, S., Clark, A., & Florey, C. (1990). Protective effect of breastfeeding against infection. British Medical Journal, 300, 11–16. Institute of Medicine. (1991). Nutrition during Lactation. Subcommittee on Nutrition During Lactation, Committee on Nutritional Status During Pregnancy and Lactation Food and Nutrition Board. Washington, D.C.: National Academy Press.

Breastfeeding in the United States

213

International Code of Marketing of Breast-Milk Substitutes. World Health Organization. 1991: Geneva. Jacobson, S., Chiofo, L., & Jacobson, J. (1999). Breastfeeding effects on intelligence quotient in 4- and 11 year old children. Pediatrics, 103(5), e71. Johnson, D.L., & Swank, P.R. (1996). Breastfeeding and children’s intelligence. Psychiatric Reports, 79, 1179. Jones, E.G. & Matheny, R.J. (1993). Relationship between infant feeding and exclusion rate from child care because of illness. Journal of the American Dietetic Association, 93(7), 809–11. Kammerman, S.B., & Kahn, A.J. (1995). Starting right: How America neglects its youngest children and what we can do about it. New York: Oxford University Press. Kearney, M.H., & Cronenwett, L. (1991). Breastfeeding and employment. Journal of Obstetric, Gynecologic & Neonatal Nursing, 20(6), 471–480. Klaus, M.H. (1999). Perinatal Care in the 21st Century: Evidence that Supports Changing the Management for Both Mother and Infant. Conference Report: National Breastfeeding Policy Conference. Los Angeles: UCLA Center for Healthier Children, Families, and Communities. Kogan, M., Kotelchuck, M., Alexander, G., & Johnson, W. (1994). Racial Disparities in Reported Prenatal Care Advice from Health Care Providers. American Journal of Public Health, 84(1), 82–88. Koopman, J.S. (1985). Infant Formulas and Gastrointestinal Illness. American Journal of Public Health, 75(5), 477–480. Kritz-Silverstein, D., Barrett-Connor, E., & Hollenbach, K.A. (1992). Pregnancy and lactation as determinants of bone mineral density in postmenopausal women. American Journal of Epidemiology, 136, 1052–1059. Kurinji, N., Shiono, P.H., Ezrine, S.F. & Rhoads, G.G. (1989). Does maternal employment affect breastfeeding? American Journal of Public Health, 79(9), 1247– 50. Labbok, M.H. (1995). Costs of not breastfeeding in the US. Academy of Breastfeeding. News and Views, 1(1), 7. Lanting, C., Fidler, V., Huismand, M., Touwen, B., & Boersma, E. (1994). Neurological Differences between 9 year old children fed breastmilk or formula-milk as babies. The Lancet, 344, 1319–1322. Lawrence, R. (1999). Breastfeeding Advantages with a Focus on Child Development. Conference Report: National Breastfeeding Policy Conference. Los Angeles: UCLA Center for Healthier Children, Families and Communities. Lindberg, L.D. (1996). Women’s decisions about breastfeeding and maternal employment. Journal of Marriage and the Family, 58, 239–251. Long, S.H., & Marquis, M.S. (1998). The Effects of Florida’s Medicaid Eligibility Expansion for Pregnant Women. American Journal of Public Health, 88(3), 371–77. Lucas, A., Morley, R., Cole, T., & Leeson-Payne, C. (1992). Breastmilk and subsequent intelligence quotient in children born preterm. The Lancet, 339, 261–264. Moore, J.F., & Jansa, N. (1987). A survey of policies and practices in support of breastfeeding mothers in the workplace. Birth, 14(4), 191–195. National Center for Education in Maternal and Child Health. (1998). Racial and Ethnic Disparities in Maternal and Child Health. Policy Brief No. 3.

214

Slusser, Lange

Newcomb, P., Storer, B., Longnecker, M., Mittendorf, R., Greenberg, R., Clapp, R., Burke, K., Willett, W., & MacMahon, B. (1994). Lactation and a reduced risk of premenopausal breast cancer. New England Journal of Medicine, 330, 81–87. Newton, E. (1992). Breastfeeding/lactation and the medical school curriculum. Journal of Human Lactation, 8(3), 122–24. Nissen, E., Lilja, G., Widstrom, A.M., & Uvnas-Moberg, K. (1995). Elevation of oxytocin levels early post partum in women. Acta Obstet Gynecol Scand, 74(7), 530–3. O’Gara, C., Canahuti, J., & Martin, A.M. (1994). Every mother is a working mother: breastfeeding and women’s work. International Journal of Gynecology and Obstetrics, 47, S33–39. Paradise, J., Rockette, H., Colborn, D., Bernard, B., Smith, C., Kurs-Lasky, M., & Janosky, J. (1997). Otitis media in 2253 Pittsburgh-area infants: prevalence and risk factors during the first two years of life. Pediatrics, 99(3), 318–333. Patton, C., Beaman, M., Scar, N., & Lewinski, C. (1996). Nurses’ attitudes and behaviors that promote breastfeeding. Journal of Human Lactation, 12(2), 111–15. Paulley, J.G., & Gaines, S.K. (1993). Preventing day-care-related illnesses. Journal of Pediatric Health Care, 7, 207–211. Perez-Escamilla, R., Plooitt, E., L’onnerdal, B., & Dewey, K. (1994). Infant feeding policies in maternity wards and their effect on breastfeeding success: an analytical overview. American Journal of Public Health, 84(1), 89–97. Piper, S., & Parks, P. (1996). Predicting the duration of lactation: Evidence from a national survey. Birth, 23(1), 7–12. Ponka, A. (1991). Infections and other illnesses of children in day-care centers in Helsinki: incidence and effect of home and daycare center variables. Infection, 4, 230–36. Rassin, D.K., Markides, K.S., Baranowski, T., Richardson, C.J., Mikrut, W.D., & Bee, D.E. (1994). Acculturation and the initiation of breastfeeding. Journal of Clinical Epidemiology, 47(7), 739–46. Reamer, S.B., & Sugarman. M. (1987). Breastfeeding beyond six months: mothers’ perceptions of the positive and negative consequences. Journal of Tropical Pediatrics, 33, 93–97. Report of the Surgeon General’s Workshop on Breastfeeding and Human Lactation. (1984). U.S. Department of Health and Human Services, National Center for Education in Maternal and Child Health, Washington, DC. Report of the Second Surgeon General’s Workshop on Breastfeeding and Human Lactation. (1991). Washington, DC, National Center for Education in Maternal and Child Health. Rogan, W.J., & Gladen, B.C. (1993). Breast-feeding and cognitive development. Early Human Development, 31, 181–193. Romero-Gwynn, E., & Carias, L. (1989). Breastfeeding intentions and practice among Hispanic mothers in Southern California. Pediatrics, 84(4), 626–32. Ross, S.M., Loening, W.E., & Mbele, B.E. (1987). Breast-feeding support. South African Medical Journal, 72(5), 357–8. Rush, D., Leighton, J., Sloan, N.L., Alvir, J.M., & Garbowski, G.C. (1988). The National WIC Evaluation: Evaluation of the Special Supplemental Food Program for Women, Infants, and Children. II. Review of Past Studies of WIC. American Journal of Clinical Nutrition, 48, 394–411.

Breastfeeding in the United States

215

Ryan, A.S., & Martinez, G.A. 1989. Breast-feeding and the working mother: a profile. Pediatrics, 83(4), 524–531. Ryan, A.S. (1997). The resurgence of breastfeeding in the United States. Pediatrics, http://pediatrics.org/cgi/content/full/99/4/e12. Salariya, E.M., Easoton, P.M., & Cater, J.I. (1978). Duration of breastfeeding after early initiation and frequent feeding. Lancet, 25, 1141–1143. Scarr, S., Phillips, D., McCartney, K., Abbott-Shim, M. (1993). Quality of child care as an aspect of family and child care policy in the United States. Pediatrics, 91(1 Pt 2), 182–188. Scrimshaw, S., Engle, P., Arnold, L., & Haynes, K. (1987). Factors affecting breastfeeding among women of Mexican origin or descent in Los Angeles. American Journal of Public Health, 77(4), 467–70. Sempertegui, F., Estrella, B., Egas, J., Carrion, P., Yerori, L., Diaz, S., Lascano, M., Aranha, R., Ortiz, W., Zalabla, A., et al. (1995). Risk of diarrheal disease in Ecuadorian day-care centers. Pediatric Infectious Disease Journal, 14, 606–12. Slusser, W., Lange, L., & Thomas, S. (Eds.) (1999). Report of the National Breastfeeding Policy Conference. Los Angeles: UCLA Center for Healthier Children, Families and Communities. Slusser, W.M., & Powers, N.P. (1997). Breastfeeding update: immunology, nutrition, and advocacy. Pediatrics in Review, 18(4), 111–119. Spisak, S., & Gross, S.S. (1991). Second Follow-up Report. Surgeon General’s Workshop on Breastfeeding and Human Lactation. Washington, DC: National Center for Education in Maternal and Child Health. Super, C.M., Keefer, C.H., & Harkness, S. (1994). Child care and infectious respiratory disease during the first two years of life in a rural Kenyan community. Soc. Sci. Med, 38(2), 227–29. Teele, D. (1991). Strategies to control recurrent otitis media in infants and children. Pediatr Ann, 20(11), 609–616. Teele, D.W. (1989). Epidemiology of otitis media during the first seven years of life in children in greater Boston: a prospective, cohort study. Journal of Infectious Disease, 160(1), 83–94. Thompson, J.E., Walsh, L.V., & Merkatz, I.R. (1990). The history of prenatal care: cultural, social and medical contexts. In Merkatz, I.R., & Thompson, J.E. (eds.), New Perspectives on Prenatal Care. New York: Elsevier. Udon, N.U., & Betley, C.L. (1998). Effects of maternity-stay legislation on “drivethrough” deliveries. Health Affairs, 17(5), 208–215. U.S. Department of Education. (1995). National Center for Education Statistics. Child Care and Early Education Program Participation of Infants, Toddlers, and Preschoolers. Washington, DC: US Department of Education. U.S. Department of Health and Human Services. (1998). Child Health USA ‘97–’98. Washington, DC: US DHHS. U.S. Department of Health and Human Services. (1994). Vital and Health Statistics, Prenatal Care in the United States, 1980–94. Series 21, no 54. Washington, DC: U.S. DHHS. Victoria, C., Smith, P., Vaughan, J., Nobre, L., Lombardi, C., Teixera, A., Fuchs, S., Moreira, L., Gigante, L., & Barros, F. (1989). Infant feedings and deaths due to diarrhea. American Journal of Epidemiology, 129, 1032–41.

216

Slusser, Lange

Walker, M. (1992). Letter. Birth, 18(1), 57. Weimann, C.M., DuBois, J.C., & Berenson, A.B. (1998). Racial and ethnic differences in the decision to breastfeed among adolescent mothers. Pediatrics, 101(6), E11. Whittemore, A., Harris, R., Itnyre, J., & the Collaborative Ovarian CA Group. (1992). Characteristics relating to ovarian cancer risk: collaborative analysis of 12 US case-control studies. American Journal of Epidemiology, 136, 1184–1203. Williams, E.L. & Hammer, L.D. (1995). Breastfeeding attitudes and knowledge of pediatricians-in-training. American Journal of Preventive Medicine, 11(1), 26–33. Wong, R., Marwues, L., Pinoz, E., & Huffman, S. (1994). Guide to assessing the economic value of breastfeeding in El Salvador and suggestions for future modifications to the guide. World Health Organization. (1981). International Code of Marketing Breastmilk Substitutes. Geneva, Switzerland: World Health Organization. World Health Organization. (1989). Protecting, promoting and supporting breastfeeding: The special role of maternity services (a joint WHO/UNICEF statement). Geneva, Switzerland: World Health Organization. Wright, A.L., Holberg, C.J., Taussig, L.M., & Martinez, F.D. (1995). Relationship of infant feeding to recurrent wheezing at age 6 years. Archives of Pediatrics and Adolescent Medicine, 149, 758–763. Zeiger, R.S., Heller, S., Mellon, M., O’Connor, R., & Hamburger, R. (1986). Effectiveness of dietary manipulation in the prevention of food allergy in infants. Journal of Allergy Clinical Immunology, 78(1), 224–238.

8 Depressive Symptoms in Parents of Children Under Age 3: Sociodemographic Predictors, Current Correlates, and Associated Parenting Behaviors Karlen Lyons-Ruth, Rebecca Wolfe, Amy Lyubchik, and Ronald Steingard

Depressive symptoms are quite prevalent among adults in the United States. Eaton and Kessler (1981), reporting on a nationally representative sample of 3,000 households, found that 20.7 percent of adult women in the U.S. and 10.9 percent of men reported depressive symptoms over the level for possible clinical depression (for comparable data, see also Comstock and Helsing 1976). In contrast to rates of depressive symptoms, rates of current, psychiatrically defined depressive disorders, including major depressive disorder and dysthymia, are much lower (5–8 percent), in part because a psychiatric diagnosis requires the application of several simultaneous criteria, including a specified duration of depressive symptoms (American Psychiatric Association 1994; Robins and Regier 1991). The literature suggests that the stresses of parenting may increase the risk of parental depression. In addition, a large body of evidence indicates that parental depression increases the vulnerability of children to a variety of negative developmental outcomes, including poorer problem-solving skills, impaired social competence, and increased rates of psychiatric disorders (Beardslee et al. 1983; Cummings and Davies 1994; Downey and Coyne 1990; Zuckerman et al. 1987). Risks to children are found whether parental depression is assessed on the basis of current depressive symptoms or psychiatric diagnostic criteria for depression. In studies that have included both types of assessments, the chronicity of parental depressive symptoms over time has been a more sensitive predictor of child dysfunction than the parent’s current or prior history of depression meeting psychiatric diagnostic criteria (Campbell et al. 1995; Lyons-Ruth 1992). Parental depressive symptoms are therefore of interest because of both their prevalence and their potential as a marker for at-risk children. This chapter presents analyses of data from the Commonwealth Fund Survey of Parents with Young Children regarding the prevalence and correlates of depressive symptoms among fathers and mothers with at least one child under age three. Depressive symptoms were assessed via a brief 217

218

Lyons-Ruth, Wolfe, Lyubchik, Steingard

screening inventory, administered as part of the telephone survey of a representative stratified random sample of 2,017 parents. The inclusion of responses from a representative sample of fathers of young children constitutes a unique feature of the Commonwealth Survey.

parental depression and young children Prevalence of Adult Depressive Symptoms Epidemiologic studies of depression, whether they use symptom checklists or diagnostic interviews, indicate that depression is about twice as prevalent among women as among men. In addition, depression is more common in younger than in older adults (Eaton and Kessler 1981; Robins and Regier 1991). In Eaton and Kessler’s epidemiologic survey of the prevalence of depressive symptoms, young adults during the child-bearing years (ages 25–44) were more likely to be depressed than older adults; averaged across both genders, 19 percent of younger adults were depressed, compared with 8 percent of adults ages 65–74. Lack of employment also increased the rate of depression, from 16 percent among working adults to 26 percent among the unemployed. Adults in two-person households experienced the lowest rate of depression, 13 percent; the rate among adults in single-person households was 17 percent, while the rate in households of three or more (presumably most including children) was 18 percent. In Eaton and Kessler’s data, including both men and women, race had a major impact on the prevalence of depressive symptoms, with 33 percent of black women and 23 percent of black men reporting high symptom levels, compared with 20 percent and 10 percent, respectively, among their white counterparts and 21 percent and 11 percent of all U.S. adults. (These comparisons have been adjusted for gender or racial differences in education, income, employment, age, and marital status, so they represent the additional contributions of gender and race in and of themselves, beyond the contribution of other correlated factors.) Other sociodemographic factors were also related to the prevalence of depressive symptoms, including level of education, income, employment status, age, household size, and urban/rural residence. Low education doubled rates of depression with other factors controlled, as did low income. Low income was found to be a particularly potent risk factor for blacks, with almost 50 percent of blacks in the lowest income category reporting depression compared with 24 percent of low income whites and 18 percent of high income blacks. Similar sociodemographic factors have been shown to be related to increased prevalence of a psychiatric diagnosis of depression (Robins and Regier 1991). Eaton and Kessler (1981) did not gather information on parental status, but their data on increased depression during the child-rearing years

Depressive Symptoms in Parents

219

suggest that the presence of children may increase the prevalence of adult depression. Richman and colleagues (1982), using symptom scales, reported a 30 percent rate of depression among working-class mothers with 3-year-olds who were matched demographically with families of three-year-olds with behavior problems. Orr and James (1984) found that 35 percent of a sample of 240 urban mothers seen for pediatric health care reported depressive symptoms over the established level indicating possible clinical depression. In that sample, low income, minority status, and female head-of-household status substantially increased the risk for maternal depression to 50 percent for women without partners who had low incomes or were black. Brown and Harris (1975), examining rates of depressive diagnosis in an urban borough of London, also found that the number of young children under age 6 increased the rate of maternal depression, and the combination of working-class status, lack of a confiding relationship with a partner, and the presence of more than one young child increased the rate exponentially. Rates of depression may be increasing in a number of Western countries (Cross-National Collaborative Group 1992; Grof, 1997; Hagnell et al. 1982; Klerman and Weissman 1989), although some of the evidence has been controversial. The Epidemiological Catchment Area Study, a five-site study of rates of psychiatric disorders in the United States, found that younger adults at all sites reported higher lifetime rates of major depressive disorder than did older adults (Regier et al. 1988; Robins and Regier 1991). The controversy centered on whether older adults had truly been less vulnerable to depression over their lifetimes or whether they were simply less likely to recall episodes of depression that had occurred years or even decades earlier (Parker 1987). A recently published article based on longitudinal evidence from the Stirling County Study addresses the question of increase without relying on recall (Murphy et al. 2000). Examining reports of depression in the previous month by samples of adults selected and questioned in 1952, 1970, and 1992, that study found that the overall prevalence of depression remained steady across the three samples. A redistribution by age and gender in the 1992 sample, however, indicates an increase of depression among women under the age of 45. The rate of depression among these women, all born after the Second World War, was twice that for their age counterparts in the two earlier samples. Although parenting status was not reported, these findings again suggest that stresses associated with parenting may be significant contributors to the increased rates of depression among young adults, particularly women, of child-bearing age. In summary, whether depressive symptoms or depressive disorders are assessed, young adults under age 45 show elevated rates of depression, with women and those in households of three or more people showing particularly elevated rates. Indicators of social adversity, including lower

220

Lyons-Ruth, Wolfe, Lyubchik, Steingard

income or education and ethnic minority status, further increase the prevalence of clinically significant depressive symptoms to rates as high as 50 percent in some studies. Parental Depression as a Risk Factor for Child Outcomes A wide range of developmental studies document a relationship between parental depressive symptoms and impaired developmental outcomes for children. Children of depressed mothers may be at risk for a variety of cognitive deficits, including poor problem-solving skills and impaired school performance and intellectual acuity (Field et al. 1995; Gross et al. 1995; Sharp et al. 1995). For example, Sharp and colleagues (1995) found that 3year-old boys of postnatally depressed mothers scored one standard deviation below boys of nondepressed mothers on a measure of general cognitive functioning, even when controlling for behavior problems, birth weight, parental IQ, measures of home environment, parent-child interaction, and breastfeeding during infancy. However, Murray and co-authors (1996) did not replicate those results, finding instead that cognitive impairment is related to social class, schooling experience, and stimulation at home and is independent of maternal depression. Although there is as yet no consensus in the literature as to the role of maternal depression in cognitive impairment, researchers agree that deficits in stimulation and verbal interaction at home contribute to cognitive impairment in children. The research findings do show consensus that parental depression is associated with impairments in problem-solving, social competence, and overall adaptive functioning. Weissman and colleagues (1987) found that children of depressed parents displayed poorer overall functioning, poorer social competence, and more school problems than children of nondepressed controls. Likewise, Hammen and colleagues (1987) reported that children of unipolar mothers had greater social, behavioral, and academic impairment than children of nondepressed mothers. Nolen-Hoeksema and colleagues (1994) found that children of depressed mothers were more impaired in problem-solving and more prone to helpless behaviors than children of nondepressed mothers, as demonstrated through direct problem-solving observation, interviews, and teacher ratings. Field and co-authors (1996) reported that infants of depressed mothers engage in less exploratory play and have lower Bayley mental and motor scores than children of nondepressed mothers. Children of depressed mothers are also at risk for social skill deficits, including less active play, less exploratory play, and more generalized negative affect toward others (Nolen-Hoeksema et al. 1994). Increased negative affect when interacting with parents has been found in several studies of infants of depressed mothers (Cohn et al. 1986; Downey and Coyne 1990; Field 1995; Martinez et al. 1996). In addition, Martinez and colleagues found that infants of depressed mothers, when interacting with other mothers,

Depressive Symptoms in Parents

221

demonstrated less head orientation, facial expressiveness, and vocalizations and more fussiness than children of nondepressed mothers. Not surprisingly, given these multiple developmental correlates, children of depressed mothers face an increased risk for the development of psychopathology (Beardslee et al. 1983; Downey and Coyne 1990; Hammen et al. 1987). In a study by Weissman and colleagues (1984), children of depressed mothers were found to be at increased risk for psychological symptoms, treatment for emotional problems, suicidal behavior, school problems, and DSM-III psychiatric diagnoses. Children of depressed mothers faced three times the risk of any diagnosis than control children, with the most common diagnoses being major depression, attention deficit disorder, and separation anxiety. Similarly, in nine studies of child psychiatric diagnoses associated with maternal depression reviewed by Downey and Coyne (1990), children of depressed parents were more likely to receive a diagnosis than control children. In particular, children of depressed parents were three times more likely to be diagnosed with any affective disorder than controls. However, rates of nonaffective disorders were also significantly different between the groups, indicating that the risk for psychopathology was quite broad, and included risk for externalizing and aggressive behaviors as well as depression and anxiety. In addition to diagnosed psychopathology, children of depressed mothers are at increased risk for poor emotion regulation strategies, lower selfesteem, more fear and anxiety, and more externalizing and aggressive behaviors than children of nondepressed mothers (Downey and Coyne 1990; Gross et al. 1995; Alpern and Lyons-Ruth 1993; Lyons-Ruth et al. 1997). Goodman and co-authors (1994) have also reported that children of depressed mothers who use negative-affective statements and critical statements are at particular risk for low self-esteem and psychological impairment. Children of chronically depressed mothers may also be at greater risk for disorganized or insecure early attachment behaviors than children of nondepressed mothers. A study of low-income women found that high levels of maternal depressive symptoms were independently related to infant disorganized attachment patterns at 18 months of age, after controlling for six measures of sociodemographic adversity (Lyons-Ruth et al. 1990). Sociodemographic adversity did not predict disorganized attachment behaviors independent of depressive symptoms. A study of middle-class women also found that attachment disorganization was related to maternal depressive diagnosis among both infants and preschoolers (Teti et al. 1994). It should be noted, however, that depressed mothers in both studies had chronic symptoms and had been referred for treatment. Transient depressive symptoms have not been associated with attachment disorganization (Campbell et al. 1995; van IJzendoorn et al. 1999). Children of depressed mothers have also demonstrated various psychophysiological disturbances as early as the postpartum period. Abrams

222

Lyons-Ruth, Wolfe, Lyubchik, Steingard

and colleagues (1994) found that newborns of depressed mothers performed more poorly on the Brazelton Neonatal Assessment Scale as a whole, and on the orientation cluster in particular, than newborns of nondepressed mothers. In addition to inferior orientation, newborns also demonstrated less motor tone, activity, robustness, and endurance, and more irritability (see also Field et al. 1995; Zuckerman et al. 1990). Differences in the biobehavioral organization of infants of depressed parents are accompanied by differences in the organization of the brain. Dawson and colleagues (1997) found that 13-to-15-month-old infants of depressed mothers exhibited reduced relative left frontal electroencephalogram (EEG) activity when compared with infants of nondepressed mothers. Additionally, infants of currently depressed mothers displayed greater EEG abnormality than infants of mothers in remission from depression. Another study demonstrated similar findings when studying 3-to6-month-old infants of depressed mothers (Field et al. 1995). The authors interpreted these findings in light of prior research implicating the left frontal brain region in the experience of positive emotions and the right frontal brain region in the experience of negative emotions. The findings suggest that children as young as 6 months of age of depressed mothers experience more negative emotion. These findings are similar to findings of reduced relative left frontal activity in adolescents and adults with depressive disorders (Dawson et al. 1997). Field also demonstrated that infants of depressed mothers displayed dysregulation of sleep-wake cycles and abnormal norepinephrine levels at birth. However, Field (1995) also reported that in cases where the mother’s depressive symptoms subsided in the first few months, the infant’s symptoms also subsided. Much less data are available concerning the correlates of depressive symptoms in fathers. Developmental data have documented fathers’ somewhat different styles of interaction with their young children. Fathers tend to engage in more physical and active play with their children, to vary the level of arousal more, and to orient their children more to the outside environment than to the relationship itself (MacDonald and Parke 1986). Infants tend to seek their fathers for playful interactions but to seek their mothers preferentially for comfort and security when stressed or frightened (Clarke-Stewart 1978; Lamb 1982). Fathers’ involvement, nurturance, and intimacy are also related to later positive outcomes for children (Wagner and Phillips 1992). There is not yet a reliable body of data exploring how a father’s depression might alter interactions with his child and influence the child’s later development. However, Phares and Compas (1992) reviewed eleven studies that included depressed fathers as well as mothers; they concluded that although the sample sizes were small, children of depressed fathers were at risk for later emotional and behavioral problems. In summary, an extensive research literature documents associations between maternal depressive symptoms or diagnosis and impairments

Depressive Symptoms in Parents

223

in children’s social, adaptive, and emotional functioning. Links between maternal depression and child functioning have been demonstrated across the spectrum of child age from the postpartum period through adolescence and across parental income and educational levels. It has also been clearly demonstrated that maladaptive child behaviors associated with maternal depression are not confined to depressed or withdrawn behaviors but include anxious, hyperactive, and aggressive behaviors as well. Child outcomes seem to be more strongly related to the chronicity of maternal depressive symptoms than to whether or not the depressive symptoms meet criteria for a defined psychiatric disorder. Fathers’ depression is also related to children’s increased risk for emotional problems, but this risk has not been documented as systematically as the risk conferred by mothers’ symptoms. Given both the increased incidence of adult depressive symptoms during the child-bearing years and the risks these parental symptoms pose to children, the environmental contexts associated with parental depression have come under increased scrutiny by public health and social policy analysts. Yet no large-scale epidemiologic data have been available to anchor the myriad smaller-scale research findings to a representative sample of parents with young children in the United States. The Commonwealth Survey provides a data set that enables us to examine the independent contributions of sociodemographic factors and current circumstances in accounting for depressive symptoms among young parents. In addition, although our literature review does not address the mechanisms through which parental depressive symptoms translate into risks for child development, the Commonwealth Survey inquired about several parenting behaviors that may be related to both depressive symptoms and child development. These include positive behaviors, such as playing and reading to the child, and conflict-related behaviors, such as yelling and feeling aggravated with the child. The size and representative nature of the survey makes it possible to disentangle parenting behaviors that are correlates of parental depressive symptoms from parenting behaviors that are equally well explained by other aspects of the parents’ sociodemographic circumstances.

survey analysis Survey Questions For this analysis, we grouped selected questions from the survey into three broad classes according to their potential relation to parents’ depressive symptoms. The first class includes aspects of the family’s sociodemographic status and the parent’s past experiences that are likely to have been in effect for some time prior to the birth of the child. These more enduring circumstances, viewed as potential predictors of parental

224

Lyons-Ruth, Wolfe, Lyubchik, Steingard

depressive symptoms, include sociodemographic factors such as education, income, employment status, source of health insurance, receipt of government aid, single parenthood, teen parenthood, ethnicity (black, white, or Hispanic), and birth order of the child. Stressful prebirth events are also included, such as childhood experiences of emotional, physical, and/or sexual abuse, and the parent’s report of whether or not the pregnancy was wanted at the time it occurred. The second class of family circumstances includes events or perceptions surrounding or subsequent to the birth of the child that could be current correlates of the parent’s depressive symptoms and might be viewed as plausible causal contributors to the parent’s symptoms, although no such strong interpretation could be made on the basis of the survey data alone. These potential current correlates include parental and child physical health and sources of support for the parent. The parent’s confidence in his or her own parenting and general mental health are also viewed as potentially contributing current circumstances. The third class of survey responses includes current aspects of parentchild interaction potentially associated with the parent’s depressive symptoms. This class of responses includes the frequency of playing, reading, cuddling, and playing music, the maintenance of daily routines, the use of positive disciplinary techniques, and the frequency of yelling, spanking, hitting, and feeling aggravated with the child. Definition of Parental Depression To probe the presence of depressive symptoms, the Commonwealth Survey employed five questions from the well-validated Center for Epidemiologic Studies depressive symptom self-report inventory. Parents were asked to rate how often the following depressive symptom statements applied to them during the previous week: “I felt depressed,” “I enjoyed life” (scored in reverse), “I had crying spells,” “I felt sad,” and “I felt that people disliked me.” Response choices were “never,” “rarely,” “some of the time,” and “most of the time.” The latter two responses – “some of the time” and “most of the time” – were considered positive responses to a given symptom statement. Since few parents endorsed all five statements, the number of depressive symptoms was collapsed to a four-level variable (0/1/2/3+) for subsequent analysis. Initial bivariate analyses used this four-level variable. Roughly 20 percent of mothers and 10 percent of fathers endorsed two or more of the five screening questions positively. Since these percentages correspond with the prevalence of depressive symptoms needing clinical follow-up among adults of parenting age found in epidemiologic studies, we adopted this cut-off point for identifying potentially depressed parents needing further clinical follow-up.

Depressive Symptoms in Parents

225

It should be noted that our categorization of depressive symptoms is not equivalent to a clinical diagnosis of depression. A screening questionnaire like that used in the survey is not adequate to assess whether or not a respondent would be considered clinically depressed. An important goal of the study, however, was to assess whether or not a brief screening instrument for depressive symptoms can be used to identify a group of parents with characteristics similar to those emerging from studies using longer symptom inventories or clinical interviews. We would expect the current instrument to be sensitive to clinically depressed parents but would lack specificity, in that less depressed parents would also be included. Methods of Analysis Response categories with small numbers of respondents were collapsed to ensure adequate cell sizes, and statistics were weighted to correct for the oversampling of black and Hispanic families in the design of the survey, unless otherwise noted. Initial screening analyses were conducted to determine whether or not the prevalence of depressive symptoms varied by parent gender or child age in months (0–36 months). Bivariate analyses were then conducted within the predictor and current circumstances domains to identify predictors, and current circumstances associated with mothers’ or fathers’ depressive symptoms. Finally, a hierarchical logistic regression analysis was conducted using depressed status (as indicated by symptoms) as the dependent variable, to evaluate which predictors and current circumstances made independent contributions to the prediction of mothers’ or fathers’ depressed status. More enduring sociodemographic predictors were entered into the model first, then stressful prebirth events, then current circumstances. In the section on parent-child interactions, the analytical strategy was reversed, and parental depressive symptoms themselves were used as a predictor of parental behaviors. When parental depressive symptoms were significantly associated with a parental behavior, further logistical regression analyses were conducted to evaluate whether the parental behavior could be accounted for by more enduring sociodemographic circumstances or by stressful prebirth events or whether depressive symptoms made a unique contribution to the prediction of the behavior.

survey findings Of the 2,017 parents who participated in the survey, 1,320 were mothers and 697 were fathers. Consistent with other literature, reports of depression (defined as two or more symptoms) were significantly lower among fathers (12 percent) than among mothers (19 percent) (p < .0004). The child’s age

226

Lyons-Ruth, Wolfe, Lyubchik, Steingard

in months (1–36 months) was not significantly correlated with the number of depressive symptoms reported (p < .19). Three Profiles of American Families: The Prevalence of Depression As an aid to the reader, the findings regarding predictors of parental depression will be presented first in relation to three summary profiles of American families. Profile 1 consists of a one-parent household with an income of less than $20,000 (N = 289); profile 2 consists of a two-parent household with annual income between $30,000 and $40,000 (N = 257); and profile 3 consists of a two-parent household with income greater than $60,000 (N = 264). Parents who fit profile 1 are much more likely to be depressed (33 percent) than parents in profile 2 (13 percent) or profile 3 (6 percent), who have partners and more adequate incomes. Being a poor, single parent increases the odds of being depressed by 6.7 times for women and 5.8 times for men, respectively, compared with parents in profile 3, who are in two-parent households with incomes greater than $60,000. Parents in profile 2, in twoparent households with income between $30,000 and $40,000, have twice the odds (2.28 for women; 2.32 for men) of being depressed as do the more financially well-off families in profile 3, but these increased odds missed conventional levels of significance (p > .07; p > .08). The picture that emerges is that parental rates of depression increase dramatically as financial and personal resources available to parents decrease. Mothers’ rates rose from 7 percent to 15 percent and 34 percent across the three profiles, while fathers’ rates rose from 5 percent to 12 percent and 25 percent (Figure 8.1). It should be noted that data for single fathers may be unreliable because of the particularly small number of respondents in that category. Sociodemographic Circumstances as Predictors of Parental Depressive Symptoms Bivariate Analyses. Socioeconomic circumstances are associated with the number of depressive symptoms reported by mothers and fathers and with the proportion of parents who reach the criterion for depression. Parents experience significantly more depressive symptoms when their incomes are lower, when they have lower educational levels, when they are not living in a household with at least one full-time employed worker, when they do not have access to private health insurance, and when they are supported by government aid. Strength of association and significance levels for depressive symptoms are presented in Table 8.1. Percentages of parents

Single Parent <$20,000 Income

Two Parents $30,000-$40,000 Income

Two Parents >$60,000 Income

figure 8.1 Prevalence of Depression Among Mothers and Fathers in Three Types of American Families.

% Of Parents Who Are Depressed

50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0%

Mothers Fathers

228

Lyons-Ruth, Wolfe, Lyubchik, Steingard

table 8.1. Potential Predictors and Correlates of Parental Depressive Symptoms: Bivariate Statistics Number of depressive symptoms




Sociodemographic predictors Lower household income No education beyond high school Single parent Teenage parent No full-time worker in household No private health insurance Government aid Ethnicity Hispanic vs. white Black vs. white Child not firstborn Stressful prebirth events Severity of abuse in childhood Pregnancy mistimed Current circumstances Parent and Child Health Parent’s health not excellent Child’s health not excellent Partner and Community Support Responsible for child’s care Few sources of help No one to talk to Dissatisfied with amount of time with child Parenting Confidence Coping with parenthood Confidence in parenting Overall mental health

Mothers



Fathers



Strength of association

Strength of association

.21∗∗∗ .22∗∗∗ .20∗∗∗ .10∗∗∗ .21∗∗∗ .16∗∗∗ .22∗∗∗

.19∗∗∗ .14∗∗∗ n.s. n.s. .19∗∗∗ .11∗∗ .14∗∗∗

.13∗∗∗ .17∗∗∗ .10∗∗∗

.15∗∗∗ n.s. .11∗∗

.16∗∗∗ .16∗∗∗

.13∗∗∗ .15∗∗∗

.22∗∗∗ .13∗∗∗

.18∗∗∗ .13∗∗∗

n.s. .15∗∗∗ .11∗∗ .13∗∗∗

.09∗ .11∗ n.s. n.s.

.21∗∗∗ .07∗ .24∗∗∗

.20∗∗∗ n.s. n.s.

Note: Income was a 6-level variable and severity of abuse was a 3-level variable; all other variables had two levels. Strength of association was assessed for multilevel predictors by Pearson’s r and for two-level predictors by the Eta statistic with significance assessed by F-test. Maternal n’s varied from 1049 to 1311; paternal n’s from 539 to 688, except for ethnic comparisons where n’s varied from 947–989 (mothers) and from 517–581 (fathers). ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001.

Depressive Symptoms in Parents

229

meeting the standard for depressed status by each sociodemographic predictor are shown in Table 8.2. Mothers’ but not fathers’ depressive symptoms increase when they are single or teen parents (Table 8.1). The rate of depression for single mothers is 32 percent, compared with 15 percent among mothers with partners. Similarly, among teen mothers, 29 percent are depressed, compared with 18 percent of mothers 20 years of age or older (Table 8.2). In addition to socioeconomic level, another predictor of both mothers’ and fathers’ depressive symptoms is whether or not the child under 3 is the only child in the household. Parents experience more depressive symptoms when they have two or more children. It is not possible to test whether or not having a first child increases adults’ depressive symptoms, since all respondents to the survey were parents of at least one young child. Even so, this finding reinforces the finding by Brown and Harris (1975) that the number of young children increases the risk of major depressive episodes in women. The significant risk of depression created by the burdens of caring for children has received less attention in the broader literature on depression than it deserves. Ethnicity is also related to increased depressive symptoms. Rates of depression for white mothers, black mothers, and Hispanic mothers, respectively, are 16.0 percent, 29.5 percent, and 22.2 percent. Rates for white, black, and Hispanic fathers are 9.1 percent, 13.3 percent, and 21.5 percent, with Hispanic fathers diverging from the characteristic gender pattern. This finding deserves follow-up in subsequent studies, as it may provide a window on factors that lead some males not to report depressive symptoms. Multivariate Analyses. When all ten sociodemographic predictors are considered together in a logistic regression analysis, considerable overlap appears. Tables 8.3 and 8.4 display the final adjusted odds ratios and significance levels resulting from the model 1 multivariate logistic regression analyses for mothers (8.3) and for fathers (8.4). For mothers, educational level, rather than household income, best captures the variation in depression status associated with socioeconomic position. No other socioeconomic indicator adds to the prediction once educational level is entered. A mother’s odds of being depressed are twice as high if she had no education past high school. The birth order of the child also adds significantly to the prediction of the mother’s depression, beyond the variation in depression related to educational level. The mother’s odds of being depressed if her young child is not the firstborn are 1.65 times higher than the odds among mothers of firstborn children. The model 1 multivariate results for mothers indicate that birth order adds to the effect of low education on maternal depression, so that a mother

230

Lyons-Ruth, Wolfe, Lyubchik, Steingard

table 8.2. Percentage of Parents Endorsing Two or More Depressive Symptoms by Sociodemographic Factors, Prebirth Events, and Current Circumstances Sociodemographic and prebirth predictors and current circumstances Sociodemographic factors Household income Under $10,000 $10,000–19,999 $20,000–29,999 $30,000–39,999 $40,000–59,999 $60,000+ Education High school or less More than high school Partner status Single parent Two parents Parent’s age at child’s birth 19 or less 20 or older Employment status No full-time worker in household Full-time worker in household Health insurance Uninsured or government aid Private insurance Government aid Recipient of aid No government aid Ethnicity Black Hispanic White Birth order Laterborn Firstborn Prebirth stressful events Childhood abuse Physical/sexual abuse Verbal/emotional abuse only No abuse Pregnancy mistimed Pregnancy not wanted at this time Pregnancy well-timed

Percentage depressed Fathers

Mothers

37 15 13 12 10 5

35 25 19 15 17 8

16 8

26 12

18 11

32 15

21 12

29 18

31 10

32 15

17 9

26 15

17 8

25 13

13 22 9

29 22 16

14 9

23 14

24 16 10

30 25 15

18 9

25 16 (continued)

Depressive Symptoms in Parents

231

table 8.2. (Continued ) Sociodemographic and prebirth predictors and current circumstances Current circumstances: health Parent’s health Health not excellent Health excellent Child’s health Health not excellent Health excellent Current circumstances: social support Child care responsibility Primary responsibility for care Care shared equally with spouse Sources of emergency help Few or none Many Sources of child-rearing advice No one Relative, friend, professional Time with child Dissatisified with amount Right amount Current circumstances: parenting confidence Coping with parenthood Less well Very well Parenting confidence Less confident Very confident Overall mental health Poor Good to excellent

Percentage depressed Fathers

Mothers

16 7

25 12

20 10

26 16

20 11

20 15

15 10

27 15

20 12

38 18

13 7

23 14

17 8

27 12

12 12

27 18

22 12

54 17

with low education and a later born child has more than three times the odds of depression, compared with a mother with neither risk factor. This incremental effect of multiple risk factors on maternal depression rates is illustrated graphically in Figure 8.2, reporting on the three profiles of American families presented earlier. Figure 8.2 confirms that at both ends of the economic spectrum, rates of maternal depression are related to the birth order of the child, with 37 percent of poor, single mothers of laterborn children and 13 percent of the most advantaged mothers of later-born children classified as depressed, compared with 25 percent and 1.5 percent of the same two groups with firstborns.

Current circumstances Health Parent health not excellent Child health not excellent

Prebirth stressful events Abused in childhood Pregnancy mistimed

Sociodemographic predictors Lower income High school or less Single parent Teenage parent No full-time worker No private insurance Government aid Ethnicity Hispanic vs. white Black vs. white Child not firstborn

Model 2

Model 3

Model 4

0.97–1.34 1.29–3.18 0.86–2.45 0.48–2.44 0.66–2.16 0.55–1.45 0.51–1.62

0.72–1.92 0.89–2.18 1.09–2.49

1.14 2.02∗∗ 1.45 1.09 1.19 0.09 0.91

1.18 1.40 1.65∗ 1.56∗∗∗ 1.40

1.27 1.39 1.64∗

1.11 1.93∗∗ 1.40 1.11 1.23 1.01 0.83

1.25–1.96 .92–2.14

.77–2.09 .86–2.24 1.07–2.53

.94–1.31 1.23–3.04 .81–2.44 .50–2.47 .67–2.27 .62–1.62 .45–1.51

1.60 ∗∗ 1.40∗

1.79∗∗∗ 1.15

1.16 1.34 1.38

1.03 2.25∗∗∗ 1.25 0.95 1.20 0.94 1.16

1.14–2.24 0.99–1.97

1.28–2.52 0.81–1.62

0.76–1.78 0.88–2.03 0.97–1.96

0.89–1.19 1.55–3.24 0.77–2.03 0.50–1.82 0.77–1.89 0.59–1.48 0.71–1.88

1.30 1.31

1.70∗∗ 1.03

1.17 1.44 1.48∗

1.06 2.15∗∗∗ 1.36 0.90 1.10 1.00 1.13

(continued)

0.92–1.85 0.92–1.87

1.20–2.41 0.72–1.46

0.76–1.80 0.94–2.22 1.01–2.15

0.91–1.22 1.48–3.13 0.83–2.23 0.47–1.75 0.70–1.75 0.63–1.59 0.69–1.85

95 % 95 % 95 % 95 % Adjusted confidence Adjusted confidence Adjusted confidence Adjusted confidence odds ratio interval odds ratio interval odds ratio interval odds ratio interval

Model 1

table 8.3. Hierarchical Regression Model of Factors Associated with Mothers’ Depression: Sociodemographic Factors, Stressful Events, Health, Social Support, and Parenting Confidence

233

Model 2

Model 3

Model 4

1.06 1.44∗ 2.84∗ 1.29

0.68–1.66 1.03–2.00 1.11–7.24 0.92–1.80

0.59–1.49 0.93–1.85 1.02–6.61 0.81–1.60

1.51–3.00 0.88–2.86 0.99–3.55

0.94 1.31 2.60∗ 1.14 2.13∗∗∗ 1.58 1.88∗

95 % 95 % 95 % 95 % Adjusted confidence Adjusted confidence Adjusted confidence Adjusted confidence odds ratio interval odds ratio interval odds ratio interval odds ratio interval

Model 1

Note: n’s ranged from 1045 to 1036. Income was a 6-level variable; all others were two-level variables. Preliminary analyses established that those with missing data on income did not differ significantly in depression rate. However, income missing was included as a variable in all analyses to ensure that families with missing income data were not excluded from the analysis. The significant variation in education related to depression was in whether the parent had education beyond a high school diploma. Therefore, education was represented by the contrast with high school or less versus more than high school in all multivariate analyses. The significant variation in insurance status related to depression lay in whether the parent had access to private insurance compared to no insurance or government-sponsored insurance. This contrast was included in multivariate analyses. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001.

Social support Responsible for child’s care Few sources of help No one to talk to Dissatisfied with amount of time with child Parenting confidence Coping less well Less parenting confidence Poorer overall mental health

table 8.3. (Continued )

Later Born

Single Parent $20,000 Income

First Born

Later Born

Two Parents $30,000-$40,000 Income

First Born

Later Born Two Parents >$60,000

First Born

Depressed

Not Depressed

figure 8.2 When the Child Is Not Firstborn: Prevalence of Depression among Mothers by Family Profile.

% Of Mothers Who Are Depressed

50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0%

Depressive Symptoms in Parents

235

Given the importance of child birth order that emerged from these analyses, follow-up analyses were carried out to explore certain questions related to the impact of birth order and additional children in the household. These results were somewhat surprising. The number of additional children under 3 in the family is not related to increased maternal depression, regardless of whether there are one, two, or three or more children under three. If the family has only one child, and that child is under 3, the maternal depression rate is 13 percent. If there are two or more children under 3 but no older children, the rate is similar, at 15 percent. In contrast, the presence of one or more older children, in addition to one or more children under 3, is related to increased rates of depression. If there are two children, and one is over 3, the rate of maternal depression rises to 23 percent and remains near that figure regardless of the number of additional children. These results suggest that the child-rearing burden experienced by mothers is not a simple function of the number of children at home, but is also partly a function of the number of years devoted to parenting. These effects deserve further follow-up in a sample that is not restricted to parents with young children and includes parents-to-be prior to the birth of the first child. For fathers, income is the best predictor of depression status, and no other variable adds to the prediction of depression once income is accounted for (Table 8.4). Income is a six-level ordinal variable; following the common practice in developmental literature, this variable was specified as a linear variable in the multivariate analyses, with the ordinal categories treated as numerical values. Therefore, the odds ratio indicates that fathers’ odds of being depressed rise by an average of .31 for each step down in income from the top level of $60,000 or more, so that fathers at the lowest level have almost four times the odds of depression (3.86) as fathers at the highest income level. Profiles presented in Figure 8.1 roughly illustrate these effects on fathers’ depressed status. Stressful Prebirth Events and Parental Depression Other potential predictors of parental depression include the parent’s history of emotional, physical, or sexual abuse in childhood and the parent’s attitude toward whether or not the pregnancy was wanted at the time it occurred. Both circumstances exist prior to the baby’s birth and could be candidates for inclusion in a prenatal screening instrument for depression. Thirty-four percent of mothers and 27 percent of fathers report that they had not wanted the pregnancy to occur when it did. Among mothers, 15 percent report a childhood history of verbal or emotional abuse, and another 14 percent report a history of physical or sexual abuse, for an overall rate of abuse of 29 percent. Among fathers, corresponding rates are 13 percent and 11 percent, or 24 percent overall. Table 8.2 displays the

236

Current circumstances Health Parent health not excellent Child health not excellent

Prebirth stressful events Abused in childhood Pregnancy mistimed

Sociodemographic predictors Lower income High school or less Single parent Teenage parent No full-time worker No private insurance Government aid Ethnicity Hispanic vs. white Black vs. white Child not firstborn

Model 2

Model 3

Model 4

1.03–1.67 0.86–3.08 0.37–2.67 0.05–12.80 0.93–6.42 0.38–2.10 0.35–1.96

0.82–3.29 0.62–4.08 0.89–2.95

1.31∗ 1.62 0.990 0.78 2.44 0.90 0.83

1.64 1.59 1.62 2.23∗∗ 1.54

1.73 1.52 1.31

1.18 1.63 —a 2.09 1.46 1.13 0.86

1.22–4.07 0.84–2.85

0.91–3.31 0.56–4.15 0.70–2.43

0.91–1.53 0.85–3.10 —a 0.15–28.86 0.54–3.95 0.52–2.48 0.38–1.92

1.75 1.59

1.92∗ 1.28

1.63 1.69 1.18

1.30∗ 1.78 —a 2.28 0.90 1.07 0.70

0.92–3.30 0.82–3.11

1.01–3.63 0.66–2.50

0.82–3.25 0.59–4.86 0.61–2.27

1.00–1.69 0.91–3.48 —a 0.15–34.15 0.28–2.85 0.47–2.43 0.30–1.63

1.70 1.55

1.90∗ 1.22

1.70 1.75 1.27

1.32∗ 1.73 —a 2.16 0.93 1.08 0.71

(continued)

0.89–3.25 0.78–3.08

1.00–3.63 0.62–2.41

0.84–3.43 0.60–5.11 0.64–2.50

1.01–1.72 0.88–3.39 —a 0.14–32.29 0.28–3.03 0.47–2.47 0.30–1.64

95 % 95 % 95 % 95 % Adjusted confidence Adjusted confidence Adjusted confidence Adjusted confidence odds ratio interval odds ratio interval odds ratio interval odds ratio interval

Model 1

table 8.4. Hierarchical Regression Model of Factors Associated with Fathers’ Depression: Sociodemographic Factors, Stressful Events, Health, Social Support, and Parenting Confidence

237

Model 2

Model 3

Model 4

3.15 2.03∗ 1.59 1.88

0.75–13.16 1.09–3.76 0.17–14.81 0.86–4.09

0.74–13.13 1.08–3.73 0.12–12.74 0.80–3.9

0.72–2.43 0.41–5.17 0.22–4.19

3.11 2.00∗ 1.21 1.77

1.32 1.46 0.95

95 % 95 % 95 % 95 % Adjusted confidence Adjusted confidence Adjusted confidence Adjusted confidence odds ratio interval odds ratio interval odds ratio interval odds ratio interval

Model 1

Note: n’s ranged from 493 to 483. Income was a 6-level variable; all others were two-level variables. Preliminary analyses established that those with missing data on income did not differ significantly in depression rate. However, income missing was included as a variable in all analyses to ensure that families with missing income data were not excluded from the analysis. The significant variation in education related to depression was in whether the parent had education beyond a high school diploma. Therefore, education was represented by the contrast high school or less versus more than high school in all multivariate analyses. The significant variation in insurance status related to depression lay in whether the parent had access to private insurance compared to no insurance or government-sponsored insurance. This contrast was included in multivariate analyses. a Variable dropped from regression due to collinearity. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001.

Social Support Responsible for child’s care Few sources of help No one to talk to Dissatisfied with amount of time with child Parenting Confidence Coping less well Less parenting confidence Poorer overall mental health

table 8.4. (Continued)

238

Lyons-Ruth, Wolfe, Lyubchik, Steingard

percentage depressed by type of abuse. In other analyses, both types of abuse are collapsed into a single category. In bivariate analyses, both prebirth predictors were significantly related to the number of depressive symptoms endorsed by mothers and fathers (Table 8.1). Childhood abuse history and mistimed pregnancy were then added to the sociodemographic predictors in the model 2 logistic regression analyses to evaluate whether they added to the prediction of depression independently of the sociodemographic predictors. As shown in Tables 8.3 and 8.4, for both mothers and fathers, a history of abuse in childhood predicts depressed status independently of all sociodemographic indicators. The odds of being depressed are 56 percent higher for mothers and 123 percent higher for fathers who had experienced physical, sexual, or emotional abuse during childhood. In relation to the three profiles of American families described earlier, these results mean that parents in all three types of families, from the poorest to the most advantaged, have an added risk of depression if they have suffered physical, sexual, or emotional abuse in childhood. Among poor, single mothers, if the mother experienced abuse in childhood, depression rates rise from 34 percent to 46 percent. Among working-class mothers, rates rise from 15 percent to 23 percent, and among more advantaged mothers rates increase from 7 percent to 10 percent. The numbers of abused fathers in the three family types are too small to yield reliable data. Current Circumstances and Parental Depressive Symptoms Three types of current circumstances were analyzed in relation to parental depression: parent and child health; support for parenting from the respondent’s partner, extended family, and community; and the respondent’s confidence as a parent, coping skills, and general mental health. Since many response categories for these variables have small cell sizes, all parental responses have been dichotomized (Table 8.2). Parent and Child Health. The health-related variables analyzed include the parent’s rating of the parent’s own and the child’s physical health. Parents were asked to rate their own and their child’s overall health on a four-point scale from “excellent” to “poor.” Less than 7 percent used the “fair” or “poor” categories in response to either question. Therefore, responses have been collapsed to two categories: “excellent” and “not excellent.” Bivariate statistics for the parent and child health variables are displayed in Tables 8.1 and 8.2. The state of the parent’s health is significantly correlated with depressive symptoms for both mothers and fathers. Less-than-excellent parental health increases the odds of parental depression by 2.38 for mothers and 2.46 for fathers. The parent’s depressive symptoms are also significantly

Depressive Symptoms in Parents

239

related to the child’s health. Less-than-excellent child health increases the odds of parental depression by 1.86 for mothers and 2.25 for fathers. Although only 2 percent of parents rated their child’s health as fair to poor, mothers of those children incurred a particularly heavy cost, with 45 percent reporting depression. Partner and Community Support for Parenting. Survey questions regarding social support include whether or not the parent does more of the daily child care than his or her partner, how many friends or relatives can be counted on to help out in an emergency, whether or not the parent feels that he or she has someone to talk to about raising the child, and whether or not the parent feels that he or she is able to spend enough time with the child. Both fathers and mothers who report having few or no friends or relatives to help out in an emergency experience more depressive symptoms (Tables 8.1 and 8.2). When mothers have few or no sources of help, their odds of depression are 2.13 times greater; for fathers, the odds are 1.61 times greater. In addition, for mothers but not fathers, the lack of someone to talk to about child rearing is a significant contributor to depressive symptoms. Having no one to talk to was a relatively rare circumstance, affecting only 2 percent of parents. With 38 percent of these mothers reporting depression, however, the finding underscores the high risk of depression among mothers without support from partners or extended family. Parents were also asked whether the amount of time they spend with their young child is about right or whether they need more or less time with the child. Although not directly explored in the survey, work hours constitute the most prevalent constraint on time for parenting. Because the demands of the work environment on young parents are regulated by the formal and informal policies of the larger community, allowing parents adequate time with their young children constitutes a central form of community support or lack of support for parents. It is notable that among all parents, only 37 percent reported that they spend the right amount of time with their child, while 38 percent of parents said they needed “a lot more time,” and another 19 percent said they needed “a little more time.” Only 6 percent were unsure or felt they needed less time with their child. Mothers with increased depressive symptoms were more likely to report that they do not have enough time to spend with their young children. Among mothers who said they are satisfied with the amount of time with the child, 14 percent were depressed, compared with 23 percent of mothers who were dissatisfied. Fathers’ feelings that they need more time with their children are not related to depression. Seventy-five percent of fathers said that they do not have enough time to spend with their child. For fathers, having the primary responsibility for the child’s care is associated with increased depressive symptoms (Table 8.4). Only 11 percent of

240

Lyons-Ruth, Wolfe, Lyubchik, Steingard

fathers, compared with 73 percent of mothers, are primarily responsible for child care. Twenty percent of fathers with primary child care responsibility were depressed, compared with only 11 percent of other fathers (odds ratio 2.11). It is notable that the rates of depression among fathers and mothers equalize when the parent in question is shouldering the primary burden of child care. Among the 101 fathers with daily care responsibility for a young child, 20 percent were depressed, equaling the 20 percent depression rate among the 823 mothers with primary child care responsibility. This is a unique finding of the current survey, since few studies have explored the correlates of depression among fathers of young children. When the six health and social support variables were added to the sociodemographic and prenatal predictors in the model 3 logistical regression analysis, both maternal and child health made unique contributions to the prediction of mothers’ depression. In addition, having no one to talk to about parenting and few sources of supportive help in an emergency also independently increased the odds of maternal depression (Table 8.3). Having few sources of help also remained a significant contributor to fathers’ depression after other predictors were accounted for, doubling the odds of depression among fathers (Table 8.4). Confidence in Parenting. Confidence and coping as a parent are psychological resources with a somewhat different conceptual relation to depression than more external contributing circumstances; they were therefore added as a separate step in model 4 of the regression. Variables include parents’ reports about how well they believe they are coping with the demands of parenthood, how confident they feel about caring for the baby, and their overall evaluation of their mental health. Mothers’ depressive symptoms are significantly related to all three psychological resources questions (Tables 8.1 and 8.2). For fathers, only coping with parenthood is related to depressive symptoms. One might expect that the three psychological resources questions would be redundant and overlap with one another in predicting maternal depression. In fact, each question made a significant, unique contribution to the prediction of depression beyond the variance already accounted for by other sociodemographic factors, prebirth events, health, and support variables. With each question, a less optimal response more than doubled the odds of maternal depression (Table 8.3). For fathers, the coping variable also contributed uniquely to the cumulative prediction of depression and doubled the odds of depression (Table 8.4). In summary, for both mothers and fathers, diverse current circumstances are associated with rates of depression among parents with children under age three. The current context of child rearing, including parent and child health, supports for parenting, and parental confidence, has a more pervasive influence on mothers’ depressive symptoms than on fathers’.

Depressive Symptoms in Parents

241

This is most likely related to the continued centrality of the parenting role for women and inequalities between parents regarding who shoulders the largest share of child-rearing responsibilities. Positive and Negative Parenting Behaviors and Parental Depressive Symptoms In the final set of analyses, parents’ depressive symptoms were examined in relation to their interactions with their young children. The frequency of positive parent-child interactions, including cuddling, playing with, reading to, and singing or playing music to the child, were analyzed, as were parents’ reports of providing regular routines for napping, mealtimes, and bedtimes. Conflict-related or limit-setting interactions were also examined, including how often the parent feels frustrated with the child, yells at the child, spanks the child, and hits, slaps, or shakes the child. For parents with children over 18 months of age, the analysis also explored parents’ use of positive limit-setting techniques, including giving explanations, giving a time out, or taking away a toy or treat as part of dealing with the child’s misbehavior. Causal inferences cannot be made on the basis of survey data alone, yet the accumulated body of longitudinal developmental findings suggests that these interactive correlates of depressive symptoms can plausibly be seen as potential outcomes of the parent’s depression. It should be noted that parents’ depressive symptoms were used as a predictor variable in this analysis, rather than as an outcome variable as in the preceding analyses. The number of depressive symptoms, a variable with four levels (0/1/2/3+), was used to generate all odds ratios, so that the odds ratios refer to the increase in likelihood of a parental behavior that occurs with each additional depressive symptom reported by the parent. Analyses were conducted both on the multilevel frequency ratings for parental behaviors and on dichotomized versions of those variables, yielding similar results. Because of the very small cell sizes for some response categories, Table 8.5 presents the odds ratios for the dichotomized parental behavior variables. The positive interaction variables were dichotomized as follows: (1) play with child once a day (84 percent)/ less than once a day; (2) read to child several times a week or more (59 percent)/less than several; (3) hug or cuddle child once a day or more (88 percent)/less than once a day; (4) sing or play music several times a week (79 percent)/less than several; (5) bedtime, naptime, and mealtime routines usually the same every day (68 percent); one or more routines changes from day to day. The conflict-related interactions were dichotomized as follows: (1) feel frustrated or aggravated with child two or more times a day (27 percent)/ less often; (2) yell at child sometimes or often (40 percent)/ rarely or never; (3) spank child sometimes or often (17 percent)/rarely or never; (4) hit, slap, or shake child sometimes or rarely (7 percent)/never;

242

Lyons-Ruth, Wolfe, Lyubchik, Steingard

table 8.5. Parent-Child Interactions Associated with the Level of the Parent’s Depressive Symptoms: Bivariate Odds Ratios Odds ratios

Parent-child interactions

Mothers’ depressive symptoms

Fathers’ depressive symptoms

Positive interactions Playing less often with child Providing fewer routines for child Reading less often to child Cuddling child less often Playing music or singing to child less often Fewer positive interactions overall

1.33∗∗ 1.38∗∗∗ 1.36∗∗∗ 1.22∗ n.s. 1.52∗∗∗

1.40∗∗ 1.40∗ 1.34∗∗ 1.34∗ n.s. 1.48∗∗∗

1.32∗∗∗ 1.30∗∗∗ 1.30∗∗∗ n.s. 1.25∗∗

1.66∗∗∗ 1.32∗∗ n.s. 1.40∗ 1.43∗∗

n.s. n.s. n.s.

1.41∗∗ n.s. n.s.

Conflict-related interactions Asked of all parents Feeling aggravated with child Yelling at child Spanking child Hitting, slapping, or shaking child Negative interactions overall Asked of parents of children 18 + months Explaining misbehavior less often Taking away a toy or treat less often Giving time out less often

Note: All parent-child interaction variables were two-level; depressive symptoms was a fourlevel variable (0/1/2/3+). Odds ratios refer to the increased likelihood of less positive interaction with each additional depressive symptom. For mothers, n’s ranged from 1284 to 1320 (641 to 645 among mothers of toddlers only); for fathers, from 678 to 697 (361 to 365 among fathers of toddlers only). ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001.

(5) explain to child often why his or her behavior was not appropriate (66 percent)/sometimes, rarely, never; (6) take away a toy or treat sometimes or often (52 percent)/rarely, never; (7) give a time-out sometimes or often (66 percent)/rarely, never. Positive Parent-Child Interactions. Mothers’ and fathers’ levels of depressive symptoms affect both positive and negative aspects of their interactions with their children. As depressive symptoms increase, both mothers and fathers play with their children less, read to their children less, and hug or cuddle their children less. Both mothers and fathers are also less likely to provide regular routines for mealtimes, naptimes, and bedtimes if they are more depressed (Table 8.5).

Depressive Symptoms in Parents

243

A composite overall positive interaction dichotomous score (fewer positive interactions/more positive interactions) was also created. Parents were classified as having fewer positive interactions if they scored low on two or more of the following positive interaction variables: (1) maintained regular routines for the child (yes/no); (2) read to child at least several times a week (yes/no); (3) played with child once a day or more (yes/no); and (4) hugged or cuddled child more than once a day (yes/no). (Singing was omitted because singing was not felt to play as central a role in early parentchild interaction as the other variables, based on current developmental literature.) The likelihood of low scores on two or more forms of positive interaction increases significantly as depressive symptoms increase and is related to depressive symptoms more strongly than any single parenting behavior analyzed separately (Table 8.5). For each additional depressive symptom endorsed, the likelihood that the parent will be low on two or more forms of positive interaction increases by 52 percent for mothers and 48 percent for fathers (Figure 8.3). Multiple logistic regression analyses were then performed to assess whether the associations between depressive symptoms and positive parenting behaviors could be accounted for by other sociodemographic or prebirth family characteristics associated with depression. When the parents’ positive interactive behaviors are adjusted for the effects of all these significant predictors of parental depression, the importance of depressive symptoms as a unique correlate of mothers’ less positive parenting behaviors is further underscored (Table 8.6). For fathers, depressive symptoms no longer add to the prediction of less positive interaction once sociodemographic and prebirth events are accounted for. For each depressive symptom endorsed by mothers, the odds of lack of routines rises by an average of 31 percent, the risk of reading to the child less than several times a week rises by 21 percent, and the risk of playing with the child less than once a day rises by 33 percent. The risk of engaging in low levels of two or more positive behaviors increases by 40 percent for each additional depressive symptom. The contribution of maternal depression to decreased levels of positive interactions with the child across income levels can be illustrated using the three types of American families profiled earlier. Among poor, single mothers, 48 percent of depressed mothers compared with 33 percent of nondepressed mothers reported low rates of two or more forms of positive interaction. Among middle-income mothers with partners, the rates were 30 percent and 21 percent, respectively. Among the most advantaged mothers, rates were 20 percent and 11 percent. Conflict-Related Parent-Child Interactions. Conflict-related behaviors are also associated with the level of the parents’ depressive symptoms (Table 8.5). Both mothers and fathers are more likely to feel aggravated by

244

0%

5%

10%

15%

20%

25%

30%

35%

0

2

3+

Number of Depressive Symptoms

1

Mothers Fathers

figure 8.3 Percentage of Parents Who Feel Aggravated with Their Young Children Two or More Times a Day by Parental Depressive Symptoms.

% of Parents Who Feel Aggravated with the Child Two or More Times a Day

40%

45%

50%

Depressive Symptoms in Parents

245

table 8.6. Predictors of Less Optimal Parent-Child Interactions: Multivariate Analyses

Parent-child interactions Mothers Less play Lower income Black parent Depressive symptoms Fewer routines Hispanic parent Black parent Depressive symptoms Less reading Lower income Teen parent Black parent Hispanic parent Laterborn child Depressive symptoms Fewer positive interactions overall Lower income Teen parent No full time worker in household Black parent Hispanic parent Depressive symptoms Feeling aggravated Depressive symptoms Yelling at child Teen parent Depressive symptoms Spanking child Lower income Teen parent Black parent Depressive symptoms More negative interactions overall Teen parent Depressive symptoms

N 1075

1076

1070

1066

1061 1069

1066

1049

Adjusted odds ratio

95 % confidence interval

1.21∗ 1.86∗ 1.33∗∗

1.00–1.46 1.09–3.16 1.09–1.62

1.86∗∗ 1.78∗∗ 1.32∗∗∗

1.23–2.81 1.13–2.77 1.11–1.55

1.21∗∗∗ 4.30∗∗ 2.00∗∗ 2.31∗∗∗ 1.95∗∗∗ 1.20∗

1.04–1.41 2.09–8.82 1.27–3.17 1.51–3.54 1.36–2.81 1.01–1.42

1.21∗∗ 3.57∗∗ 0.58∗ 3.18∗∗∗ 2.46∗∗∗ 1.40∗∗∗

1.03–1.42 1.69–7.55 0.33–1.01 1.97–5.11 1.59–3.82 1.17–1.67

1.34∗∗∗

1.13–1.59

0.25∗∗∗ 1.39∗∗∗

0.12–0.53 1.17–1.64

1.21∗ 0.35∗ 1.72∗ 1.38∗∗∗

1.03–1.42 0.13–0.93 1.02–2.86 1.14–1.68

0.30∗∗ 1.32∗∗

0.14–0.62 1.10–1.57 (continued)

246

Lyons-Ruth, Wolfe, Lyubchik, Steingard

table 8.6. (Continued)

Parent-child interactions Fathersa Feeling aggravated Childhood abuse Depressive symptoms Hitting, slapping, or shaking child Depressive symptoms

N 502

499

Adjusted odds ratio

95 % confidence interval

1.77 ∗ 1.75∗∗∗

1.00–3.13 1.30–2.34

1.41∗

1.02–1.95

Note: Only significant predictors are shown. Models included all sociodemographic factors and stressful prebirth events listed in Tables 8.3 and 8.4. All parent-child interaction variables were two-level; depressive symptoms was a four-level variable (0/1/2/3+); predictor variables were two-level, except for income, which was a six-level variable, as shown in Table 8.2. Odds ratios for depressive symptoms refer to the increased likelihood of less positive interaction with each additional depressive symptom. a Depressive symptoms did not account for independent variance in fathers’ positive behaviors. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001.

their young child’s behavior twice a day or more if they experience more depressive symptoms. These effects are particularly strong for fathers, for whom each additional depressive symptom increases the odds of frequent aggravation with the child by an average of 54 percent (Figure 8.4). Both parents are also more likely to yell at the child as their depressive symptoms increase. In addition, mothers are more likely to spank the child as depressive symptoms increase, while fathers are more likely to hit, slap, or shake the child. Among mothers with no depressive symptoms, 11 percent reported spanking their child sometimes or often, compared with 20 percent of mothers with three or more symptoms. Among fathers with no depressive symptoms, 5 percent reported hitting, slapping, or shaking the child, while 10 percent of fathers with three or more depressive symptoms reported engaging in those behaviors. A composite negative interaction dichotomous score was also created and scored as positive if the parent engaged in one or more of the following behaviors: (1) aggravated with the child twice a day or more (yes/no); (2) sometimes or often yells at the child (yes/no); (3) sometimes or often spanks the child (yes/no); and (4) has hit, slapped, or shaken the child (yes/no). The composite negative interaction score is also reliably related to the level of depressive symptoms for both mothers and fathers (Table 8.5). For each additional depressive symptom, the odds that the parent shows at least one form of negative interaction rises by 25 percent for mothers and 43 percent for fathers.

247

1

2

3+

Number of Depressive Symptoms

0

Mothers Fathers

figure 8.4 Percentage of Parents Who Are Low on Two or More Types of Positive Interaction by Parental Depressive Symptoms.

% of Parents Who Are Low on 2 or More Forms of Positive Interaction

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

248

Lyons-Ruth, Wolfe, Lyubchik, Steingard

Parents with children 18 months of age or older were also asked how frequently they use non-aggressive forms of discipline, such as explaining to the child why its behavior is not appropriate, taking away a toy or treat, or giving a time-out (asking the child to take a break from the current activity). Only the frequency of giving explanations (often/not often) is related to depressive symptoms, and this relation is reliable for fathers only (Table 8.5). Among nondepressed fathers, 66 percent said they give explanations often, while only 49 percent of depressed fathers explain often. None of the other positive disciplinary practices is significantly related to depressive symptoms. However, inspection of the data indicates that two contradictory trends may have affected the data. Nondepressed parents may be more likely to use positive forms of discipline in preference to negative forms. However, depressed parents may be more likely to discipline the child frequently, leading to higher rates of all disciplinary practices. In fact, depressed parents reported both using time-outs and taking away toys or treats slightly more often than nondepressed parents. In future surveys, it will be important to ask first how often parents discipline the child each day and then ask what proportion of their disciplinary interventions involve a particular practice, so that rates of all disciplinary practices can be adjusted for the overall rate of disciplinary interventions. Multiple logistic regression analyses were then conducted on all conflictrelated parental behaviors to evaluate whether the set of sociodemographic and prebirth predictors could account for the associations between depressive symptoms and conflict-related parenting behaviors. Again, the results underscore the central role of depressive symptoms in predicting negative parent-child interactions (Table 8.6). These results are particularly dramatic because they pertain to parents’ negative interactions with very young children. In addition, the data are stratified and weighted to represent the entire U.S. population of parents with young children. Depressive symptoms are the most consistent predictors of parents’ negative behaviors toward their young children. The odds of mothers’ yelling, spanking, feeling aggravated, and composite negative behavior rise by 31 to 38 percent with each additional depressive symptom endorsed. In fathers, the odds of hitting, slapping, or shaking the child increase by 47 percent for each depressive symptom reported and the odds of feeling aggravated with the child twice a day or more rise by 74 percent. With depressive symptoms as part of the model, socioeconomic and prebirth stressful events account for relatively little additional variance in conflict-related parental behavior. Teen mothers are less likely to yell and less likely to engage in any form of negative interaction. Black mothers and mothers with lower incomes are more likely to spank their children, but they are not more likely to display increased rates of other negative behaviors. Teen mothers’ low rates of negative behaviors, combined with their low rates of maintaining routines, reading, and overall positive behaviors

Depressive Symptoms in Parents

249

(Table 8.6), may indicate relatively low involvement with the child in general. This, in turn, may be attributable to receiving help from grandmothers and extended family or other reasons. For fathers, a history of childhood abuse also increases the odds of frequent aggravation by 74 percent, independent of the effect of depressive symptoms. Therefore the influence of childhood abuse on fathers’ behavior is not mediated by depressive symptoms, as appears to be the case among mothers. These survey results converge well with the results of smaller-scale studies that have found increased irritability and more frequent negative interactions with children among parents across income levels who report depressive symptoms or are clinically depressed. Parental depressive symptoms are more frequent under conditions of economic stress or childhood abuse, so it is not surprising that a number of studies have found associations between low income or abuse history and negative parenting behaviors (see McLoyd 1990). The Commonwealth Survey provides the first available data on an adequately large and representative sample of U.S. parents to demonstrate that previous associations in the research literature between low income, abuse history, or unwanted pregnancy and negative parental behaviors are mediated in part by the level of the parents’ current depressive symptoms. It is important to note, however, that childhood history was represented in the survey only by a single question on childhood abuse, while the contributions of sociodemographic factors were more thoroughly explored. Developmental research suggests that more extensive exploration of the patterns of family interaction experienced by parents in their own childhoods could reveal significant effects of childhood experience that are independent of the contributions of depressive symptoms. Even so, the primacy of depressive symptoms as mediators of irritable parenting across socioeconomic levels is an important finding of the current survey. What remains unclear is whether a more withdrawn stance and a more irritable and punitive stance are alternative organizations of parenting displayed by different subgroups of depressed parents, or whether most depressed parents display all of these less optimal behaviors. For fathers, previous analyses had already revealed that increased negative behaviors were related to depressive symptoms while variation in positive behaviors was not. To address this question for mothers, an additional multiple regression analysis was conducted on the frequency of mother’s depressive symptoms, entering first the set of decreased positive behaviors related to depressive symptoms, which included frequency of play, frequency of reading, and consistency of routines, then entering the set of increased negative behaviors related to depressive symptoms, which included feeling aggravated, yelling, and spanking. If the same mothers were reporting both the negative and positive sets of behaviors related to depression, controlling for the association between a mother’s decreased positive

250

Lyons-Ruth, Wolfe, Lyubchik, Steingard

behaviors and her depression score should greatly attenuate the relation between the same mother’s increased negative behaviors and her depression score. As expected, the three positive behaviors negatively associated with maternal depressive symptoms – frequency of play, frequency of reading, and consistency of routines – accounted for a significant proportion of variance in maternal depressive symptoms when entered as a block, F (3,990) = 13.07 p < .0001. However, when the set of negative maternal behaviors was then added to the model, negative behaviors continued to account for unique variance in mother’s depressive symptoms, after variance associated with decreased positive behaviors was accounted for, F chg (3,987) = 16.11 p < .0001. This analysis for mothers confirms that the same depressed mothers are not reporting both decreased positive behaviors and increased negative behaviors. Instead, these two sets of behaviors relate independently to maternal depressive symptoms.

discussion and policy implications Recent epidemiologic data document a higher rate of clinical depression among women of child-bearing age than among women surveyed in the 1950s and 1970s (Murphy et al. 2000). Depressive symptoms are at their height among young adults ages 25–44 (Eaton and Kessler 1981). Further, developmental researchers have begun to document the adverse child developmental outcomes associated with parental depression and with sustained high levels of depressive symptoms. The clear implication of the findings from these two research domains is that higher rates of adverse child outcomes will result from the higher rates of depression among adults of parenting age apparent in recent surveys. The Commonwealth Survey data offer a unique and expanded window onto the predictors, current correlates, and parenting behaviors associated with increased depressive symptoms among parents with young children. Parental depressive symptoms can be thought of as functioning in relation to mental health much as fever functions in relation to physical health. Depressive symptoms are elevated in a variety of serious mental disorders, including but not limited to depressive disorders, and can also be elevated in response to a variety of stressful conditions that never result in a fullblown psychiatric diagnosis. Depressive symptoms, then, can be viewed as indicators of the degree of imbalance between a parent’s psychological coping resources and his or her sources of psychological stress. Elevated symptoms indicate that coping resources are wearing thin and may be in danger of being overwhelmed, with resultant mental or behavioral disorder. This conceptualization provides a context for understanding the multiple predictors, current correlates, and parental behaviors associated with parents’ depressive symptoms in these survey results.

Depressive Symptoms in Parents

251

The first important implication of the Commonwealth data is that while parental depression constitutes a risk factor for children, children also constitute a risk factor for adult depression. The increased risk for parental depression conferred by young children was first demonstrated in the British epidemiologic study by Brown and Harris (1975). This point is further underscored in the current survey data in the finding that having a later-born child increases a parent’s risk for depression. It is also notable that fathers who assume primary child care responsibility show depression rates of 20 percent, equal to that of mothers with primary child care duties. These results suggest that higher rates of depression among women may be partly related to society’s allocation of child care responsibilities to women. This hypothesis receives further support from the finding that male-female depression rates do not diverge until adolescence, when young people begin to assume adult sexual and social roles (Ge et al. 1994; Nolen-Hoeksema and Girgus 1994). The care of young children also fits well the “high demand-low control” work-role profile that Hlatky and colleagues have shown to be associated with high rates of depression, as well as increased rates of stress-related physical conditions such as cardiovascular disease (1995). As demonstrated in their data, hostility and depression are associated with one another, and both are associated with alterations in hypothalamic-pituitary-adrenal axis function. Persons in high demand-low control jobs also show similar alterations in psychophysiological functioning. For example, working women with young children living in the home show higher levels of cortisol excretion in a 24-hour period than working women with no children in the home (Luecken, et al. 1997; Williams et al. 1997). The relationship between depression and parenting now begs for even more sophisticated followup in surveys that ask in depth about the ages and circumstances of all children in the household, the pressures they create for their parents, and the current cultural stresses and supports that may mediate the translation of these parenting pressures into higher clinical symptoms among those caring for young children. The second major implication from the survey is that even a limited inventory of depressive symptoms yields data that replicate the results of other large and small-scale studies. Increasing the inventory of symptoms would undoubtedly increase the precision and power of the findings. However, the convergence of the data generated by the current survey with the body of well-replicated findings from other epidemiologic and developmental research allows greater confidence in generalizing those findings to the entire U.S. population of parents with young children. In addition, the data identify the population subgroups of parents at greatest risk for depressive symptoms and their associated negative outcomes. The third major implication of the Commonwealth Survey data is that parental depression is a condition with relational components that can

252

Lyons-Ruth, Wolfe, Lyubchik, Steingard

affect one’s ability to parent effectively. In the literature review at the beginning of this chapter, we considered the epidemiologic data on adult prevalence of depressive symptoms and the developmental data on child outcomes associated with parental symptoms. Omitted was a discussion of potential mediating mechanisms between the parent’s symptoms and the child’s maladaptive outcomes. The Commonwealth data converge powerfully with other accumulated developmental research data in identifying problematic parenting behaviors associated with depression that may potentially mediate the negative impact of parental depression on the child. Depressed parents in the Commonwealth Survey report more feelings of aggravation and frustration with their small child’s behavior, more yelling, more spanking, and more hitting. Developmental studies of observed parent-child interactions also document the increased irritability, negative affect, and intrusive control of the child displayed by depressed parents (Lyons-Ruth 1992; Downey and Coyne 1990; Weissman and Paykel 1974). While the popular stereotype of depression is that it leads to sadness and withdrawal, clinicians need to revise this stereotype in light of the large amount of recent data indicating that depressed parents are likely to be irritable and critical with their children and are likely to use more coercive or intrusive methods of discipline. Conversely, when irritable behaviors are displayed by parents in the clinician’s office, they should be a signal to the clinician to inquire further about the possibility of depression. Developmental studies have documented increased negative parental affect and intrusive behavior among depressed parents in interaction with their infants as early as six months of age (Cohn et al. 1986; Field 1995), with reciprocal negative affect being displayed on the part of the infant (Dawson et al. 1997; Field et al. 1995). This negative affect continues to be expressed in parental behavior through the preschool years (see, for example, De Mulder and Radke-Yarrow 1991; Downey and Coyne 1990; Lyons-Ruth et al. 1990, 1993). By age 3, child behavioral problems, including negative behavior toward parents and aggression toward peers, begin to be reported among children of depressed parents by both parents and preschool teachers (Alpern and Lyons-Ruth 1993; Richman et al. 1982). The survey also makes a unique contribution in documenting the depressed parent’s greater difficulty in setting up and following through with child care routines. This difficulty can be seen as a manifestation of the alternate, withdrawn face of depression, in which parents are not only more angry and irritable but also less able to mobilize the energy to be involved in positive ways with their children. Decreased involvement in verbal and emotional interaction with the child among depressed parents has also been documented in a number of developmental studies (see LyonsRuth 1992; Downey and Coyne 1990; Field 1995). The Commonwealth data further document the depressed parent’s decreased frequency of playing, reading, and hugging or cuddling the child. Finally, the Commonwealth

Depressive Symptoms in Parents

253

Survey data confirm widespread developmental findings that the negative correlates of parental depression are evident during the first three years of the child’s life. The policy implications of this convergence of survey data and observational research findings are pressing. Parents need more information and support in their parenting roles in the early years, particularly in light of the evidence that parenting responsibilities increase rates of depression. However, there have been very few systematic attempts to provide supports for parents of young children in the United States through the medical, social service, or mental health systems. Nurturing our youngest and most vulnerable citizens and their caregivers needs to become a more central tenet of U.S. health care policy. Obstetric and pediatric clinicians are in consistent contact with young parents and, as such, are natural points for increased delivery of services to parents or of referrals to other health services or parenting supports (see the chapter by Zuckerman and Parker, in this volume). What should those services provide? Simple screening procedures for depression among expectant parents and parents with young children constitute an easy-to-implement firstlevel intervention to identify depressed parents and provide further services. In a pilot study, a Boston-based HMO administered three depression screening questions to all adult primary care patients during 2,030 outpatient visits and identified 4.8 percent of patients with previously undetected depression. The first-line treatment for those patients was a trial of antidepressants (Sokol, Stelovich and Simeone 1997). Yet medication alone is unlikely to ameliorate the parenting difficulties associated with depression. In many developmental studies documenting adverse parenting practices associated with depression, depressed parents were already receiving medications and other standard forms of mental health care (see Weissman and Paykel 1974; Gordon et al. 1989; LyonsRuth et al. 1990; Weinberg and Tronick 1998; Teti et al. 1994). In addition, in four major longitudinal studies of the predictors of parenting behaviors, more proximal correlates of parenting, such as depressive symptoms, marital conflict, or parental self-reported personality characteristics, were less powerful predictors of parenting behaviors than parents’ descriptions of the parenting they received as children (Belsky et al. 1986; Caspi and Elder 1988; Cox et al. 1985; Lyons-Ruth 1992). In light of these data, Lyons-Ruth (1992) pointed out that a “strong theory” of depression as a primary causal agent for parenting problems does not fit the data. Instead, a “weak theory,” positing that depression is one correlate or indicator of parenting problems, provides a better fit. Adverse childhood experiences predict both parental depressive symptoms and negative parenting behaviors (Lyons-Ruth 1992). Both parenting problems and depressive symptoms are influenced by internal models of prior parent-child relationships. Even when these models are unsatisfactory to

254

Lyons-Ruth, Wolfe, Lyubchik, Steingard

parents and consciously disavowed, they are grounded in early experience and strong emotions and tend to be repeated when alternative models are not available (see, for example, van IJzendoorn et al. 1999). Organized patterns of thinking and behavior change slowly and require extended periods of exposure to new models and reworking of old ways to become stable as new parenting capabilities. While health services are well equipped to provide medication, neither pediatric nor mental health services are currently prepared to provide extended help to parents with infants and young children. However, it is increasingly clear that help with problematic parenting practices should be provided early in the relationship of parent and infant, before the child reaches school age and becomes an entrenched contributor to a negative cycle of interaction. Developing family support teams that are connected with all obstetric and pediatric services could address this critical gap in services for young families. These teams need to provide cross-disciplinary expertise in psychopharmacology, in up-to-date developmental knowledge, and in clinical skill in working with parents and infants. Both the aspects of parental history that undermine the parent’s ability to parent positively and the forms of parental behavior associated with secure and insecure attachment relationships need to be a part of the developmental expertise of professionals or paraprofessionals on the team. The availability of outreach to the home is also critical to engaging isolated and depressed parents, particularly those with more than one young child. A first wave of home-visiting studies targeting the infant’s social and emotional as well as cognitive development have demonstrated some success in affecting parent-child interaction and attachment (Beckwith 1988; Lyons-Ruth et al. 1990; Lieberman et al. 1991) and child abuse (Olds and Kitzman 1993). In the one intervention study that examined parental depressive symptoms, however, Lyons-Ruth et al. (1990) found that 9–18 months of weekly home visiting services produced significant changes in parent-infant interaction and attachment but not in maternal depressive symptoms, which remained at high levels. Despite the continuation of maternal depressive symptoms, however, three and a half years later children who had received home-visiting as infants were rated by teachers as less aggressive toward peers in the classroom (Lyons-Ruth, Lyubchik & DiLallo 2000). A second wave of longer-term intervention studies, some specifically targeted at depressed mothers, is now in progress, influenced by the early findings of several researchers regarding the centrality of depressive symptoms as a marker for impaired parent-infant interaction (Lyons-Ruth et al. 1986; Field 1984; Radke-Yarrow et al. 1985). These newer studies are only beginning to publish results, but preliminary reports indicate that positive effects on parent-infant interaction and attachment occur by 12–18 months, whether or not changes occur in parental depressive symptoms

Depressive Symptoms in Parents

255

(see Heinicke, 1999; Cicchetti et al. 2000; Murray et al. 1996; also see Olds and Kitzman 1993 for a review of intervention studies not focused on depression). Stable, long-term change in depressive symptoms among parents with young children may require, among other things, working out new and more rewarding ways of interacting as a family. These more rewarding models need to be made available to young families through sustained helping relationships that allow time for this reworking process to occur. With emerging research results supporting the efficacy of such services, a more focused national commitment to the development of comprehensive support systems for young families needs to follow.

references American Psychiatric Association. (1994). Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition. Washington, D.C.: American Psychiatric Association. Abrams, S.M., Field, T., Scafidi, F., & Prodromidis, M. (1995). Newborns of depressed mothers. Infant Mental Health Journal, 16(3), 233–39. Alpern, L., & Lyons-Ruth, K. (1993). Preschool children at social risk: Chronicity and timing of maternal depressive symptoms and child behavior problems at school and at home. Development and Psychopathology, 5, 371–87. Beardslee, W.R., Bemporad, J., Keller, M., & Klerman, G.L. (1983). Children of parents with major affective disorder: A review. American Journal of Psychiatry, 140(7), 825–32. Beckwith, L. (1988). Early intervention with parents of preterm infants. Psychiatry, 51, 242–47. Belsky, J., Hertzog, C., & Rovine, M. (1986). Causal analyses of multiple determinants of parenting: Empirical and methodological advances. In Lamb, M., Brown, A., & Rogoff, B. (eds.), Advances in developmental psychology, vol. 4. Hillsdale, NJ: Erlbaum. Brown, G., & Harris, T. (1975). Social Origins of Depression: A Study of Psychiatric Disorders in Women. New York, NY: Free Press. Campbell, S.B., Cohn, J.F., & Meyers, T. (1995) Depression in first-time mothers: Mother-infant interaction and depression chronicity. Developmental Psychology, 31, 349–357. Caspi, A., & Elder, G. (1988). Childhood precursors of the life course: Early personality and life disorganization. In Hetherington, E.M., Lerner, R., & Perlmutter, M. (eds.), Child development in life-span perspective (pp. 115–142). Hillsdale, NJ: Erlbaum. Cicchetti, D., Rogosch, F., & Toth, S.L. (2000). The efficacy of toddler-parent psychotherapy for fostering cognitive development in offspring of depressed mothers. Journal of Abnormal Child Psychology, 28, 135–148. Clarke-Stewart, A. (1978). And daddy makes three: The father’s impact on mother and young child. Child Development, 49, 466–478. Cohn, F., Matias, R., Tronick, E.Z., Connell, D., & Lyons-Ruth, K. (1986). Face-to-face interactions of depressed mothers and their infants. In Tronick, E.Z., & Field, T.

256

Lyons-Ruth, Wolfe, Lyubchik, Steingard

(eds.), Maternal depression and infant disturbance (pp. 31–45). San Francisco, CA: Jossey-Bass. Comstock, G.W., & Helsing, K.J. (1976). Symptoms of depression in two communities. Psychological Medicine, 6, 551–563. Cox, M.J., Owen, M.T., Lewis, J.M., Riedel, C., Scalf-McIver, L., & Suster, A. (1985). Intergenerational influences on the parent-infant relationship in the transition to parenthood. Journal of Family Issues, 6, 543–564. Cross-National Collaborative Group (1992). The changing rate of major depression: cross-national comparisons. JAMA, 268, 3098 – 3105. Cummings, E.M., & Davies, P.T. (1994). Maternal depression and child development. Journal of Child Psychology and Psychiatry, 35(1), 73–112. Dawson, G., Frey, K., Panagiotides, H., Osterling, H., & Hessl, D. (1997). Infants of depressed mothers exhibit atypical frontal brain activity: A replication and extension of previous findings. Journal of Child Psychology and Psychiatry, 38(2), 179–186. DeMulder, E.K. & Radke-Yarrow, M. (1991). Attachment with affectively ill and well mothers: Concurrent behavioral correlates. Development and Psychopathology, 3, 227–242. Downey, G., & Coyne., J.C. (1990). Children of depressed parents: An integrative review. Psychological Bulletin, 108(1), 50 – 76. Eaton, W.W., & Kessler, L.G. (1981). Rates of symptoms of depression in a national sample. American Journal of Epidemiology, 114, 528–538. Field, T. (1995). Infants of depressed mothers. Infant Behavior and Development, 18, 1–13. Field, T., Fox, N.A., Pickens, J., & Nawrocki, T. (1995). Relative right frontal EEG activation in 3- to 6-month-old infants of “depressed” mothers. Developmental Psychology, 31(3), 358–363. Field, T., Estroff, D.B., Yando, R., del Valle, C., Malphurs, J., & Hart, S. (1996). “Depressed” mothers’ perceptions of infant vulnerability are related to later development. Child Psychiatry and Human Development, 27(1), 43–53. Field, T.M. (1984). Early interactions between infants and their postpartum depressed mothers. Infant Behavior and Development, 7, 517–22. Ge, X., Lorenz, F.O., Conger, R.D., Elder, G.H., & Simons, R.L. (1994). Trajectories of stressful life events and depressive symptoms during adolescence. Developmental Psychology, 30, 467–483. Goodman, S.H., Adamson, L.B., Riniti, J., & Cole, S. (1994). Mothers’ expressed attitudes: Associations with maternal depression and children’s self-esteem and psychopathology. Journal of the Academy of Child and Adolescent Psychiatry, 33(9), 1265–1274. Gordon, D., Burge, D., Hammen, C., Adrian, C., Jaenicke, C., & Hiroto, D. (1989). Observations of interactions of depressed women with their children. American Journal of Psychiatry 146, 50–55. Grof, P. (1997). The age of misery: Affective disorders on the rise. Canadian Journal of Psychiatry, 42, 361–362. Gross, D., Conrad, B., Fogg, L., Willis, L., & Garvey, C. (1995). A longitudinal study of maternal depression and preschool children’s mental health. Nursing Research, 44(2), 96–101.

Depressive Symptoms in Parents

257

¨ Hagnell, O., Lanke, J., Rorsman, B., & Ojesj o, ¨ L. (1982). Are we entering an age of melancholy? Depressive illnesses in a prospective epidemiological study over 25 years: The Lundby Study, Sweden. Psychological Medicine, 12, 279–289. Hammen, C., Gordon, G., Burge, D., Jaenicke, C., & Hiroto, D. (1987). Maternal affective disorders, illness, and stress: Risk for children’s psychopathology. American Journal of Psychiatry, 144, 736 –741. Heinicke, C., Fineman, N., Ruth, G., Ricchia, S., & Guthrie, D. (1999). Relationshipbased intervention with at-risk mothers: Outcome in the first year of life. Infant Mental Health Journal. Hlatky, M., Lam, L., Lee, K., Clapp-Channing, N.E., Williams, R.B., Pryor, D.B., Califf, R.M., & Mark, D.B. (1995). Job strain and the prevalence and outcome of coronary heart disease. Circulation, 92, 327–333. Kendall, R., Rennie, D., Clarke, J., & Dean, C. (1981). The social and obstetric correlates of psychiatric admission in the puerperium. Psychological Medicine, 11, 341–350. Klerman, G.L., & Weissman, M.M. (1989). Increasing rates of depression. Journal of the American Medical Association 261, 2229–35. Lamb, M.E. (1982). The father-child relationship: A synthesis of biological, evolutionary, and social perspectives. In Hoffman, L.W., R. Gandelman, & H.R. Schoffman (eds.), Parenting: Its causes and consequences (pp. 55–73). Hillsdale, NJ: Erlbaum. Lieberman, A.F., Weston, D.R., & Pawl, J.H. (1991). Preventive intervention and outcome with anxiously attached dyads. Child Development 62, 199–209, 1991. Luecken, I., Suarez, E.C., Kuhn, C., Barefoot, J.C., Blumenthal, J.A., Siegler, I.C., & Williams, R.B. (1997). Stress in employed women: Impact of marital status and children at home on neurohormone output and home strain. Psychosomatic Medicine, 59, 352–359. Lyons-Ruth, K. (1992). Maternal depressive symptoms, disorganized infant-mother attachment relationships and hostile-aggressive behavior in the preschool classroom: A prospective longitudinal view from infancy to age five. In Cicchetti, D., & Toth, S. (eds.), Rochester symposium on developmental psychopathology: Vol 4, A developmental approach to affective disorders (pp. 131–171). Rochester, NY: University of Rochester Press. Lyons-Ruth, K., Alpern, L., & Repacholi, B. (1993). Disorganized infant attachment classification and maternal psychosocial problems as predictors of hostileaggressive behavior in the preschool classroom. Child Development, 64, 572–585. Lyons-Ruth, K., Connell, D., Grunebaum, H., & Botein, S. (1990). Infants at social risk: Maternal depression and family support services as mediators of infant development and security of attachment. Child Development, 61, 85– 98. Lyons-Ruth, K., Easterbrooks, M.A., & Cibelli, C.D. (1997). Infant attachment strategies, infant mental lag, and maternal depressive symptoms: Predictors of internalizing and externalizing problems at age 7. Developmental Psychology, 33(4), 681– 692. Lyons-Ruth, K., Lyubchik, A., & DiLallo, J (2000). The impact of early home-visiting services on behavior problems and social competence in Kindergarten: A three and a half year followup. Congress Abstracts, 7th World Conference of the World Association for Infant Mental Health, Infant Mental Health Journal, 21, (4–5), 292.

258

Lyons-Ruth, Wolfe, Lyubchik, Steingard

Lyons-Ruth, K., Zoll, D., Connell, D., & Grunebaum, H. (1986). The depressed mother an her one-year-old infant: Environmental context, mother-infant interaction and attachment, and infant development. In Tronick, E., & Field, T. (eds.), Maternal depression and infant disturbance (pp. 61–82). San Francisco, CA: Jossey Bass. MacDonald, K., & Parke, R.D. (1986). Parent-child physical play: The effects of sex and age of children and parents. Sex Roles, 15, 367–378. Martinez, A., Malphurs, J., Field, T., Pickens, J., Yando, R., Bendell, D., Valle, C., & Messinger, D. (1996). Depressed mothers’ and their infants’ interactions with nondepressed parents. Infant Mental Health Journal, 17(1), 74–80. McLoyd, V.C. (1990). The impact of economic hardship on black families and children: Psychological distress, parenting, and socioemotional development. Child Development, 61, 311–346. Murphy, J., Laird, N., Monson, R., Sobol, A., & Leighton, A. (2000). A forty-year perspective on the prevalence of depression from the Stirling County Study. Archives of General Psychiatry, 57(3), 209–215. Murray, L., Hipwell, A., Hooper, R., Stein, A., & Cooper, P. (1996). The cognitive development of 5-year-old children of postnatally depressed mothers. Journal of Child Psychology and Psychiatry, 37(8), 927– 935. Nolen-Hoeksema, S., & Girgus, J.S. (1994). The emergence of gender differences in depression during adolescence. Psychological Bulletin, 115, 424–443. Nolen-Hoeksema, S., Wolfson, A., Mumme, D., & Guskin, K. (1994). Helplessness in children of depressed and nondepressed mothers. Special Section: Parental depression and distress: Implications for development in infancy, childhood and adolescence. Developmental Psychology, 31(3), 377–387. Olds, D.L., & Kitzman, H. (1993). Review of research on home visiting for pregnant women and parents of young children. The Future of Children, 3, 53–92. Orr, S., & James, S. (1984). Maternal depression in an urban pediatric practice: Implications for health care delivery. American Journal of Public Health, 74(4), 363–365. Parker, G. (1987). Are the lifetime prevalence estimates in the ECA study accurate? Psychological Medicine, 17, 275–282. Phares, V., & Compas, B.E., (1992). The roles of fathers in child and adolescent psychopathology: Make room for daddy. Psychological Bulletin, 111, 387–412 Radke-Yarrow, M., Cummings, E.M., Kuczynski, L., & Chapman, M. (1985). Patterns of attachment in two- and three- year olds in normal families and families with parental depression. Child Development, 36, 884–893. Regier, D.A., Boyd, J.H., Burke, J.D., Rae, D.S., Myers, J.K., Kramer, M., Robins, L.N., George L.K., Karno, M., & Locke, B.Z. (1988). One month prevalence of mental disorders in the United States; Based on five epidemiological catchment area sites. Archives of General Psychiatry, 45, 977–986. Richman, N., Stevenson, J., & Graham, P. (1982). Preschool to school: A behavioral study. London: Academic Press. Robins L.N., & Regier, D.A. (eds.), (1991). Psychiatric Disorders in America: The Epidemiologic Catchment Area Study. New York. The Free Press. Sharp, D., Hay, D.F., Pawlby, S., Schmucker, G, Allen, H., & Kumar, R. (1995). The impact of postnatal depression on boys’ intellectual development. Journal of Child Psychology and Pscyhiatry, 36(8), 1315–1336.

Depressive Symptoms in Parents

259

Sokol, H.N., Stelovich, S., & Simeone, P. (1997). The Concord Hillside Medical Associates “Early Detection of Depression” Pilot Study. Harvard Community Health Plan Practice Forum, 2, 6–7. Teti, D., Gefland, D.M., Messinger, D. S., & Isabella, R. (1994). Maternal depression and the quality of early attachment: An examination of infants, preschoolers, and their mothers. Special Section: Parental depression and distress: Implications of development in infancy, childhood, and adolescence. Developmental Psychology, 31(3), 364–376. van IJzendoorn, M.H., Schuengel, C., & Bakermans-Kranenburg, M.K. (1999). Disorganized attachment in early childhood: Meta-analysis of precursors, concomitants and sequelae. Development and Psychopathology, 11, 225–249. Wagner, B., & Phillips, D. (1992). Beyond beliefs: Parent and child behaviors and children’s perceived academic competence. Child Development, 63, 1380–1391. Weinberg, M.K., & Tronick, E. (1998). The impact of maternal psychiatric illness on infant development. Journal of Clinical Psychiatry, 59, 53–61 Weissman, M., & Paykel, E. (1974). The depressed woman: A study of social relationships. Chicago: University of Chicago Press. Weissman, M.M., Prusoff, B.A., Gammon, D., Merikangas, K.R., Leckman, J.F., & Kidd, K.K. (1984). Psychopathology in the children (ages 6–18) of depressed and normal parents. Journal of the American Academy of Child Psychiatry, 23(1), 78–84. Weissman, M., Gammon, D., John, K., Merikangas, K.R., Warner, V., Prusoff, B.A., & Sholomskas, D. (1987). Children of depressed parents: Increased psychopathology and early onset of major depression. Archives of General Psychiatry, 44, 847–853. Williams, R.B., Barefoot, J.C., Blumenthal, J., Helms, M.J., Luecken, L., Pieper, C.F., Siegler, I.C., & Suarez, E.C. (1997). Psychosocial correlates of job strain in a sample of working women. Archives of General Psychiatry, 54, 543–548. Zuckerman, B.S., Amaro, H., & Beardslee, W. (1987). Mental health of adolescent mothers: The implications of depression and drug use. Developmental and Behavioral Pediatrics, 8(2), 11–116. Zuckerman, B., Bauchner, H., Parker, S., & Cabral, H. (1990). Maternal depressive symptoms during pregnancy, and newborn irritability. Developmental and Behavioral Pediatrics, 11(4), 190–194.

part

III

DELIVERY OF HEALTH SERVICES TO MOTHERS AND CHILDREN

9 Prenatal Care, Delivery, and Birth Outcomes Paul H. Wise

Childbirth, throughout history, has been shaped by the dual influences of technical expertise and the societal commitment to the promise of life. Current patterns of childbirth in America have been deeply affected by unprecedented innovations in clinical capability to prevent and treat many of the most important traditional threats to the health of women and newborns during pregnancy, labor and delivery, and early infancy. Improvements in preconceptual screening, preventive nutritional supplementation, prenatal diagnosis and intervention, management of high-risk labor and delivery, treatment of critically ill newborns, and integration of preventive educational and social services into primary family health services have all contributed to major changes in both the nature and the risks of childbirth in the United States. Yet the evolution of childbirth in America is not only a product of technical progress; it is also a reflection of changes in American society and family life. This chapter outlines some of the more important emerging patterns of service utilization and health outcomes related to childbirth in the United States, with a focus on how technical capabilities interact with social forces and how services and outcomes are actually distributed in an increasingly diverse and socially stratified population. Findings from the Commonwealth Fund Survey of Parents with Young Children are then discussed as they reflect on several important areas: prenatal care, attendance at childbirth classes, type of delivery, prematurity and low birthweight, and length of postpartum hospital stay. The chapter ends with a discussion of policy implications and other conclusions.

trends in prenatal, childbirth, and postnatal services The Significance of High-Quality Prenatal Services Although prenatal care refers generally to services provided to women during pregnancy, there remains considerable variation in how the term 263

264

Wise

is used in both the scientific literature and everyday speech. Recent efforts to identify important elements of prenatal care have been helpful in guiding the delivery of high-quality services to pregnant women (American College of Obstetricians and Gynecologists 1993). The precise content of high-quality prenatal services is likely to defy strict standardization, however, and should be welcomed as both varied and dynamic: variable in the sense that some populations of women may have different needs than others, and dynamic because risks in pregnancy and our capability to modify those risks are evolving continuously. Nevertheless, for most purposes, including those of this chapter, prenatal care can be broadly viewed as a rich blend of medical and social services designed to enhance maternal health and infant outcomes (Brown 1993). The variability and dynamic quality of prenatal care has in turn made it difficult to gauge the ultimate effectiveness of that care. Although few commentators question the importance of prenatal care as a part of a comprehensive system of health care, less consensus exists as to its utility in preventing specific types of adverse outcomes, particularly prematurity. It has been noted that, despite improvements in the early initiation of prenatal care in the United States, prematurity rates have not fallen (Kogan et al. 1998a). In addition, evaluations of enhanced prenatal care programs have revealed decidedly inconclusive results. While programs designed to reduce financial barriers and expand outreach efforts have generally been associated with improved utilization of prenatal services and reduced rates of low birthweight (Buescher et al. 1991; Ray et al. 1997; Baldwin et al. 1998), their impact on extreme prematurity and very low birthweight has not been impressive (Piper et al. 1994; Haas et al. 1993; Mercer et al. 1996; Dyson et al. 1998; Ray et al. 1997, 1998). New clinical insights have underscored the complex etiology of prematurity and have raised questions about the ability of standard prenatal care to prevent premature births (Goldenberg and Andrews 1996; Goldenberg and Rouse 1998; Iams 1998). Not surprisingly, these reports have been met by some controversy in the policy world since they tend to contradict the view advanced by many advocates that expanded prenatal care programs would prevent prematurity, low birthweight, and infant death (National Commission to Prevent Infant Mortality 1988). This dynamic has only intensified longstanding tensions between the clinical and public health approaches to improving birth outcomes (Meckel 1990; Wise 1993, 1999). Nevertheless, these recent evaluations have helped reframe the capabilities of prenatal care as decidedly mixed, with greater efficacy in addressing certain health problems than others. In turn, this enhanced caution has generated much-needed interest in more refined means of measuring both the use and the content of prenatal care, as well as the quality of delivered services (Kogan et al. 1994; Misra and Guyer 1998). The evidence questioning the efficacy of prenatal care has therefore played a constructive role, both by highlighting the need

Prenatal Care, Delivery, and Birth Outcomes

265

for new clinical interventions and by laying a foundation for the provision of a more comprehensive range of services to populations at risk (Hauth et al. 1995; Alexander and Howell 1997). Prenatal Care and Women’s Health Although ensuring healthy childbirth is a central concern of prenatal care, the purposes of prenatal care extend beyond the health of pregnancy. Most fundamentally, the general health needs of young women in the United States remain poorly addressed by current programs and policies, and prenatal care can play an important role in identifying needs and linking women to appropriate services (Expert Panel on the Content of Prenatal Care 1989). These functions could prove quite important, as many of the most important threats to the health of young women in the United States are largely preventable. Injuries (particularly from motor vehicle collisions), HIV, homicide, and suicide are the leading causes of death for women ages 15–45 (Anderson et al. 1998). Sexually transmitted diseases, ectopic pregnancies, depression, and anxiety disorders can all be addressed in the prenatal environment with important long-term ramifications (MMWR 1996, 1995; Walters 1993). In addition, chronic conditions such as asthma and obesity, as well as adverse health behaviors, including smoking and illicit drug use, affect large portions of women of reproductive age. Eligibility for public programs may also be first recognized through prenatal care. In a sense, therefore, prenatal care is best defined not as an isolated set of specialized services but as an integral component of comprehensive women’s health services over a lifetime (Chavkin 1995). A number of women’s health conditions and behaviors can also have important effects on child health and functioning. Chronic illness in women can affect family functioning and the quality of maternal-child interactions (McGauhey and Starfield 1993). Depression is a common mental health problem with well-documented adverse effects on children’s health (Beardslee et al. 1983; Zuckerman and Beardslee 1987; Beardslee 1989). Parental depression can also affect the risk that children will experience similar symptoms, as well as other behavioral and educational problems (Downey and Coyne 1990). Maternal smoking has been linked to sudden infant death syndrome (Blair et al. 1996), hospitalization for lower respiratory tract infections and asthma (Chilmonczyk et al. 1993), and otitis media (Ey et al. 1995). Children of alcoholics are at increased risk for injuries, depression, and low self-esteem, and may themselves drink more heavily than their peers in adolescence (Margolis et al. 1986; Epstein et al. 1995), and parental behaviors can influence their children’s attitudes to diet, exercise, tobacco, and alcohol use (Sallis et al. 1992). In addition, there is some evidence that a mother’s care-seeking behavior during the prenatal

266

Wise

period may characterize pediatric care-seeking patterns for her children in the first few years of life (Freed et al. 1999; Kogan et al. 1998b) Measuring Prenatal Care In general, the measurement of prenatal care utilization remains relatively crude, relying most heavily on simple indicators of visits to a physician, midwife, or nurse. This approach was developed to utilize birth certificate information, permitting standard assessments of large populations. Because such data sources do not provide detailed information on the actual services delivered as part of prenatal care, the principal measures have been quantitative in nature, focusing on two items: when in pregnancy the first prenatal care visit occurred, and how many total visits occurred prior to delivery. By assessing prenatal care initiation and the number of prenatal care visits, a general description of prenatal care utilization in a population can be developed. Several methods exist to link initiation and visits as part of an index of the adequacy of prenatal care visitation. One widely used index uses a nomogram that categorizes each pregnancy into adequate, intermediate, or inadequate, based on the trimester of initiation and the number of visits (Kessner 1972). A newer index developed by Kotelchuck assesses the adequacy of care based on the recommendations of the American College of Obstetricians and Gynecologists and generally provides a more accurate determination of adequacy (Kotelchuck 1992). Both methods build in corrections for infants born prematurely (short pregnancy duration), as they will likely have fewer prenatal care visits than a full-term birth even if prenatal care visitation has been appropriate. Both indices, however, tend to reflect when prenatal care first began. Other indices have been introduced and may provide more refined insights into patterns of prenatal care use in some analytic settings (Alexander and Kotelchuck 1996; Kogan et al. 1998a). Due to the overwhelming impact of extreme prematurity, however, initiation of care remains particularly important in determining the opportunity to prevent the most serious birth outcomes in the United States. Approximately two-thirds of all neonatal deaths in the United States are associated with prematurity (National Center for Health Statistics 1995). Severe congenital anomalies, many of which can now be identified early in pregnancy, also contribute substantially. Recent Trends in Prenatal Care Use In 1996, 81.8 percent of all women delivering liveborn infants in the United States received prenatal care in the first three months of pregnancy. Although this represents a modest increase over the 75.8 percent of women to receive such care in 1990, it reflects far slower progress than had occurred during the 1970s. In 1969, first trimester prenatal care initiation was 68.5 percent; by 1980, it had reached 76.3 percent. During the 1980s, the

Prenatal Care, Delivery, and Birth Outcomes

267

figure actually fell slightly, reaching 75.8 percent by 1990. Current rates therefore represent a welcome, though small, improvement in early prenatal care rates (National Center for Health Statistics 1995). Although there is evidence that these improvements have occurred in all social groups, they have been less profound in certain high-risk groups, including African Americans and women of low education. This finding implies that even as national initiation rates improve, special efforts are likely to be necessary to enhance utilization in certain groups. Although national initiation trends are important, it is also useful to examine trends in prenatal visitation once prenatal care has begun. Recently developed indices directed at measuring the intensity of prenatal care offer some evidence that intensity has increased faster than early initiation. In addition, it appears that increases in intensive prenatal care use have been less profound for African Americans and women of low education (Kogan et al. 1998a). The trend in the portion of all births associated with third trimester or no prenatal care represents the mirror opposite of the pattern observed for early initiation. During the 1960s and 1970s, this rate fell significantly; during the 1980s, the rate of late or no prenatal care actually rose from 5.1 percent in 1980 to 6.4 percent in 1989. By 1995, the rate had fallen to 4.4 percent of all births. Among the more troubling aspects of prenatal care use in the United States is the persistence of major social differences in early initiation. In 1996, 83.9 percent of all white births were associated with first trimester initiation; the corresponding figures for black births and Hispanic births were 71.3 percent and 71.9 percent, respectively. An analysis of 1993 data suggests that 90 percent of mothers 20 years of age and over with more than 12 years of education received early prenatal care, while only 63 percent of mothers with fewer than 12 years of education received such care. Even for women of similar education levels, black mothers received significantly less early prenatal care than their white counterparts. Addressing Barriers to Prenatal Care Use Barriers to the use of prenatal care can be grouped into four general categories: financial barriers, including inadequate insurance coverage; inadequate capacity of local health care systems; ineffective organization or delivery of high-quality services; and cultural and personal attributes that influence care-seeking behaviors. Financial barriers occur where the provision of services is highly dependent on the financial resources of a patient. Women with private health insurance have traditionally had the highest rates of early prenatal care initiation. Medicaid, the largest public program of health coverage for the poor in the United States, has had a major impact on prenatal care initiation and is largely credited with documented increases in prenatal care use by

268

Wise

poor women during the 1960s and 1970s (Expert Panel on the Content of Prenatal Care 1989; Brown 1993). Over the past decade, a number of states have attempted to reduce financial barriers to prenatal care use, either by extending Medicaid coverage to a wider range of women or by creating new coverage mechanisms for women with incomes slightly above the official poverty line or otherwise ineligible for Medicaid (illegal immigrants, for example, in some states). Despite the utility of these insurance-focused initiatives, a large literature has documented serious inadequacies in the capacity of local health care systems to provide timely care (Brown 1993; Chavkin 1995). Long waiting times are the rule in many communities, and represent a problem of special importance for the early initiation of prenatal care. In some areas, the lack of adequate system capacity is due to the choice by certain institutions and obstetrical providers not to accept patients covered by Medicaid. In some communities, poor organization and coordination of services can undermine access to prenatal care (Brown 1993; Andersen et al. 1983). Women’s health services are among the most fragmented in the American health care system. Obstetrical care may be completely unconnected to contraception or primary care services, while abortion services have little contact with other elements of women’s health care in many communities. Beyond issues of coordination, provider knowledge and attitudes toward patients may create functional barriers to appropriate services (Aday and Andersen 1974). These factors may become particularly important in determining the nature and quality of services once prenatal care has begun. The use of prenatal care can also be influenced by beliefs regarding pregnancy, the efficacy of health services, and awareness of programs designed to enhance access to these services (Weissman and Epstein 1996). Great caution should be exercised, however, whenever personal factors are used to explain documented differences in prenatal care utilization, particularly when those factors are associated with specific social groups. The danger is that an illusion of personal decision making can obscure deeper structural determinants of access, including financial burdens, logistical problems such as transportation or traveling in an unsafe neighborhood, or language barriers. Past experiences with the health care system can also shape subsequent care-seeking behaviors. From this perspective, access must be defined not only in terms of capacity or availability but also in terms of dignity (Daniels 1985). Here, access should not be considered adequate if the emotional toll or practical hardship inherent in seeking care is so great that subsequent utilization is affected. Attendance at Childbirth Classes Although not usually recognized as a component of prenatal care, childbirth classes should be viewed as a useful element of comprehensive

Prenatal Care, Delivery, and Birth Outcomes

269

preparation for labor, delivery, and infant care. Classes generally involve several sessions during middle to late pregnancy and usually address the mechanisms of labor and delivery, methods of breath control and relaxation, and health, safety, and general care issues related to the arrival of a newborn. The classes convey important information, and parents’ attendance may also serve as a reflection of their broader receipt of information about pregnancy, delivery, and young child care. Social differences in childbirth class attendance, in themselves an important point of concern, may indicate more far-reaching disparities in the provision of information related to optimal maternal and infant outcomes. Type of Delivery Vaginal delivery is generally considered the normal delivery method and is often termed “simple” or “uncomplicated,” although women who have experienced vaginal delivery might well refute those characterizations. The surgical removal of the fetus from the uterus may be required under a variety of circumstances and is generally termed a cesarean delivery, so named because of the traditional belief that Julius Caesar (or his eponymous ancestor) was born in this manner. Although the vast majority of all births are delivered vaginally, the rate of cesarean deliveries has been quite dynamic over the past 30 years. Rates of cesarean deliveries in the United States grew significantly between the early 1960s and 1986, with a relative plateauing since then (Taffel et al. 1990). Cesarean section is among the most common surgical procedures in the United States, occurring in 20.6 percent of all births in 1996 (National Center for Health Statistics 1995). The concern associated with the dramatic increase in cesarean deliveries is related to the relative increase in morbidity and cost associated with unnecessary cesarean procedures. Underscoring the potential that some portion of current cesarean deliveries are not necessary is the wide variation in cesarean rates in different geographical areas. Although it is difficult to ascertain whether these differences reflect real differences in the clinical need for cesarean deliveries, the general view is that the variation in cesarean section rates reflects, at least in some measure, the influence of nonclinical factors on patterns of delivery methods (Haas et al. 1993). Such nonclinical factors could include patient preferences, obstetrical practice patterns, financial considerations, and discriminatory decision making, among many others (Goyert et al. 1989). A special arena of concern is the impact of nonclinical factors on repeat cesarean deliveries, or cesarean deliveries in second or later pregnancies done because of a prior cesarean delivery. For many years, the common wisdom in obstetrical practice was that once a cesarean section had been performed, subsequent deliveries would have to be delivered by cesarean

270

Wise

section for fear that the weakened uterine wall would preclude active labor. This approach has been modified over recent years to allow a trial of active labor in most cases, even if a prior birth was delivered by cesarean section. Despite the demonstrated safety of vaginal birth after cesarean (VBAC) in most cases, considerable variation in its use persists. Nonclinical factors appear to have a considerable impact on VBAC rates, as they do for all cesarean deliveries (Stafford 1991). Prematurity and Low Birthweight Premature and low birthweight newborns currently account for approximately two-thirds of all neonatal deaths in the United States (Overpeck et al. 1992), owing to the greatly enhanced risk of a variety of serious medical problems among infants born too early or too small. Increased rates of premature and low birthweight infants largely account for the relatively poor ranking of the United States in measures of infant mortality when compared with other industrialized countries. In addition, elevated rates of premature and low birthweight infants are the primary reason for the tragically elevated mortality rate of black infants in the United States (Wise et al. 1995). Even among surviving newborns, prematurity and low birthweight can be associated with an increased risk for chronic illness and developmental disorders. Social disparities in prematurity and low birthweight have long been documented. In addition to race, low income and maternal education, very young and older maternal age, late or no prenatal care, and a variety of maternal behaviors such as smoking, illicit drug use, and heavy alcohol use have all been associated with prematurity and low birthweight. However, the clinical determinants of prematurity and low birthweight remain poorly defined. Recent work has suggested a heterogeneous etiology including infectious, vascular, and other obstetrical causes, many of which may have their roots in maternal conditions that exist prior to pregnancy (Kempe et al. 1992; Hillier et al. 1995). The relationship between social factors and clinical pathways remains largely unexplored. However, these social factors are likely to find clinical expression by elevating the medical risk of a pregnancy or by reducing access to effective preventive and therapeutic interventions (Wise 1993). Postpartum Length of Hospital Stay Prior to the twentieth century, childbirth almost always occurred in or close to the home. However, with growing recognition of infection control, the emergence of modern obstetrics, and the growing power of the medical profession, childbirth moved into the domain of the hospital. In 1995, almost 96 percent of all births in the United States were hospital-based.

Prenatal Care, Delivery, and Birth Outcomes

271

During the 1960s and early 1970s, hospitalization for uncomplicated vaginal childbirth generally entailed a three-day stay. Beginning in the 1970s, however, the length of stay associated with normal childbirth started to decline. Although the reasons for this decline were complex, concern for the cost of hospitalization seems to have been an important determinant. Pressures to reduce costs have accelerated over the past few years and have provided growing financial incentives for reduced hospital stays after childbirth. Several studies have suggested that these pressures may be strongest in managed care or other prospective payment systems. This push to reduce hospital stays after childbirth, however, has been met by important concerns regarding the potential for serious medical consequences, as well as a considerable public outcry over the prospect of forcing women and their newborns out of the hospital too quickly. In response, public officials and politicians have turned a critical eye toward what have been labeled “drive-through” deliveries. More than half the states have passed legislation designed to protect reimbursement for hospital stays, generally for approximately 48 hours after delivery, for families covered by insurance plans that fall under state jurisdiction (Seaman 1997). However, about half of all families in the United States are covered by employerbased systems that fall under the Employee Retirement Income Security Act (ERISA), which is regulated under federal law. Accordingly, the federal Newborns’ and Mothers’ Health Protection Act was passed by Congress; as of the beginning of 1998, the act restricts the power of insurance plans that fall under ERISA to mandate hospital stays of below 48 hours. Although such legislative initiatives have been greeted with broad appreciation, there remains considerable controversy as to whether there exists sufficient evidence that short hospital stays after delivery are actually detrimental to the health of women or infants. The available studies tend to suffer from inconsistencies in methodology or inadequate sample sizes to determine meaningful results with confidence. In order to assess the impact of shortened obstetrical stays and the potential impact of these legislative responses, it is useful to take a step back and examine the central purposes of hospitalization after childbirth. In general, the purposes fall into three categories: identification and treatment of perinatal conditions, education and assistance with the requirements of caring for a young infant, and mobilization of ongoing medical and social services for families in need. The identification and treatment of pathological conditions speaks directly to the enormous physiological changes that women and infants undergo soon after delivery. The transformation from intrauterine life to the postnatal environment carries with it major new challenges to the anatomy and metabolism of the newborn. Congenital anomalies, particularly of the heart, may be identified only after a period of extrauterine

272

Wise

life. Jaundice may reach its greatest threat to neonatal health only after the first two days of life. The effects of unsuccessful breastfeeding, particularly dehydration, are most likely to develop after several days of life. Important screening techniques for such serious metabolic disorders as phenylketonuria may not be particularly accurate until regular feedings have been established, usually after the first day of life. In addition to the issues that surround the health of the newborn, maternal conditions such as postpartum infection and complications from procedures such as episiotomy may not be exhibited until several days after delivery. Maternal mental health needs, particularly those associated with depression, may also not be readily observable shortly after birth. The hospital may also serve as an important setting for parental education on such topics as newborn care, breastfeeding, and home safety. For parents, providing practical care for the newborn under the eye of hospital staff may yield invaluable information and inspire confidence prior to discharge. Supportive supervision of early breastfeeding may prove particularly important in assuring that breastfeeding will be successful and continue through early infancy. The importance and use of infant car seats can also be effectively conveyed through hospital-based efforts. The obstetrical stay also represents a crucial opportunity to link women and their children with a variety of medical and social services. To fulfill this role, the hospital must have the interest and capacity to identify the health and social needs of patients, along with mechanisms for connecting families with ongoing health and social services. Short hospital stays may make identification and linkages more difficult. Given that the three central purposes of postpartum hospitalization are unquestionably important elements of high-quality postpartum care, the prospect of shortened hospital stays raises two primary questions: Can these purposes still be met by hospital-based care, even when the length of stay is relatively brief? And if these purposes cannot be met adequately by short hospital stays, what mechanisms can be instituted to replace previously available hospital-based services? Despite the importance of these questions, the scientific evidence necessary to address them remains largely inadequate. In a comprehensive review, Braveman and colleagues pointed out that measuring the impact of shortened stays can be technically difficult (1995). There is recent evidence, however, that relatively short stays for low-risk births may not be associated with serious adverse outcomes, particularly if accompanied by appropriate outpatient services (Mandl et al. 1998). Even so, these and other studies have documented that short obstetric stays may not be confined only to women with low risk characteristics (Margolis et al. 1997). Several studies suggest that the integration of hospital and supplemental services, such as enhanced prenatal education and postpartum home visits

Prenatal Care, Delivery, and Birth Outcomes

273

by trained nursing personnel, could be of some benefit (Braveman et al. 1996; Mandl et al. 1998). Hospital and home-based services could be coupled to facilitate early identification of pathological conditions, parental education and breastfeeding support, and linkages to needed services. Particularly when implemented for families selected because of their low risk status and interest in early discharge, such integrated approaches may prove to be safe improvements over traditional hospital-based care (Carty and Bradley 1990; Seidman et al. 1995; Braveman et al. 1996; Soskolne et al. 1996).

the commonwealth survey and analytic methods The primary means of comparing the birth experiences of the families participating in The Commonwealth Fund Survey of Families with Young Children was through simple, stratified tabulations. To produce national estimates, variables were weighted to reflect national demographic patterns. The analysis of the Commonwealth Survey is best interpreted within a context shaped by other available data sets containing information on childbearing. Among the most useful of the national data sets is the National Longitudinal Survey of Youth (NLSY), which has followed a cohort of young people through early adulthood and contains information on the personal experiences of those who have had children.1 The Census Bureau Survey of Income and Program Participation (SIPP) regularly collects information on economic, social, and demographic changes in a national sample.2 The National Health Interview Survey (NHIS),3 the National Survey of Family Growth (NSFG),4 and the National Health and Nutrition Examination Survey (NHANES),5 as well as data sets based on vital statistics,6 provide ongoing data on the childbearing experiences of American families. A national survey with detailed information on pregnancy, birth outcomes, and young child health is the National Maternal and Infant Health Survey (NMIHS). Although somewhat dated (it was conducted in 1988), this data source continues to provide important information on a number of demographic, social, and health-related variables (Sanderson and Gonzalez 1998). Its utility was enhanced recently by

1 2 3 4 5

Information available on website www.bls.gov/nlsnew.htm Information available on website www.sipp.census.gov/sipp/sipphome.htm Information available on website www.cdc.gov/nchswww/about/nhis/nhis.htm Information available on website www.cdc.gov/nchswww/about/nsfg/nsfg.htm Information available on website www.cdc.gov/nchswww/prducts/catalogs/subject/ nhanes3/nhanes.htm 6 Information available on website www.cdc.gov/nchswww/about/major/natality/ natality.htm

274

Wise

three-year follow-up information on the infants described in the original 1988 survey. Together, these data sets provide important insights into a variety of factors influencing childbirth in the United States. Childbearing, however, is rarely the principal focus of the surveys, and they provide little information regarding parental perceptions of the birth experience and satisfaction with medical services. The current survey therefore not only represents a rich source for demographic and service-based analysis but also provides special empirical understanding of the more qualitative aspects of childbearing, thereby presenting a far more textured insight into childbearing in the United States today. Prenatal Care Use The Commonwealth Survey reveals that some 91 percent of women began prenatal care within the first three months of pregnancy. Forty-one percent began care in the first month of pregnancy, and another 38 percent during the second month. Approximately 1 percent of all women in the survey reported that they had started prenatal care just before birth or had received no prenatal care at all. When these prenatal care utilization data are grouped into two categories – care beginning in the first trimester and care beginning in the second trimester or later, including no prenatal care – approximately 7 percent, or roughly one in every 14 surveyed births, began prenatal care relatively late in pregnancy. Although these overall findings suggest broad provision of early prenatal care, a closer look reveals significant social variations in prenatal care utilization in the surveyed population. Prenatal Care Use and Social Factors. Household income is highly associated with the initiation of prenatal care (Figure 9.1). For women with household incomes of less than $20,000 per year, only 84 percent began prenatal care in the first trimester, while 95 percent with incomes above $40,000 began care during this early part of pregnancy (p < 0.01). This disparity is even more prominent for initiation of care in the first month of pregnancy: 36 percent of lower income respondents started care in the first month, compared with 46 percent of the higher income group (p < 0.01). Late initiation or no prenatal care was significantly higher among the lower income groups. For women with household incomes of less than $20,000 per year, one out of every eight women initiated care later than the first trimester or received none at all; for women with incomes less than $10,000, the figure was almost one in five. For women with incomes greater than $40,000, the comparable figure was one out of 50 (p < 0.001). In other words, women from the poorest group had late or no prenatal care at almost ten times the rate of those in the wealthier group (p < 0.001).

Prenatal Care, Delivery, and Birth Outcomes

275

50 45 40 35 Percent

30

first second third late or none

25 20 15 10 5 0 <$20,000

$20-39,000

>$40,000

Household Income

figure 9.1 Month of Prenatal Care Initiation by Household Income.

Racial disparities in prenatal care are also apparent (Table 9.1). While 93.7 percent of white non-Hispanic women began prenatal care in the first trimester, 88.6 percent of black non-Hispanic women (p < 0.01) and 82 percent of Hispanic women (p < 0.01) began care during this period. Only 3.7 percent of white non-Hispanic women started care late or had no care, while 8.1 percent of black non-Hispanic women (p < 0.01) and 10.1 percent of Hispanic women had late or no care (p < 0.001). Although the number of Asian women included in the survey is too small to provide definitive insight into their prenatal care use, the data suggest that Asian women tended to start care relatively later in pregnancy than women of other racial groups, but rarely received very late or no prenatal care. Women who reported planning their pregnancies were more likely to initiate prenatal care early. This was particularly evident during the first month of pregnancy, with 43 percent of women with planned pregnancies starting care in the first month as compared with 27 percent for women who did not plan their pregnancies (p < 0.01). However, by the end of the first trimester, the figures for planned and unplanned pregnancies were 93 percent and 87 percent, respectively, suggesting some catch-up in prenatal care use by women who did not plan their pregnancies (not statistically significant). Maternal age is also associated with prenatal care initiation. Among women less than 20 years of age, 85 percent began care in the first trimester; the figure for women greater than 34 years of age was 94 percent. Rates of

Wise

276 table 9.1. Prenatal Care Initiation by Health Insurance Coverage (%) Prenatal care initiation a

Early

Lateb

Unknown

Race/ethnicity White/non-Hispanic Black/non-Hispanic Hispanic

93.7 88.6 85.0

3.7 8.1 10.1

2.4 3.2 4.9

Insurance status Employer HMO PPO Non-managed care Other private HMO PPO Non-managed care Public HMO PPO Non-managed care Uninsured

94.4 92.7 95.4 92.1 91.1 89.3 91.8 90.8 83.3 89.9 94.2 83.1 75.4

3.0 2.8 2.5 3.1 4.2 5.1 4.0 4.3 11.0 6.1 4.5 11.1 18.1

2.6 4.5 2.2 4.8 3.9 4.5 3.5 4.0 5.6 4.0 1.3 5.8 6.5

= First trimester. = Second trimester or later (including no prenatal care).

a Early b Late

receiving care late or not at all were 14 percent for teenagers and 6 percent for women over age 34 (p < 0.01). Young teens – those less than 18 years of age – had the lowest rate of first trimester initiation, at just 64 percent, while 21 percent received late or no care (p < 0.01). The relationship between maternal education and prenatal care initiation is similar to that found for household income. Women with less than a high school diploma had a first trimester initiation rate of 74 percent, while that for college graduates was 97 percent (p < 0.001). The late or no prenatal care figures were a striking 17 percent for women with less than high school education and 1 percent for those who graduated college (p < 0.01). Slight differences may be noted in prenatal care initiation by region of the country. The eastern part of the United States had the best first trimester initiation record, at 92.6 percent, while women in the Pacific region reported a comparable 87.8 percent (p < 0.01). Health Insurance and Prenatal Care Use. Lack of health insurance coverage significantly elevates the chance that a woman will begin prenatal care late or receive no prenatal care at all (Table 9.1). Approximately 75 percent of women without health insurance began prenatal

Prenatal Care, Delivery, and Birth Outcomes

277

care in the first trimester of pregnancy, while 94.4 percent of women covered by employer-sponsored insurance began prenatal care during the first trimester (p < 0.001). Women with other forms of private insurance had a first trimester rate of 91.1 percent (not significantly different from employer-sponsored), while the rate for women with publicly funded insurance (primarily Medicaid) was 83.3 percent (p < 0.001). Women without insurance were particularly likely to have late or no prenatal care. Three percent of women with employer-sponsored insurance began late or had no prenatal care, while other privately insured women reported 4.2 percent late or no care (not significantly different); for publicly insured women, the figure was 11.1 percent (p < 0.001), and 18.1 percent for uninsured women (p < 0.001). Enrollment in managed care has a mixed impact on prenatal care utilization. Among women with employer-based coverage, first trimester prenatal care initiation was similar for those belonging to a health maintenance organization (HMO), at 92.7 percent; preferred provider organization (PPO), at 95.4 percent; and neither type of managed care system, at 92.1 percent (not significantly different from HMO or PPO). Among those covered by publicly sponsored insurance, however, women in HMO and PPO plans had better first trimester initiation rates (89.9 percent and 94.2 percent, respectively) than women with non-managed care insurance coverage (83.1 percent). However, because the number of publicly insured women enrolled in managed care systems is relatively small, this difference is not statistically different. When insurance covered only part of prenatal care services, slight differences in prenatal care utilization patterns were noted; a far greater effect occurred when no prenatal services were covered or the woman was uninsured. Multivariate Analysis of Prenatal Care Use. Factors found to be associated with the late initiation of prenatal care tend to cluster together in American society, and indeed were highly interactive within the surveyed population. Multivariate models were therefore constructed to examine insurance coverage, race, maternal age, education, and income simultaneously. The results of these models are summarized in Table 9.2. Not surprisingly, the variables examined were found to be closely linked. Very low income (below $10,000 per year) and lack of a high school diploma continued to be highly associated with poor prenatal care utilization, even after other factors, including insurance coverage, were included in the model. Similarly, lack of health insurance remained an important factor. However, the importance of teen pregnancy fell away once income and insurance status were assessed. (Maternal education was truncated because women less than 18 years of age, virtually by definition, have less than a high school diploma.) Whether or not the pregnancy was planned was also examined using a variety of models, and could not be found to

Wise

278

table 9.2. Association of Selected Variables with Late Initiation of Prenatal Care: Unadjusted (Univariate) and Adjusted (Logistic Regression Model) Odds Ratios Unadjusted Adjusted 95% Confidence odds ratio odds ratio interval Significance Insurancea Public insurance Uninsured

3.9 7.8

1.4 3.2

(1.1, 2.8) (1.8, 5.1)

p < 0.05 p < 0.01

Race / ethnicityb Black Non-Hispanic Hispanic Maternal agec < 20 years < 18 years

2.2 3.1

1.7 1.4

(1.2, 2.5) (0.8, 2.3)

p < 0.05 n.s.

2.5 4.1

1.2 1.9

(0.7, 2.4) (1.1, 3.2)

n.s. p < 0.05

Incomed < $20,000 < $10,000

6.8 9.4

2.2 4.2

(1.0, 4.1) (2.1, 6.6)

p = 0.05 p < 0.001

15.3

3.5

(1.3, 5.8)

p < 0.05

Maternal educatione < High school

= employer-based. = White Non-Hispanic. c Standard = 20–30 years. d Standard = > $40,000. e Standard = some college.

a Standard b Standard

be related significantly to late prenatal care initiation. These findings suggest that great caution should be exercised when invoking isolated factors to explain poor prenatal care utilization. Rather, demographic and health service variables are likely to be highly interactive, reflecting the pervasive expression of social status in American daily life. For several reasons, care should also be taken in interpreting the meaning of risk factors such as young maternal age, low income, or low maternal education and translating those findings into highly targeted service delivery strategies. First, although demographic variables may be interrelated, they are not necessarily coincident. For example, while black women in the survey were more likely than other groups to have less than a high school education, the vast majority of black women have at least a high school diploma. Second, a variable may be associated with an elevated risk for a particular adverse outcome, yet may make only a small contribution to the actual number of those outcomes. For example, although having no insurance coverage was associated with an elevated risk for late prenatal care initiation, almost 80 percent of all those starting prenatal care late were, in fact, covered by some form of insurance. Similarly, although adolescents have an elevated risk for low birthweight, they contribute only a small portion of all low birthweight infants in the United States. These relationships occur

Prenatal Care, Delivery, and Birth Outcomes

279

because the vast majority of births occur to women who are over 20 years of age and have health insurance. Therefore, while highly targeted initiatives could address the needs of selected high-risk groups, their overall effectiveness could be enhanced by coupling them with broader strategies likely to improve the outcomes of larger populations. Attendance at Childbirth Classes Overall, 71 percent of the total surveyed population reported attending some form of childbirth classes, although attendance was heavily influenced by social factors. Seventy-seven percent of white non-Hispanic parents attended classes, compared with 54 percent of black non-Hispanics (p < 0.001) and 53 percent of Hispanics (p < 0.001). Parental education is also an important influence on attendance at childbirth classes. College graduates reported an attendance rate of 86 percent, while those with less than a high school diploma reported only 34 percent attendance (p < 0.001). Comparable disparities emerge when attendance is stratified by household income and parental age, with the poorest and youngest groups reporting attendance of 43 percent (p < 0.001) and 33 percent (p < 0.001), respectively. Childbirth class attendance was higher among parents with planned pregnancies (75 percent) than among those with unplanned pregnancies (61 percent) (p < 0.01). The type of insurance coverage is also associated with attendance at childbirth classes. Parents in public insurance programs were less likely to attend childbirth classes (46 percent) than parents with employersponsored insurance (75 percent) (p < 0.001). Uninsured parents were the least likely to attend childbirth classes, with only 42 percent reporting attendance (p < 0.001). Parents enrolled in managed care systems were slightly more likely to attend classes than either the employer-sponsored or the publicly insured groups (not significantly different). Type of Delivery Among the survey sample, 77 percent of all births were delivered vaginally and 23 percent were delivered by cesarean section, rates roughly similar to those reported nationally on birth certificates. Slight and not significant regional differences in cesarean delivery were noted, with the Eastern region reporting a rate of 24.4 percent; the Central region, 26.1 percent; the Mountain region, 18.3 percent; and the Pacific region, 21.2 percent. Births to black non-Hispanic women were more likely to have been delivered by cesarean section than were births to white non-Hispanic or Hispanic women. For deliveries among white non-Hispanic women, the cesarean section rate was 20.7, while the rate for black non-Hispanic women was 26.6 percent (p < 0.05) and the rate for Hispanic deliveries

280

Wise

was 21.7 percent (not significantly different from white non-Hispanic). These racial differences are related to the influences of several other social variables. Using multivariate models, wealthier women were found to be slightly more likely than poorer women to undergo cesarean delivery, but the disparity is largely accounted for by the fact that wealthier women tend to be somewhat older than the poorer group. Uninsured women had a cesarean section rate of 25.6 percent, some 17 percent higher than women with employer-sponsored insurance (p < 0.05) and 34 percent higher than women covered by public insurance (p < 0.01). Significantly, the highest rates of cesarean section occurred in women who received very late (ninth month of pregnancy) or no prenatal care (29.1 percent) (p < 0.01)). Although the precise mechanisms linking prenatal care and cesarean section remain unclear, there exists the possibility that the lack of prenatal care may elevate the risk of clinical conditions necessitating cesarean section (such as fetal distress, maternal hypertension, or hemorrhage) or that the lack of a clinical relationship between patient and clinician may affect clinical decision making regarding delivery method. Although specific information on vaginal birth after cesarean ( VBAC) is not included in the survey, data on the presence of other children in the household were used as a proxy for parity to estimate cesarean deliveries in second or higher order pregnancies. Using this method in multivariate models that included age, race, income, insurance coverage, and parity, large differences in cesarean section rates were noted. Black non-Hispanic, Hispanic, and lower income women had higher relative cesarean rates in second or higher order pregnancies than they did for first births. Although these estimates must be interpreted with caution, they suggest that VBAC may be far more likely to occur in white non-Hispanic and wealthier groups. Prematurity and Low Birthweight The rate of premature and low birthweight births in the survey sample is 8.4 percent. Although this figure is somewhat lower than national low birthweight rates, it may include premature infants who did not have low birthweights. Analysis of prematurity and low birthweight in the study sample is useful, but the relatively small number of such infants makes detailed, stratified analyses difficult to interpret. Prematurity and low birthweight were more common among black non-Hispanic births (12.7 percent) than among white non-Hispanic (8.3 percent) (p < 0.05) or Hispanic births (9.7 percent) (not significantly different from white non-Hispanic). Household income was also associated with prematurity and low birthweight: 12.2 percent of births to women with incomes less than $20,000 were premature or low birthweight,

Prenatal Care, Delivery, and Birth Outcomes

281

compared with only 7.8 percent among women with incomes greater than $40,000 (p < 0.05). In part, these disparities were mediated by differences in prenatal care utilization patterns. When prenatal care initiation was controlled (along with lesser influences, such as maternal age), racial and income differentials were reduced but not eliminated. Findings on Postpartum Length of Hospital Stay Approximately half of all women in the survey who delivered their infants vaginally remained in the hospital for one day or less (Figure 9.2). Another 37 percent remained two days, while the remaining 12 percent remained in the hospital for three days or more. Among women who delivered by cesarean section, 21 percent remained hospitalized for two days or less, 41 percent for three days, and 37 percent for four days or more. In accordance with the documented national trend toward reduced postpartum stays, women who gave birth earlier in the survey period reported longer hospital stays than did those who had delivered more recently. This was noted for both vaginal and cesarean deliveries. Demographic Differences in Early Postpartum Discharge. Early discharge rates showed major regional differences. In the Pacific region, 66 percent of all vaginal births were discharged within one day of birth, while the comparable figure for the Eastern region was 41 percent (p < 0.001). The Pacific region also had the highest concentration of managed care. Although managed care systems were associated with somewhat higher rates of first-day discharge, early discharges were higher for all forms of insurance coverage in the Pacific region. Early discharge was not confined to wealthier women of apparent low risk. Hispanic women, including those not speaking English, were more likely to be discharged from the hospital within one day of delivery (55 percent) than were white non-Hispanic women (48 percent) (p < 0.05) or black non-Hispanic women (38 percent) (p < 0.05). These differences moderated somewhat by the second postpartum day. High rates of early discharge among Hispanic women and low rates among black non-Hispanic women were associated with a greater concentration of Hispanic interviewees in the Pacific region (the area with the highest rates of early discharge) and black non-Hispanic women in the Northeast (the area with the lowest early discharge rates). Indeed, when multivariate models were constructed, no significant differences in early discharge were found among the examined groups (Table 9.3). Women and infants with household incomes of less than $20,000 were as likely to be sent home within one day of birth as women with incomes greater than $40,000. Similar findings were noted for levels of maternal education. Even women with late or no prenatal care were as likely to be

Wise

282 Length of Postpartum Stays Vaginal Deliveries

3 Days or More 12%

2 Days

1 Day or Less

37%

51%

Length of Postpartum Stays Cesarean Delivery Unk

2 Days or Less 21% 4 Days or More 37%

3 Days 41%

figure 9.2 Maternal Hospital Stay for Vaginal and Cesarean Section Deliveries.

Prenatal Care, Delivery, and Birth Outcomes

283

table 9.3. Association of Selected Variables with Early Discharge, Readmission, and Not Breastfeeding: Adjusted (Logistic Regression Model) Odds Ratios and 95% Confidence Intervals

Early Dischargea Blackb Hispanicb Maternal age < 20c Unmarriedd Income < $20,000e Paternal unemployment f Ruralg Public insuranceh No insuranceh

Early discharge odds ratio (95% confidential interval)

Readmission odds ratio (95% confidential interval)

Not breastfeeding odds ratio (95% confidential interval)

— 0.8 (0.5, 1.6) 1.3 (0.8, 1.9) 1.2 (0.5, 2.2) 0.9 (0.6, 1.6) 1.4 (0.8, 2.1) 1.4 (0.7, 2.3)

1.3 (0.8, 2.4) 0.6 (0.3, 1.5) 0.8 (0.4, 1.7) 1.2 (0.5, 2.2) 1.5 (0.8, 3.1) 0.5 (0.2, 0.9)∗ 2.4 (1.1, 4.2)∗

0.8 (0.6, 0.9)∗ 3.0 (2.2, 4.1)∗∗ 0.9 (0.7, 1.2) 1.7 (1.0, 2.9)∗ 1.8 (1.3, 2.4)∗ 1.1 (0.8, 1.9) 1.9 (0.9, 4.4)

0.7 (0.3, 1.8) 0.9 (0.6, 1.7) 1.4 (0.8, 3.2)

2.2 (0.7, 6.6) 2.9 (2.3, 6.5)∗∗ 1.8 (0.7, 4.6)

0.7 (0.3, 1.9) 1.8 (1.1, 3.2)∗ 1.5 (0.9, 3.6)

a Maternal

hospital stay < 48 hours; standard > 48 hours. = white non-Hispanic. c Standard = maternal age 35+. d Standard = married. e Standard = income > $40,000. f Standard = employed. g Standard = urban. h Standard = employer-based. ∗ p < 0.05. ∗∗ p < 0.01. ∗∗∗ p < 0.001. b Standard

sent home during the first postpartum day as women who received early prenatal care, once birthweight and gestational age were controlled. In addition, early discharge was not more likely to occur among women who participated in childbirth classes. Early Discharge, Parent Information, and Infant Health. Approximately 20 percent of the surveyed families received a home visit soon after childbirth. The vast majority of women and infants discharged from the hospital less than 48 hours after birth did not receive a home visit, although they were slightly more likely to receive a home visit than those discharged later (p > 0.05). Importantly, adolescent or low income women who were discharged early were not more likely to receive a home visit than their older or wealthier counterparts. In addition, there were no differences in home visitation based on insurance coverage among women discharged early. Multivariate models with home visit as the dependent

284

Wise

variable showed that neither young maternal age, low income, nor insurance status affected home visitation among women who were discharged early. While short postpartum stays were both widespread and not apparently confined to families of low social risk, there were also indications that short stays may be associated with reduced provision of information on breastfeeding and child care. Overall, 77 percent of all surveyed interviewees reported having received some information on breastfeeding during the hospitalization. However, for stays of less than 1 day, the rate was only 60 percent (p < 0.01). Significantly, for early discharges among black non-Hispanic women and low income women, only half reported receipt of breastfeeding information. Rates for these groups were even lower for receiving information related to calming the baby and sleep position (important in preventing sudden infant death syndrome). Among the greatest concerns regarding short postpartum stays is the potential development of neonatal health problems. Although hospital readmission can occur for neonates regardless of the length of the postpartum stay, it is useful to consider whether or not readmission is more likely among infants discharged relatively soon after birth. Overall, regardless of the length of postpartum stay, infants born to unmarried women, unemployed fathers, parents living in rural areas, or parents not covered by public insurance had significantly elevated readmission rates. A simple comparison shows that infants discharged prior to 48 hours after birth were significantly more likely to be readmitted than those discharged at greater than 48 hours (p < 0.05). After controlling for social variables in the multivariate, however, this finding is no longer statistically significant, perhaps because of the relatively small number of births included in the survey. What remained significant in the multivariate models were paternal unemployment compared with employed fathers (p < 0.05) and public insurance (p < 0.05). Home visits had no apparent effect on the risk of readmission among either early or late discharges. Although women discharged less than 48 hours after delivery reported receiving less information on breastfeeding during their hospital stay, they were more likely to have breastfed their infants (Table 9.3). This relationship occurred despite controlling for the social variables in the multivariate model. Black, young, unmarried, and publicly insured women were less likely to have breastfed. Parent Concerns about Early Discharge. Many interviewees voiced concerns regarding the adequacy of the postpartum stay. Almost one in five felt that their newborn spent too little time in the hospital; one in three of those discharged in one day or less described the length of stay as too short (Figure 9.3). Among those who remained in the hospital for at least

Prenatal Care, Delivery, and Birth Outcomes

285

35

30

Percent

25

20

15

10

5

0 1 Day or Less

2 Days

3 Days or More

figure 9.3 Parents Reporting That Their Newborn Spent Too Little Time in the Hospital.

3 days, only 4 percent felt their newborn spent too little time in the hospital. Although this pattern of perceptions occurred roughly equally across all social groups, poorer and less educated women were somewhat more likely to feel that the length of stay was inadequate. Interestingly, despite concerns regarding length of stay, most respondents reported high levels of satisfaction with hospital care regardless of the number of days in hospital postpartum. Overall, 76 percent reported that they were “very satisfied” with the hospital care they received. Levels of satisfaction were similar among women who were delivered by cesarean section, who had a premature or low birthweight infant, and who were discharged within one day of birth. No significant differences in satisfaction levels were found for different social groups or types of insurance coverage, using multivariate models. Why satisfaction should be high even among parents who thought the length of stay was too short is not clear. It is possible that early discharge was a concern but not sufficiently important to affect general satisfaction with the total childbearing experience. Regardless of the precise reasons why such an apparently discordant set of opinions could occur, the complex perceptions regarding early discharge warrant cautious examination.

286

Wise

policy implications The provision of prenatal care in the United States remains an important challenge to current public policies. Although most pregnant women begin prenatal care within the first few months after conception, findings from the Commonwealth Survey underscore the profound social disparities that exist in prenatal care provision. Poor and minority women remain far more likely to start care later than their wealthier counterparts. The data suggest that uninsured women are at particularly high risk for poor prenatal care utilization. Disparities in prenatal care are particularly troubling in the current climate of rapid progress. Prenatal diagnostic techniques have made remarkable strides in recent years. Prevention of transmission of the human immunodeficiency virus through prenatal therapies represents a new and important clinical capability (Cooper et al. 1996), while insights into some of the more important determinants of prematurity may lead to major new mechanisms of prevention (Hauth et al. 1995; Hillier et al. 1995). As efficacy grows, so too does the burden of public policy to provide services equitably to all those in need. Clearly, policies that ensure insurance coverage for prenatal care services would be constructive. The survey data suggest, however, that insurance coverage is likely to be only part of the answer, since poor women with insurance reported high levels of receiving late or no prenatal care. The data confirm other assessments that barriers to prenatal care are not entirely financial in nature (Guyer 1990). Other socially defined issues such as language, lack of transportation, and lack of opportunity to seek care for working women may all represent meaningful impediments to care (Curry 1989). Perhaps the most important barrier to widely accessible prenatal care is its deep isolation from other arenas of women’s health care. For many women, the prenatal care delivery system involves both providers and bureaucracy entirely different from those that apply when seeking other forms of health care. Particularly for poor women, the fragmentation of young women’s health care in the United States means that contraception, abortion, and primary medical care services may be poorly coordinated (Chavkin 1995; Wise et al. 1995). Indeed, for most public insurance programs, young adult women become eligible for coverage only after conception and are dropped soon after delivery, greatly undermining continuity of care (Chavkin et al. 1995). The best guarantee of early initiation of prenatal care may be a close, ongoing relationship with the health care system long before pregnancy occurs. Policies that would promote such a seamless entry into prenatal care should therefore promote a more integrated vision of services and a broader commitment to the general health of reproductive-age women.

Prenatal Care, Delivery, and Birth Outcomes

287

The survey data also raise disturbing questions about the implementation of early obstetrical discharge in the United States. First, the data suggest that short obstetrical stays are equally common among women at high risk and wealthier, older, and more educated women. Although the data could not speak definitively to the impact of the pattern of early discharge, the lack of apparent differentiation in the length of stay may imply that the increased needs of certain groups of women and infants are not being fully met. The inadequate provision of information to many poor parents underscores this concern, as does the prospect of their inferior access to young child and maternal health care soon after discharge. Although these concerns require urgent policy attention, it is important to recognize that despite short hospital stays and lack of adequate information, the vast majority of parents said they are happy with the care they received. This finding provides a useful context for the deliberation of policies designed to alter current patterns of childbearing practices in the United States. The survey suggests that much is right about the current system of care for pregnant women and infants, and that caution should be exercised in protecting elements of care that are both effective and appreciated by patients and their families. More fundamentally, however, respondents’ general satisfaction with the care they received may reflect the joy and appreciation that almost always greet the birth of a healthy child.

conclusions The Commonwealth Survey suggests that the experience of childbirth in the United States reflects both the collective experiences and the social distinctions of daily life in America today. Not surprisingly, the findings convey a mixed message, one that speaks to the widespread use of many important services while at the same time providing striking testimony of the inequitable use of others. Several findings convey important lessons about the strengths and disparities inherent in our health care system: r The vast majority of women in the United States start prenatal care early in pregnancy, yet powerful social disparities exist in prenatal care utilization. r Although cesarean deliveries account for one in five births in the United States overall, major variations characterize rates of cesarean delivery by insurance type and where one lives. r Despite major strides in the care of critically ill premature and low birthweight newborns, the birth rate of these infants remains relatively high, particularly for low income and African American births. r A postpartum stay of 1 day has become common throughout the United States, but short stays are not associated with the widespread use of home-based services or confined to groups at low social risk.

288

Wise

r Despite concerns regarding the short duration of postpartum stays, most respondents reported high rates of satisfaction with their childbirth experience. Together, these findings suggest that the development and refinement of policies directed at improving the experience of childbirth in the United States should be approached with care. The juxtaposition of failure and success indicates that reform of current policies should be tied to efforts to protect services that have been beneficial. It is also essential to recognize that the effectiveness of policies designed to improve childbearing is inherently dynamic. The logic of policy is such that each new clinical advance may alter the determinants, and even the requirement, of equitable provision. These relationships are especially important in light of the promise of new, more efficacious interventions to improve maternal and infant health. The survey findings survey also suggest that profound social inequities in American society have found deep expression in the experience of childbirth. In general, patterns of utilization and outcomes were heavily influenced by race, income, and health insurance status. Therefore, although the complex findings of the survey suggest that policy reforms be developed with precision, the inherent tragedy of socially disparate experiences also demands that such reforms be advanced with urgency.

references Aday, L.A., & Andersen, R. (1974). A framework for the study of access to medical care. Health Services Research, 9, 208–20. Alexander, G.R., & Kotelchuck, M. (1996). Quantifying the adequacy of prenatal care: a comparison of indices. Public Health Reports, 111, 408–18. Alexander, G.R., & Howell, E. (1997). Preventing preterm birth and increasing access to prenatal care: Two important but distinct national goals. American Journal of Preventive Medicine, 13, 290–1. American College of Obstetricians and Gynecologists. (1993). Standards for Obstetric-Gynecologic Services, 7th ed. Washington, DC. Andersen, R.M., McCutcheon, A., Aday, L.A., Chiu, G.Y., & Bell, R. (1983). Exploring dimensions of access to medical care. Health Services Research, 18, 49–74. Anderson, R.N., Kochanek, K.D., & Murphy, S.L. (1998). Births and Deaths: US, July 1996-June 1997. Monthly Vital Statistics Report, 46, 44. Baldwin, L.M, Larson, E.H., Connell, F.A., Nordlund, D., Cain, K.C., Cawthon, M.L., Byrns, P., & Rosenblatt, R.A. (1998). The Effect of Expanding Medicaid Prenatal Services on Birth Outcomes. American Journal of Public Health, 88, 1623–1629. Beardslee, W., Bemporad, J., Keller, M., & Klerman, G.L. (1983). Children of parents with major affective disorder: a review. American Journal of Psychiatry, 140, 825–31. Beardslee, W. (1989). Children of parents with affective disorders. Pediatrics Review, 10, 313–17.

Prenatal Care, Delivery, and Birth Outcomes

289

Blair, P.S., Fleming, P.H., Bensley, D., Smith, I., Bacon, C., Taylor, E., Berry, J., Golding, J., & Tripp, J. (1996). Smoking and the sudden infant death syndrome: results from 1993–5 case-control study for confidential inquiry into stillbirths and deaths in infancy. BMJ, 313–6. Braveman, P., Egerter, S., Peral, M., Marchi, K., & Miller, C. (1995). Early discharge of newborns and mothers: a critical review of the literature. Pediatrics, 96, 716–26. Braveman, P., Miller, C., Egerter, S., Bennet, T., English, P., Katz, P., & Showstack, J. (1996). Health service use among low-risk newborns after early discharge with and without nurse home visiting. Journal of the American Board of Family Practitioners, 9, 254–60. Brown, S. (1993). Prenatal Care: Reaching Mothers, Reaching Infants. Washington, DC: Institute of Medicine. Buescher, P.A., Roth, M.S., Williams, D., & Goforth, C.M. (1991). An evaluation of the impact of maternity care coordination on Medicaid birth outcomes in North Carolina. American Journal of Public Health, 81, 1625–9. Carty, E.M., & Bradley, C. (1990). A randomized, controlled evaluation of early postpartum hospital discharge. Birth, 17, 199–206. Centers for Disease Control and Prevention. Morbidity and Mortality Weekly Report, (1995). Ectopic pregnancy, United States, 1990–1992. 44(3), 46–8. Centers for Disease Control and Prevention. Morbidity and Mortality Weekly Report, (1996). Summary of notifiable diseases, United States. 45(53), 1–87. Chavkin, W., Breitbart, V., & Wise, P.H. (1995). Efforts to reduce perinatal mortality, HIV, and drug addiction: Survey of the states. Journal of the American Medical Women’s Association, 50(5), 164–6. Chavkin, W. (1995). Prenatal care and women’s health. Journal of the American Medical Women’s Association, 50, 143–47. Chilmonczyk, B.A., Salmun, L.M., Megathlin, K.N., Neveux, L.M., Palomaki, G.E., Knight, G.J., Pulkkinen, A.J., & Haddow, J.E. (1993). Association between exposure to environmental tobacco smoke and exacerbations of asthma in children. New England Journal of Medicine, 328, 1665–69. Cooper, E.R., Nugent, R.P., Diaz, C., Pitt, J., Hanson, C., Kalish, L.A., Mendez, H., Zorrilla, C., Hershow, R., Moye, J., Smeriglio V., & Fowler, M.G. (1996). After AIDS clinical trial 076: The changing pattern of zidovudine use during pregnancy, and the subsequent reduction in the vertical transmission of human immunodeficiency virus in a cohort of infected women and their infants. Journal of Infectious Disease. 174, 1207–11. Curry, M.A. (1989). Nonfinancial barriers to prenatal care. Women’s Health, 15, 85–9. Daniels, N. (1985). Just Health Care. Cambridge: Cambridge University Press. Downey, G., Coyne, J.C. (1990). Children of depressed parents: An integrated review. Psychological Bulletin, 108(1), 50–76. Dyson, D.C., Danbe, K.H., Bamber, J.A., Crites, Y.M., Field, D.R., Maier, J.A., Newman, L.A., Ray, D.A., Walton, D.L., & Armstrong, M.A. (1998). Monitoring women at risk for preterm labor. New England Journal of Medicine, 338, 15–9. Epstein, J.A., Botvin, G.J., Diaz, T., & Schinke, S.P. (1995). The role of social factors and individual characteristics in promoting alcohol use among inner-city minority youths. Journal of the Study of Alcohol, 56, 39–46.

290

Wise

Expert Panel on the Content of Prenatal Care. (1989). The Content of Prenatal Care. Washington, DC: Department of Health and Human Services. Ey, J.L., Holberg, C.J., Aldous, M.B., Wright, A.L., Martinez, F.D., & Taussig, L.M. 1995. Passive smoke exposure and otitis media in the first year of life. Pediatrics, 95, 670–677. Freed, G.L., Clark, S.J., Pathman, D.E., & Schectman, R. (1999). Influences on the receipt of well-child visits in the first two years of life. Pediatrics, 103, 864–9. Goldenberg, R.L., & Andrews, W.W. (1996). Intrauterine infection and why preterm prevention programs have failed. American Journal of Public Health, 86, 781–3. Goldenberg R.L., & Rouse, D.J. (1998). Prevention of premature birth. New England Journal of Medicine. 339, 313–20. Goyert, G.L., Bottoms, S.F., Treadwell, M.C., & Nehra, P.C. (1989). The physician factor in cesarean birth rates. New England Journal of Medicine, 320, 706–9. Guyer, B. (1990). Medicaid and prenatal care: necessary but not sufficient. JAMA, 264, 2264–5. Haas, J.S., & Waszak, S.J. (1993). The effect of health coverage on use of cesarean section. JAMA, 270, 20–4. Haas, J.S., Udvarhelyi, I.S., Morris, C.N., & Epstein, A.M. (1993). The effect of providing health coverage to poor uninsured pregnant women in Massachusetts. JAMA, 269, 87–91. Hauth, J.C., Goldenberg, R.L., Andrews, W.W., DuBard, M.B., & Copper, R.L. (1995). Mid-trimester treatment with metronidazole plus erythromycin reduces preterm birth only in women with bacterial vaginosis. New England Journal of Medicine, 333, 1732–6. Hillier, S.L., Nugent, R.P., Eschenach, D.A., Krohn, M.A., Gibbs, R.S., & Martin, D.H. (1995). Association between bacterial vaginosis and preterm delivery of a low-birth-weight infant. New England Journal of Medicine, 333, 1737–42. Iams, J. (1998). Prevention of preterm birth. New England Journal of Medicine, 338, 54–6. Kempe, A., Wise, P.H., Barkan, S.E., Sappenfield, W.M., Sachs, B., Gortmaker, S.L., Sobol, A.M., First, L.R., Pursley, D., & Rinehart, H. (1992). Clinical determinants of the racial disparity in very low birth weight birth. New England Journal of Medicine, 327, 969–73. Kessner, D. (1972). Infant Mortality. New York, NY: Metropolitan Life. Kogan, M.D., Alexander, G.R., Kotelchuck, M., Nagey, D.A., & Jack, B.W. (1994). Comparing mothers’ reports on the content of prenatal care received with recommended national guidelines for care. Public Health Reports, 109, 637–46. Kogan, M.D. Martin, J.A., Alexander, G.R., Kotelchuck, M., Ventura, S.J., & Frigoletto, F.D. (1998a). The changing pattern of prenatal care utilization in the United States, 1981–1995, using different prenatal care indices. JAMA, 279(20), 1623–8. Kogan, M.D., Alexander, G.R., Jack, B.W., & Allen, M.C. (1998b). The association between adequacy of prenatal care utilization and subsequent pediatric care utilization in the United States. Pediatrics, 102, 25–30. Kotelchuck, M. (1992). A new index of prenatal care. American Journal of Public Health, 122, 134–41. Lee, K.S., Perlman, L., Ballantyne, M., Elliot, I., & To, T. (1995). Association between duration of neonatal hospital stay and readmission rate. Journal of Pediatrics, 127, 758–66.

Prenatal Care, Delivery, and Birth Outcomes

291

Mandl, K.D., Brennan, T.A., Wise, P.H., Tronick, E.Z., & Homer, C.J. (1998). Maternal and infant health: effects of moderate reductions in postpartum length of stay. Archives of Pediatrics and Adolescent Medicine, 152, 411–15. Margolis, L.H., Kotch, J., & Lacey, J.H. (1986). Children in alcohol-related motor vehicle crashes. Pediatrics, 77, 870–2. Margolis, L.H., Kotelchuck, M., & Chang, H.Y. (1997). Factors associated with early maternal postpartum discharge from the hospital. Archives of Pediatric and Adolescent Medicine, 151, 466–72. McGauhey, P.J., & Starfield, B. (1993). Child health and the social environment of white and black children. Social Science and Medicine, 36, 867–74. Meckel, R.A. (1990). Save the Babies: American Public Health Reform and the Prevention of Infant Mortality 1850–1929. Ann Arbor, MI: University of Michigan Press. Mercer, B.M., Goldenberg, R.L., Das, A., Moawad, A.H., Iams, J.D., Meis, P.J., Copper, R.L., Johnson, F., Thom, E., McNellis, D., Miodovnik, M., Menard, M.K., Caritis, S.N., Thurnau, G.R., Bottoms, S.F., & Roberts, J. 1996. The preterm prediction study: a clinical risk assessment system. American Journal of Obstetrics and Gynecology, 174, 1885–95. Misra, D.P., & Guyer, B. (1998). Benefits and limitations of prenatal measuring content. JAMA, 279(20), 1661–1662. National Center for Health Statistics. (1995). Annual Report of Vital Statistics of the United States. Hyattsville, MD: Public Health Service. National Center for Health Statistics. (1996). Health, United States, 1995. Hyattsville, MD. Public Health Service. Table 71. National Commission to Prevent Infant Mortality. (1988). Death Before Life: The Tragedy of Infant Mortality. Washington, DC: Government Printing Office. Overpeck, M.D., Hoffman, H.J., & Prager, K. (1992). The Lowest Birth-Weight Infants and the US Infant Mortality Rate: NCHS 1983 Linked Birth/Infant Death Data. American Journal of Public Health, 82(3), 441–443. Piper, J.M., Mitchel, E.F., & Ray, W.A. (1994). Expanded Medicaid coverage for pregnant women to 100 percent of the federal poverty level. American Journal of Preventive Medicine, 10, 97–102. Ray, W.A., Mitchel, E.F., & Piper, J.M. (1997). Effect of Medicaid expansions on preterm birth. American Journal of Preventive Medicine, 13, 292–7. Ray, W.A., Gigante, J., Mitchel, E.F., Edward, F., & Hickson, G.B. (1998). Perinatal Outcomes Following Implementation of TennCare. JAMA, 279, 314–16. Sallis, J.F., Alcarez, J.E., McKenzie, T.L., Hovell, M.F., Kolody, B., & Nader, P.R. (1992). Parental behavior in relation to physical activity and fitness in 9-year-old children. American Journal of Disease in Childhood, 146, 1383–7. Sanderson, M., & Gonzalez, J.F. (1998). 1988 National Maternal and Infant Health Survey: methods and response characteristics. Vital and Health Statistics–Series 2: Data Evaluation and Methods Research, 125, 1–39. Seaman, S. (1997). Putting the brakes on drive-through deliveries. Journal of Contemporary Health, Law and Policy, 12, 497–521. Seidman, D.S., Stevenson, D.K., Ergaz, Z., & Gale, R. (1995). Hospital readmission due to neonatal hyperbilirubinemia. Pediatrics, 96, 727–9. Soskolne, E.I., Schumacher, R., Fyock, C., Young, M.L., & Schork, A. (1996). The effect of early discharge and other factors on readmission rates of newborns. Archives of Pediatrics and Adolescent Medicine, 150, 373–79.

292

Wise

Stafford, R.S. (1991). The impact of nonclinical factors on repeat cesarean section. JAMA, 265, 59–63. Taffel, S.M., Placek, P.H., & Moien, M. (1990). 1988 U.S. cesarean-section rate at 24.7 per 100 births: a plateau? New England Journal of Medicine, 323, 199–200. Walters, V. (1993). Stress, anxiety and depression: women’s accounts of their health problems. Social Science and Medicine, 36, 393–402. Weissman, J.S., & Epstein, A.M. (1996). A framework for thinking about insurance status and access to care. In Weissman, J.S., & A.M. Epstein, (eds.), Falling Through the Safety Net. Baltimore, MD: Johns Hopkins Univeristy Press. Wise, P.H. (1993). Confronting racial disparities in infant mortality: reconciling science and politics. American Journal of Preventive Medicine, 9(suppl), 7–16. Wise, P.H. (1999). Efficacy and justice: The importance of medical research and tertiary care to social disparities in infant mortality. Journal of Perinatology, 19(6 pt. 2), S24–27. Wise, P.H., Wampler, N., & Barfield, W. (1995). The importance of extreme prematurity and low birthweight to US neonatal mortality patterns: implications for prenatal care and women’s health. Journal of the American Medical Women’s Association, 50, 152–55. Zuckerman, B.S., & Beardslee, W.R. (1987). Maternal depression: A concern for pediatricians. Pediatrics, 79, 110–17.

10 Access to Health Care for Young Children in the United States Paul W. Newacheck, Miles Hochstein, Kristen S. Marchi, and Neal Halfon

The health care needs of children, and of young children in particular, reflect their underlying age-specific health needs. In general, the health and development issues of young children can be distinguished from those of older children on the basis of a number of important characteristics. These include high levels of dependency on family, passage through the developmentally critical period of early childhood, and a characteristically different set of health services needs (Halfon and Hochstein 1997). Even among young children, it is useful to distinguish among stages of development, with the health needs of neonates (birth to 30 days), infants (1 month to 1 year of age), and children in the second and third years of life differing in important ways (American Academy of Pediatrics 1997). Current literature provides a useful foundation for understanding the health needs of young children and the ability of young children to get access to needed care. Yet much of the existing literature is dated or uses less than ideal age categorizations. In our analysis, which focuses specifically on children from birth to age 3, we attempt to build on previous studies by examining the role of race and ethnicity, income, and insurance status in determining access to care using new national data on access to a broad range of health and developmental services.

health and access to care among american children The Health Status of Children in the United States The health status of all children in the United States today is generally better than it was a few decades ago, continuing a trend that began in the early part of the twentieth century. In recent years, childhood mortality has declined most sharply for young children (ages 1–4), dropping from 109 in 100,000 in 1960 to 44 in 100,000 in 1992 (Klerman and Perloff 1997). Mortality from all causes has also declined for children ages 5–9, 10–14, and 15–19, 293

294

Newacheck, Hochstein, Marchi, Halfon

although improvements for these age groups have been less significant than those for young children (Klerman and Perloff 1997). Major morbidity has also declined in some instances, with particularly significant progress in areas such as lead poisoning (National Center for Health Statistics 1998). The relatively greater improvement in the health status of young children is related, at least in part, to national and state policies targeted at younger children. Disparities in health outcomes continue to exist based on income, ethnicity, or region of the country in which a child lives. Therefore opportunities continue to exist for improving the health status of young children, particularly those in low income families, minorities, and those born to young mothers. Moreover, the United States continues to lag behind other nations in mortality rates for its youngest children (Anderson 1997). Despite the fact that mortality rates and prevalence of certain morbidities have declined, younger children still face significant health and developmental hurdles. For example, 17 percent of all children in the United States are reported to have had a developmental disability during childhood (Boyle et al. 1994). Developmental disabilities range in prevalence from cerebral palsy (0.2 percent of all children) to learning disabilities (6.5 percent) and include deafness, blindness, epilepsy, cerebral palsy, stammering and stuttering, other speech defects, developmental delay, learning disabilities, and emotional or behavioral problems (Boyle et al. 1994). Despite the wide range of prevalence and the differing clinical issues associated with developmental disabilities, children with developmental disabilities all have enhanced needs for ongoing access to routine and specialized health and developmental services. The growing awareness of developmental disability is paralleled by a growing understanding of the role of social factors in the development and health of children. For example, the development and health of children in the United States is now being affected by a vast social experiment involving group child care for approximately 60 percent of all children while their mothers and fathers work (Howes 1997). The health and developmental effects of different kinds of child care arrangements, at different developmental stages, remain controversial (Howes 1997). Access to Health and Developmental Services for Young Children Just as young children appear to have particularly good health status compared with older children, by conventional measures young children appear to enjoy superior access to health care. For example, young children are more likely to have seen a physician over a given time interval (National Center for Health Statistics 1994) and have more physician contacts than older children (National Center for Health Statistics 1998). Yet such general statistics do not take into account the age-specific utilization patterns and

Access to Health Care for Young Children

295

needs of younger children. For example, the American Academy of Pediatrics recommends much more frequent child health supervision visits for infants and toddlers than for older children (1997). For all children, access involves the ability of the parent to negotiate the system on behalf of the child. In families with fewer resources or diminished capacities due to mental health problems, drug use, or family violence, the care young children receive may depend more on how people other than the children’s parents advocate for them within the system. Moreover, because so many of the developmental health needs of children are met in the family context, the family’s ability to get access to needed advice and support constitutes an important component of the child’s access to care. Effects of Race, Ethnicity, Income, and Insurance on Access to Care While access to medical care has improved for children as a whole in the United States in recent years, children who do not have sufficient resources due to family income or insurance status and ethnic minorities still face greater problems receiving appropriate care (Newacheck et al. 1997). For example, compared with children from non-poor, white, insured families, children who are either poor, members of a minority group, or uninsured are significantly more likely to lack a usual source of care, to be unable to identify a regular clinician, to delay or miss care for economic reasons, to have infrequent physician contact, to have fewer physician contacts per 100 bed days, and to be unable to get needed medical care, dental care, vision care, or mental health services (Newacheck et al. 1997; National Center for Health Statistics 1998; Ettner 1996). Continuity of care has also been demonstrated to be a problem for Hispanics at all income levels compared with whites and blacks, and for lower income whites compared with blacks (Kogan et al. 1995; Halfon et al. 1997). Income also affects the location of physician visits, with young children below the poverty level less likely to see a physician in an office setting and more likely to see a physician in a hospital setting (National Center for Health Statistics 1998; Lewit and Monheit 1992). Compared with people who have health insurance, uninsured individuals, including children (ages 1–17), are likely to be sicker upon admission to a hospital, to use more resources during hospitalization, and to suffer from higher mortality rates while in the hospital (Hadley et al. 1991). One of the most important determinants of children’s access to appropriate and timely care is having some form of insurance to pay for such care (Newacheck et al. 1996). Although race, income, and insurance all contribute independently to access, insurance coverage is the single most powerful predictor of access to care. Insurance-based access problems follow somewhat different patterns for younger and older children. For example,

296

Newacheck, Hochstein, Marchi, Halfon

from 1988 to 1992 young children (less than 6 years of age) suffered greater relative loss of employer-based insurance than did middle children (ages 6 – 11), or older children (ages 12–17) (Newacheck et al. 1995). Yet, as private employer-provided insurance eroded, expansions in Medicaid coverage during this period were particularly valuable for younger children, with 19.3 percent covered by Medicaid in 1988 and 30.1 percent covered by 1992 – an increase of 56 percent (Newacheck et al. 1998). This increase in public coverage benefited younger children more than older children, decreasing the total percentage of uninsured younger children by 17 percent, even as the percentage of uninsured children ages 6–11 dropped by only 8 percent. As a consequence of the growth in public insurance, younger children are now more likely to be insured than older children (Fronstin 1997; Holahan 1997).

data sources and methodological approach The two primary sources of data used are The Commonwealth Fund Survey of Parents of Young Children, conducted in 1995–96, and the 1993–94 National Health Interview Survey (NHIS) and its supplements on access to care and health insurance coverage. Together, these surveys provide the data needed to generate a comprehensive profile of access to health care for young children in America, including those in certain high-risk subgroups. The surveys are complementary, in that the main weaknesses of each are compensated by strengths in the other. The NHIS provides a large, nationally representative data base on health insurance and access to care for children, but it is largely limited to access and use of traditional medical services. In contrast, the Commonwealth Survey, while much smaller, provides unique information on a broad range of health services, including services designed specifically for young children and their families. These services, including home visits, developmental consultations, and health tracking, are increasingly viewed as important for children’s healthy development. Because the methodological details of the Commonwealth Survey are described elsewhere in this book, our discussion on sample design focuses primarily on the NHIS. The NHIS is a continuing nationwide household survey that is conducted using personal interviews by the Bureau of the Census for the National Center for Health Statistics (Benson and Marano 1994). The survey is designed to collect information on the demographic characteristics, health status, and health care use patterns of the U.S. civilian noninstitutionalized population. Each year, the survey instrument includes a core questionnaire on health status, use of medical services, and demographic characteristics, as well as a set of supplemental questionnaires on topics of current interest. During the final two quarters of 1993 and all of 1994, two special supplemental questionnaires on access to care and health insurance

Access to Health Care for Young Children

297

coverage were included in the survey. The analysis of NHIS data presented in this chapter draws upon the data collected in the core questionnaire and supplements. Some 8,193 children under age 3 were included in the 1993 – 94 access and insurance supplements. In the NHIS, a knowledgeable adult member of the household, typically the mother, responds to questions about the health of child family members. The overall response rate for the survey exceeded 94 percent, and item nonresponse, including missing and unknown responses, for key dependent variables was less than 10 percent. Cases with missing responses were excluded from the relevant estimates for access and utilization variables. Variable Construction We were particularly interested in assessing access to health and developmental services for children in three at-risk groups: those in poor families, members of minority groups, and those without insurance. Consequently, we identified several subgroups of children for analysis in both data sets, based on poverty status, minority status, and insurance status. Poverty status was assessed using the Bureau of the Census definition, which takes into account family income from all sources and family size. In 1994, the poverty level for a family of three was $12,320. It should be noted that because the NHIS collects income data in bands (for example, $0 to $999, $1,000 to 1,999, $2,000 to $2,999, and so forth), the poverty status measure is approximate. The Commonwealth Survey collected income in $10,000 bands and therefore provides a less exact basis for categorizing families by poverty status. Children from families with unknown or unreported income (n = 760 in the NHIS and n = 164 in the Commonwealth Survey) were excluded from the income comparisons. Race and ethnicity are based on respondent self-reports in both the NHIS and the Commonwealth survey. For our analysis, all persons identifying themselves as nonwhite or of Hispanic origin were classified as minority. There were no missing data for race and ethnicity in the NHIS. The 13 children with missing information on race in the Commonwealth Survey were excluded from race/ethnicity comparisons. In the NHIS, insurance status was assessed based on responses to a series of questions included in the health insurance supplement about coverage during the month prior to the interview date. Persons were classified as insured if they reported coverage for that month by either private insurance, Medicare, Medicaid, CHAMPUS (now Tricare), the Indian Health Service, or other public assistance program. Persons reporting no coverage from any of those sources were classified as uninsured. Children with unknown or unreported insurance status (n = 1,014) were excluded from the insurance comparisons. In the Commonwealth Survey, children were classified as insured if they

298

Newacheck, Hochstein, Marchi, Halfon

were covered by Medicaid or by health insurance through an employer or some other source at the time of the survey. Children with unknown or unreported insurance status (n = 11) were excluded from the insurance comparisons. Statistical Analysis Estimates based on the NHIS and the Commonwealth Survey presented in the tables and text have been statistically weighted to reflect national population totals. The weights, provided by the respective data collection organizations, are equal to the inverse of the sampling probability for each case, adjusted for nonresponse. The statistical significance of differences between estimates for the three at-risk groups (poor, minority, or uninsured) and the reference group (nonpoor, white, and insured) was assessed using t-tests for differences in means and proportions. In our analysis of both surveys, standard errors and test statistics were derived using software that takes into account the complex sample design of the survey (Shah et al. 1996). The reader should recognize that the confidence intervals for estimates are often quite large, especially for the high-risk subgroups. Consequently, we focus our conclusions on patterns of results rather than individual point estimates.

findings on access to care for young children Together, The Commonwealth Fund Survey of Parents with Young Children and the National Health Interview Survey allow us to profile access to and use of a broad range of health services and to determine whether certain high-risk subgroups (uninsured, minority, and poor young children) experience differential barriers to care. The surveys also provide critical information concerning satisfaction with services received by families with young children. We begin with a presentation of findings from the National Health Interview Survey, including information on presence and type of usual source of care, waiting times, ability to obtain care on evenings and weekends, satisfaction with services, unmet health needs, and use of physician services. From the Commonwealth Survey, we present information on use of infant care services, visit content, satisfaction with services, insurance coverage of selected services, and difficulties in meeting the child’s health and medical expenses. Population Characteristics According to the NHIS, an estimated 12.0 million children under age 3 lived in the United States during 1993 – 94. Of that population, 2.6 million

Access to Health Care for Young Children

299

were in families living below the poverty level, 4.5 million were minority children, and 1.1 million were uninsured. In total, 5.9 million children were poor, minority, and/or uninsured. Consequently, almost half of all young children fell into at least one of the three at-risk groups. Usual Source of Care The vast majority of children under age three (96 percent) had a usual source of care, or a place they could go for routine medical advice and treatment when sick, during 1993–1994 (Table 10.1). Although the proportion of children without a usual source of care was small, there was substantial variability between each of the three at-risk groups and the reference group (white, nonpoor and insured children). Children from poor families were five times more likely than reference group children to be without a usual source of care. Minority children were three times more likely than their counterparts in reference group families to be without a usual source of care. However, uninsured children were at greatest risk; they were much more likely than children in the reference group to be without a usual source of care in 1993– 94. Among young children with a usual source of care, more than eight in ten were reported to receive their care in a physician’s office, private clinic, or HMO (Table 10.1). Some 10 percent received their care in community and other health centers, while smaller proportions received their care in hospital outpatient clinics (including emergency rooms) and other locations. Although the majority of children in the three at-risk groups received their care in physician’s offices, private clinics and HMOs, they were far more likely than children in the reference group to receive their care in community and other health centers, hospital outpatient clinics, and other locations. The difference is particularly striking for community and other health centers. Children in the three at-risk groups were between ten and twelve times more likely than children in the reference group to receive their care at such locations. In addition, minority and poor children were three times as likely to receive their care in hospital outpatient clinics compared with white, nonpoor and insured children. There was no significant difference in use of hospital and outpatient clinics by uninsured children and children in the reference group. Access at the Child’s Usual Source of Care We assessed three measures of access and satisfaction with care for children with a usual source of care: percentage not identifying a regular doctor, nurse or other clinician, percentage without access to care on evenings and weekends at their usual source of care, and percentage of children for whom the respondent reported not being satisfied with at least one aspect

98.7

96.4

White, nonpoor, and insured children

All children under age 3

0.3

0.2

0.8 0.5 1.9

83.5

93.8

66.0 69.1 75.0

Source: Microdata from the 1993–94 National Health Interview Survey. Note: Percentages may not total 100 due to rounding. a Includes emergency room.

93.5 95.1 83.5

0.8

1.0

1.8 1.5 2.2

SE

%

SE

%

Poor children Minority children Uninsured children

Population characteristic

Physician’s office, private clinic, and HMO (n = 5,848)

2.8

1.5

4.5 4.4 1.8

%

0.3

0.3

0.6 0.5 0.7

SE

Hospital outpatient clinic (n = 207)

9.9

1.9

25.2 20.4 19.2

%

0.5

0.3

1.6 1.2 2.0

SE

Community and other health centers (n = 777)

3.9

2.8

4.4 6.1 3.9

%

0.6

0.8

0.7 0.8 1.1

SE

All other locationsa (n = 247)

Percentage distribution by site of care (n = 7,079)

Percentage with a usual source of care (n = 7,428)

table 10.1. Usual Source of Care and Site of Care: U.S., 1993–94

Access to Health Care for Young Children

301

of the child’s care (Table 10.2). The satisfaction measure is a composite indicator that combines responses to several queries (waiting time to get an appointment, waiting time to see the doctor, satisfaction with the way questions were answered, and satisfaction with the overall care received). Sample size considerations precluded analysis of each item separately. The satisfaction questions were asked only of children who had a visit to their usual source of care in the 3 months prior to the interview. Overall, approximately 86 percent of young children had a regular clinician whom they saw on repeat visits at their usual source of care. One in seven children, or 14 percent, did not have a regular clinician. There were substantial differences in likelihood of having a regular clinician, depending on whether children were in one of the at-risk groups or in the reference group. Specifically, children in the three at-risk groups were, on average, three times more likely to be without a regular clinician than were white, nonpoor, and insured children. The vast majority of young children, 92 percent, had access to care on evenings and weekends through their usual source of care. Again, there were large differences depending on whether a child was in one of the at-risk groups or in the reference group. On average, children in each of the three at-risk groups were about three times more likely than children in the reference group to be without access to after-hours care. The NHIS data suggest that most, but not all, families were satisfied with the care their children received at the usual source of care. Family respondents reported not being satisfied with at least one aspect of care for 14 percent of all young children. Again, there were differences by risk group. Families with uninsured children were most likely to report dissatisfaction with the care received by their children (20 percent), followed by children from poor families (18 percent), and minority children (16 percent). In contrast, children from white, nonpoor, and insured families were least likely to report being dissatisfied with one or more aspects of the care received by their children (12 percent). Missed Care among Young Children The NHIS included several questions concerning whether families were able to obtain services they felt they needed for their child during the year prior to the interview (Table 10.3). Overall, only a small percentage of young children were reported as needing but not obtaining services. An estimated 2 percent of all children under age 3 were reported as being unable to obtain needed medical care, 2 percent were reported as unable to get needed dental care, and 1 percent were unable to get needed prescription medications, eyeglasses, or mental health care. Taken together, approximately 4 percent of all young children were reported as not obtaining a needed service during 1993–94.

0.8

0.8

1.7 1.4 2.3

SE

8.2

4.7

14.9 12.7 13.7

%

0.6

0.5

1.4 1.1 2.1

SE

13.8

11.6

17.9 15.6 20.1

%

0.6

0.7

1.8 1.2 3.5

SE

a Question

Source: Microdata from the 1993–94 National Health Interview Survey. asked only for children visiting their usual source of care within the previous three months; aspects of care include waiting time to get an appointment, waiting time to see the doctor, satisfaction with the way questions were answered and with the overall care received.

8.0

14.1

All children under age 3

23.2 22.9 24.3

%

White, nonpoor, and insured children

Poor children Minority children Uninsured children

Population characteristic

Percentage not identifying Percentage without Percentage not satisfied with at a regular clinician after-hours medical care least one aspect of their carea (n = 6,905) (n = 6,184) (n = 4,384)

table 10.2. Access Indicators for the Usual Source of Care: U.S., 1993–94

1.2

1.6

White, nonpoor, and insured children

All children under age 3

0.2

0.2

0.5 0.3 1.0

1.5

1.1

1.5 1.2 4.1

Source: Microdata from the 1993 – 94 National Health Interview Survey.

2.4 1.4 5.0

0.2

0.2

0.3 0.3 0.8

SE

1.0

0.6

1.7 1.3 2.7

0.1

0.2

0.4 0.2 0.8

SE

%

%

%

SE

Percentage unable to get Percentage unable to needed medical care get needed dental (n = 7,448) care (n = 7,451)

Poor children Minority children Uninsured children

Population characteristic

Percentage unable to get needed prescriptions, eyeglasses, and/or mental health care (n = 7,431)

table 10.3. Missed Care: U.S., 1993–94

3.6

2.7

4.7 3.4 9.7

%

0.3

0.3

0.6 0.4 1.4

SE

Overall percentage unable to get needed care (n = 7,391)

304

Newacheck, Hochstein, Marchi, Halfon

Although the overall proportion of young children with unmet needs is small, there were significant differences in reports of unmet needs among children in the various risk categories. Looking first at medical care, children from poor families were twice as likely and uninsured children were four times as likely as children in the reference group to go without needed services. No difference was found for minority children. Unmet dental needs were most prevalent among uninsured children. Uninsured children were more than three times as likely to go without needed dental care as children in the reference group. No significant differences in unmet dental needs were found between either minority children or poor children and children in the reference group. Differences by risk group are also apparent in children’s ability to obtain needed prescriptions, eyeglasses, or mental health care. Minorities and children from poor families were more than twice as likely and uninsured children were more than four times as likely as children in the reference group to go without one of those services when needed. Overall, uninsured children faced the greatest obstacles in obtaining needed care: one in ten uninsured children was unable to obtain at least one needed health service during 1993 – 94. Children from poor families were about half as likely as uninsured children to have an unmet need but were still nearly twice as likely as children in the reference group to go without needed services. There was no statistically significant difference in overall rates of unmet needs between minority children and children in the reference group. Delayed and Missed Care Due to Cost The NHIS included an additional series of questions regarding delays in obtaining and inability to obtain needed medical care due to cost (Table 10.4). Overall, families reported that 2 percent of young children experienced delays in obtaining needed medical because of cost considerations. Less than 1 percent of children were unable to obtain needed medical care due to cost. Uninsured children faced the greatest likelihood of delay in receipt of needed medical services. Uninsured children were four times as likely as children in the reference group to delay obtaining medical care due to cost. There was no significant difference in delay in obtaining care due to cost between children from poor families or minority children and children in the reference group. Although the absolute percentages were small, differences among the risk groups were even greater when inability to obtain needed medical care due to cost was assessed. Uninsured children were forty times as likely as white, nonpoor, and insured children to go without needed care because of its cost. Children from poor families were ten times as likely and minority children five times as likely as reference

2.1

2.3

White, nonpoor, and insured children

All children under age 3

0.2

0.3

0.5 0.3 1.2

Source: Microdata from the 1993 – 94 National Health Interview Survey.

2.4 1.8 8.5 0.5

0.1

1.1 0.5 4.0 0.1

0.1

0.4 0.1 1.0

SE

%

SE

%

Poor children Minority children Uninsured children

Population characteristic

Percentage unable to get needed medical care due to cost (n = 8,189)

Percentage delaying medical care due to cost (n = 7,456)

table 10.4. Delayed Care and Inability to Obtain Care Due to Cost: U.S., 1993–94

2.5

2.1

2.8 1.9 10.0

%

0.2

0.3

0.5 0.3 1.3

SE

Percentage delaying and/or unable to get medical care due to cost (n = 7,456)

306

Newacheck, Hochstein, Marchi, Halfon

table 10.5. Average Annual Physician Contacts: U.S., 1993–94

Population characteristic

Average number of contacts per 100 bed days (n = 8,193)

Average number of contacts (n = 8,193) Mean

SE

Mean

SE

Poor children Minority children Uninsured children

8.6 7.0 4.9

0.7 0.3 0.6

127.9 120.3 78.8

16.2 13.0 20.1

White, nonpoor, and insured children

9.0

0.3

261.3

23.5

All children under age 3

8.2

0.3

168.3

11.2

Source: Microdata from the 1993 – 94 National Health Interview Survey.

group children to be reported as unable to obtain needed medical care due to cost considerations. Use of Physician Services In Table 10.5, information on volume of physician services received by young children is presented in unadjusted form and after adjusting for need as indicated by illness. The first measure shows the average annual unadjusted number of physician contacts (including telephone consultations) for children. On average, young children had 8.2 physician contacts per year. The second measure presented is the use-disability ratio, which provides a ratio of physician contacts to bed days. Bed days are days when a child spends more than half the day confined to bed due to illness. On average, children under age 3 had 168 physician contacts for every 100 days spent ill in bed in 1993–94. There was a substantial amount of variation in use of physician services by children in the three at-risk groups and the reference group. Although children from poor families and minority children differed little in their use of physician services on an unadjusted basis from children in the reference group, the discrepancy for uninsured children was particularly great; uninsured children used only about half as many physician services as children in the reference group in 1993–94. The data showing physician contacts per 100 bed days provide a more meaningful comparison of use since this utilization indicator is adjusted for need. On that basis, white, nonpoor, and insured children used twice as many physician services as poor or minority children. As with the unadjusted data, a more pronounced discrepancy exists for uninsured children. Uninsured children used only about a third as many physician services as reference group children on an adjusted basis.

Access to Health Care for Young Children

307

Satisfaction with the Child’s Usual Provider The Commonwealth Survey provides several unique measures of satisfaction with the child’s usual provider, including the overall care received, care guidance, provision of information about the child’s growth and development, ability to reach the usual provider by phone, and how well the provider listens to and answers questions (Table 10.6). Similar to the NHIS data on satisfaction, the Commonwealth Survey reveals that white, nonpoor, and insured families were less likely to report being dissatisfied with the care their child received than were families of minority, poor, or uninsured children. In terms of the health care provided, parents of minority children were twice as likely, and parents of poor and uninsured children three times as likely, as those in the reference group to classify their provider as only fair or poor. A similar pattern exists for care guidance, with minorities being twice as likely, and poor and uninsured families being three times as likely, as families in the reference group to report their child’s provider as only fair or poor. Parents of minority children were also twice as likely, while those of poor and uninsured children were three times as likely, as parents of reference group children to consider their provider as fair or poor in providing information about the child’s growth and development. Families with children in any of the three at-risk groups were also more likely than families of white, nonpoor, and insured children to be dissatisfied with their ability to reach their child’s usual provider by telephone. The steepest gradient was for uninsured children; their parents were almost three times as likely to rate their provider’s reachability as fair or poor as were parents of children in the reference group. Large differences were also apparent for satisfaction with the way the provider listened to and answered questions. Minority children were more than three times as likely, while poor and uninsured children were four times as likely, as reference group children to consider their provider to be only fair or poor at listening to or answering questions. Receipt and Need for Information on Early Childhood Development The Commonwealth Survey collected a unique set of data concerning whether a health professional provided information on several aspects of early childhood development, including disciplining the child, helping the child learn, dealing with sleep patterns, dealing with the child’s crying, and how and when to toilet train. In addition, the survey asked parents whether or not they could have used more information in each of these areas of early childhood development. Our analysis focuses on differences in their answers according to the family’s race or ethnicity, income, and insurance status.

88.2

87.3

White, nonpoor and insured children

All children under age 3

0.9

1.3

2.5 1.6 4.7

3.5

1.8

7.5 5.8 7.4

0.5

0.5

2.1 1.1 3.3

6.9

4.8

14.9 9.3 13.7

%

0.7

0.9

2.6 1.3 4.1

SE

Source: Microdata from the 1996 Commonwealth Survey of Parents of Young Children.

86.0 85.9 73.7

SE

6.9

4.7

14.4 9.8 17.9

0.7

0.8

2.6 1.4 4.6

SE

12.3

9.9

20.6 15.5 28.0

%

0.9

1.2

2.9 1.6 5.8

SE

4.9

2.7

10.7 8.2 10.2

%

0.6

0.6

2.2 1.3 3.7

SE

%

%

%

SE

Usually see Provision of same provider health care Care guidance (n = 1,818) (n = 1,588) (n = 1,581)

Poor children Minority children Uninsured children

Population characteristic

Child’s growth and Listening and development Ability to answering (information about) reach by phone questions (n = 1,578) (n = 1,554) (n = 1,593)

Considered Provider Only Fair or Poor in Following Items

table 10.6. Satisfaction with Usual Provider among Young Children with a Usual Source of Care: U.S., 1996

Access to Health Care for Young Children

309

Families with children in the three at-risk groups were least likely to have discussed these early childhood development issues with a health professional. The largest difference was between uninsured children and children in the reference group. On average, families with uninsured children were only half as likely as families in the reference group to have discussed developmental concerns with a health professional. Families with minority or poor children were somewhat better off but still less likely than white, nonpoor, and insured families to have discussed developmental issues with a health professional. More than half of all families with young children reported that they could have used more information on the developmental issues (Table 10.7). The greatest needs for additional information were for help in disciplining a child, deciding how and when to toilet train a child, and helping the child to learn. Fewer but still sizable numbers of families reported that they could have used more information addressing sleep patterns and children’s crying behavior. Need for additional information on early childhood development varied little among the three at-risk groups or between the at-risk groups and the reference group. Only in the case of dealing with children’s crying behavior did families with at-risk children report a greater need for information than children in the reference group. Receipt and Usefulness of Early Childhood Health and Related Services A battery of questions concerning receipt of health and health-related services specific to the needs of young children was included in the Commonwealth Survey. These items included whether the family received a packet of newborn information, a postpartum home visit, 24-hour telephone access, reminders concerning periodic visits and immunizations, and a booklet to track the child’s health and health care (Table 10.8). Families receiving those services were also asked whether they found the services to be very useful. Parents in each of the three at-risk groups were about 10 percentage points less likely than parents of white, nonpoor, and insured children to receive a packet of newborn information. They were also only about onehalf to two-thirds as likely to have 24-hour telephone access or to have received a booklet to track their child’s health. In contrast, parents of both poor and minority children were somewhat more likely than parents of children in the reference group to have received a postpartum home visit or reminders concerning future visits and immunizations. A majority of families reported that they found the health and health-related services shown in Table 10.8 to be very useful. This was especially true for 24-hour telephone access and a booklet to track the child’s health, which were valued by more than eight in ten families. Differences in

24.2

All children under age 3

1.1

1.7

23.5

25.8

13.6

1.1

1.6

3.1

2.5 1.7

SE

41.1

48.8

19.3

25.5 32.0

%

1.3

1.9

3.7

2.7 1.9

SE

34.6

38.1

19.1

28.0 31.1

%

1.2

1.8

3.6

2.7 1.9

SE

29.7

34.9

18.7

19.4 24.0

%

Source: Microdata from the 1996 Commonwealth Survey of Parents of Young Children.

27.5

3.0

12.7

19.4 20.9

%

2.3 1.7

17.3 20.3

SE

%

White, nonpoor, and insured children

Poor children Minority children Uninsured children

Population characteristic

1.2

1.8

3.5

2.5 1.8

SE

How/when to toilet train (n = 2,009)

44.1

43.1

47.4

47.1 46.2

%

1.3

1.9

4.8

3.2 2.0

SE

How to discipline child (n = 1,998)

55.5

54.6

58.3

57.5 58.1

%

1.3

1.9

4.7

3.1 2.0

SE

How to help child learn (n = 2,010)

30.7

27.3

36.6

36.0 36.2

%

1.2

1.7

4.7

3.0 1.9

SE

22.6

19.0

31.5

23.6 30.6

%

1.1

1.5

4.5

2.6 1.8

SE

How to deal with child crying (n = 2,008)

How to deal with sleep patterns (n = 2,002)

How to deal with child crying (n = 2,012)

How to deal with sleep patterns (n = 2,009)

How to help child learn (n = 2,010)

How to discipline child (n = 2,010)

Could have used more information

Discussed with health professional

table 10.7. Need for and Receipt of Information on Early Childhood Development: U.S., 1996

41.7

42.5

41.5

41.9 41.6

%

1.3

1.9

4.7

3.1 2.0

SE

How/when to toilet train (n = 1,996)

1.5

3.2

0.8

0.7

82.2

83.7

White, 94.3 nonpoor, and insured children

All children 89.8 under age 3

21.0

18.3

18.9

24.9

28.9

1.1

1.5

3.8

1.8

2.9

32.9

38.3

14.4

26.8

23.9

%

1.2

1.8

3.2

1.9

2.7

SE

42.9

39.3

35.3

48.1

55.6

%

1.3

1.8

4.6

2.0

3.2

SE

66.2

74.0

46.7

54.9

51.9

%

Among parents reporting receiving the service. Source: Microdata from the 1996 Commonwealth Survey of Parents of Young Children.

a

2.3

SE

%

SE

%

82.2

Poor children Minority children Uninsured children

Population characteristic

1.2

1.6

4.8

2.0

3.2

SE

Booklet to track health (n = 2,000)

55.2

51.5

54.0

61.9

57.4

%

1.4

1.9

5.3

2.2

3.5

SE

Packet of newborn information (n = 1,737)

68.7

61.4

65.3

78.0

76.0

%

2.7

4.3

10.6

3.5

4.9

SE

Home visit post partum (n = 404)

86.9

87.0

94.2

85.2

88.4

%

1.7

2.1

5.6

3.3

4.8

SE

84.8

83.2

82.3

91.4

84.8

%

1.4

2.3

6.3

1.5

3.1

SE

Visits/ shots reminder (n = 893)

24 hr phone line access (n = 595)

Visits/ shots reminder (n = 2,006)

24 hr phone line access (n = 2,007)

Home visit post partum (n = 2,010)

Packet of newborn information (n = 2,009)

Found service very usefula

Ever received

84.5

81.3

83.5

90.0

88.6

%

1.2

1.8

5.5

1.7

3.1

SE

Booklet to track health (n = 1,238)

table 10.8. Percentage of Families Reporting Ever Receiving Specific Services and How Useful They Considered Them: U.S., 1996

312

Newacheck, Hochstein, Marchi, Halfon

perceived usefulness of health and health-related services did not vary substantially between children in each of the three at-risk groups and children in the reference group. Breadth of Covered Services under Health Insurance The Commonwealth Survey collected information on types of services covered by insurance among those with health insurance coverage at the time of the survey. Coverage information was collected on three types of services: acute care visits, well-child visits, and immunizations. As might be expected, almost all children with insurance were covered for acute care visits; only 2.4 percent of insured young children were without this type of coverage. More surprising, nearly 9 percent of insured young children were reportedly without coverage for well-child care, and a slightly higher proportion was reported without coverage for immunizations. Since Medicaid covers both well-child care and immunizations, as do most managed care plans, children without coverage for these services are likely to be those insured by traditional indemnity plans. Minority children with insurance and reference group children were equally likely to lack coverage for acute care and well-child visits. However, insured minority children were less likely than reference group children to be without coverage for immunizations. Poor children with insurance were only half as likely as reference group children to lack coverage for well-child visits and only about one-third as likely to lack coverage for immunizations. These salutary findings are likely to be attributable to more pervasive Medicaid coverage among insured poor children. Difficulty Paying for Health Care The Commonwealth Survey queried families about whether they had some or a lot of trouble in paying for their child’s health care (Table 10.9). Overall, 18 percent of respondents reported that they had difficulty in paying for care. Parents of white, nonpoor, and insured children were least likely to have difficulty in paying for care (16 percent), although they differed little from parents of minority children (18 percent) and poor children (21 percent). However, parents of uninsured children were nearly three times as likely as parents of reference group children to report some or a lot of trouble in paying for care (45 percent and 16 percent, respectively).

discussion and policy implications The release of new data from the National Health Interview Survey and the Commonwealth Fund Survey of Parents of Young Children offers a unique

Access to Health Care for Young Children

313

table 10.9. Difficulties Paying for Care: U.S., 1996

Population characteristic

Percentage reporting a lot or some trouble paying for care (n = 2,011) %

SE

Poor children Minority children Uninsured children White, nonpoor and insured children All children under age 3

20.6 18.2 44.9 15.9 17.8

2.5 1.5 4.8 1.3 1.0

Source: Microdata from the 1996 Commonwealth Survey of Parents of Young Children.

opportunity to assess access to health care and use of health services by young children in America. Never before has such a rich combination of nationally representative data been available for children from birth to age 3. These data provide a current national profile of young children’s access to and use of health services, as well as a baseline for tracking and monitoring progress in improving access and use of health services over time. Taken together, the results from our analysis reveal two messages of importance for policymakers and practitioners. First, although most young children in the United States have reasonably good access to traditional health care services, barriers remain for vulnerable subgroups of the population, including minorities, children from poor families, and uninsured children. Second, although a large proportion of families would like to receive enhanced health services, including developmentally based services, relatively few now receive them regardless of income, race, or insurance status. Consequently, all young children may benefit from an enhanced array of health and developmental services. Implications for Access to Traditional Health Services The results suggest that most young children experience relatively good access to basic medical services. For example, over 96 percent of young children were reported to have a usual source of health care. The same proportion of families reported that the health care provided to their child was either good or excellent. Moreover, only 2 percent had gone without needed medical care in the past year. Health policymakers, insurers, practitioners, and parents of young children should feel justifiably proud of these favorable statistics. Despite these indications of success, barriers in getting access to health care remain for minority children, children from poor families, and children

314

Newacheck, Hochstein, Marchi, Halfon

without insurance. Throughout our analysis of the National Health Interview Survey and the Commonwealth Survey, a pattern of reduced access and use of services emerged for children in each of these groups compared with their more advantaged counterparts. Access and utilization differentials were on the order of twofold to threefold for children in the three at-risk subpopulations compared with the white, nonpoor, and insured children in the reference group. Uninsured children experienced the greatest disadvantage. They were ten times more likely to be without a usual source of care, almost five times more likely to have unmet medical needs, and used only about half as many physician services as children in the reference group. According to the National Health Interview Survey, slightly more than 1 million children under age 3 were uninsured in America in 1993–94. Based on studies conducted by the General Accounting Office and other organizations, a substantial minority of uninsured children, as many as one-third to one-half, are eligible for Medicaid but not enrolled (General Accounting Office 1996; Selden et al. 1998). There is a clear need for expanded outreach and enrollment efforts for this group of Medicaid-eligible children. Most uninsured young children, however, reside in families with incomes above the Medicaid eligibility income limits but without sufficient resources to purchase private health insurance on their own. For this group of children, a different strategy is needed. Families of these children would greatly benefit from subsidized insurance coverage, through either public or private insurance programs. Enactment of the Balanced Budget Act of 1997 provided new avenues for addressing the needs of both groups of uninsured children – those eligible but not enrolled in Medicaid and those with incomes above Medicaid eligibility cutoffs. Specifically, the law contains a number of important provisions for expanding Medicaid and private health insurance for children in families with modest incomes and for encouraging greater enrollment and participation among those eligible for Medicaid and other publicly subsidized health insurance programs. The most important provision of the Balanced Budget Act established the Children’s Health Insurance Program (Title XXI of the Social Security Act). This program enables states to provide health insurance to uninsured children up to age 19 in low income families (defined as those with incomes below 200 percent of the federal poverty level, or up to 50 percent above the state’s income eligibility limit for Medicaid, if that is higher) through expansion of existing Medicaid programs, a separate children’s health insurance plan, or a combination of both. State participation is voluntary but Congress has provided an incentive to participate in the form of a more generous federal matching rate than applies under the regular Medicaid program. To date, forty-eight states have applied for CHIP funding.

Access to Health Care for Young Children

315

Congress granted states substantial discretion in establishing eligibility, benefits, and outreach efforts for the new program. Consequently, the effectiveness of the program in reaching uninsured young children and meeting their needs for health care will depend on how states choose to use the latitude provided for program design and implementation (Halfon et al. 1999). The critical design and implementation decisions concern the extent to which states cover some or all of the target population of children from low income families, whether the scope of benefits offered meets the needs of young children, the extent to which states impose cost-sharing requirements on participating families, and the effectiveness of outreach and enrollment efforts. In addition, the Balanced Budget Act of 1997 contains two important Medicaid provisions that can be used by states to enhance the program for young children. First, states can now provide presumptive eligibility for children. This provision allows providers to begin immediate treatment of children thought to meet Medicaid eligibility criteria without having to await the sometimes lengthy process of formal eligibility determination. Second, states may now offer guaranteed Medicaid eligibility for up to a year. This provision enables families to obtain a higher level of continuity of care, even if their incomes fluctuate over the course of a year, and could help increase the likelihood that Medicaid-enrolled children will have a “medical home,” or usual source of health care. Together, these provisions offer states new opportunities for improving access and quality of care for young children in the Medicaid program. Implications for Access to Enhanced Health and Developmental Services Although most young children enjoy good access to traditional medical services, results from our analysis of new data from the Commonwealth Survey reveal that families with young children have only moderate access to the enhanced health and developmental services advocated by professional organizations such as the American Academy of Pediatrics through its child health supervisory guidelines (1997) and the federal Maternal and Child Health Bureau through its Bright Futures initiative (Green 1994). These services include providing information and anticipatory guidance on various aspects of early childhood development (such as helping the child learn or how and when to toilet train) and provision of health and health-related services specific to the needs of young children (such as a packet of newborn information, a postpartum home visit, or a booklet to track children’s health). Receipt of those enhanced health and developmental services was low for all young children in our study, not just those at disadvantage due to income, race/ethnicity, or insurance status. For example, fewer than half

316

Newacheck, Hochstein, Marchi, Halfon

of all families had discussed developmental issues with a health professional, and more than half of all families with young children reported that they could have used more information on their child’s development. Although most families received some enhanced health services, such as an initial packet of newborn information or a booklet to track their children’s health, fewer than half of families with young children received reminders concerning future visits and immunizations, had 24-hour telephone access to their provider, or had a home visit postpartum. Not surprisingly, the majority of families receiving these services found them very useful. Taken together, the findings suggest that more can be done to enhance the provision of an array of health and developmental services to families with young children. The prospects for improving the breadth and depth of pediatric services for families with young children will depend on several factors. These include recognition of the need for enhanced services by public and private payers of health care, whether enhanced services can be provided within the context of a rapidly changing and highly competitive marketplace, and whether accountability mechanisms can be put in place to ensure their delivery. At present, the American Academy of Pediatrics and Bright Futures standards have not been universally embraced, either by the major payers of care nor by many managed care organizations. The recommendations are based on a combination of clinical expertise derived from the judgements of expert panels, the evolving research literature on the importance of early childhood on long-term health and development, and what has been the traditional model of health care provision in pediatric offices. There is a growing body of empirical literature to show that providing developmental assessments for young children can improve developmental outcomes, and that the provision of anticipatory guidance about crying, discipline, and other developmental issues can change parental behavior or behavioral outcomes in children (Regalado and Halfon 2001). This growing body of scientific evidence about the benefits of services to support and promote development in young children should be useful in advocating evidence-based standards for the provision of health care and in countering arguments that enhanced health and developmental services are not medically necessary. The provision of pediatric health care increasingly occurs in managed care settings. In many cases, this means that the actual time available for an office visit has decreased, potentially limiting the ability of pediatric providers to meet basic health needs, much less provide the enhanced services and developmental guidance suggested here. Managed care does, however, provide a much more organized process for measuring the content of quality of care. Therefore, as more and more children are cared for

Access to Health Care for Young Children

317

in managed care settings, the potential to improve quality of care based on accepted standards and guidelines is heightened. As quality of care indicators such as those incorporated in the National Committee on Quality Assurance (NCQA) Health Plan Employer Data Information Set (HEDIS) are increasingly used, our ability to monitor and measure the quality of care will also improve. Unfortunately, HEDIS measures on the provision of health services to young children focus primarily on immunizations and the receipt of wellchild visits, not health and developmental services needed by young children. If those services are eventually viewed as essential elements of medically necessary care, specific measures that target their provision could be incorporated in data collection and accountability mechanisms. This might be accomplished by creating HEDIS measures that determine whether or not children receive appropriate developmental assessments at standardized intervals. Parents could then compare performance of health plans on those measures. Recently, the Foundation for Accountability (FACCT) and NCQA have joined forces to develop new measures of quality of care for children. This initiative, the Child and Adolescent Health Measurement Initiative (CAHMI), has targeted early childhood, and specifically child development, as an area for measurement development. As part of the CAHMI, FACCT has developed a survey of parents called Promoting Healthy Development, which is designed to examine anticipatory guidance and parental education, services to children at risk for developmental problems, parental counseling and its impact on parental confidence and behavior, and developmental assessments. If implemented by health plans as part of their accountability and quality improvement activities, the Promoting Healthy Development instrument would bring greater attention to the provision of developmental services as part of routine primary pediatric care.

conclusion Expansion of the Medicaid program and other mechanisms have brought substantial improvements to young children’s access to basic health care. Still, there is room for improvement in two principal areas. First, disadvantaged children, especially uninsured children, continue to face barriers in getting access to basic health services. Expanded access to health insurance through the new Children’s Health Insurance Program should help in this regard. Nevertheless, diligence is required to ensure that all young children eventually gain health insurance coverage, not just those eligible for this new program. Second, most families with young children do not currently receive the range of enhanced health and developmental services recommended by leading pediatric professional organizations. It

318

Newacheck, Hochstein, Marchi, Halfon

is therefore important that measurement and accountability efforts focus not only on gross measures of performance, like the number of well-child visits, but on the content of those visits, including specific services that have the greatest potential to enhance health and promote development. Enhanced health and development services are keys to maximizing the value of early childhood expenditures by public and private payers. Indeed, our findings indicate that we must redouble our efforts to convince private and public payers and providers to broaden their perspectives concerning coverage of enhanced health and developmental services for young children.

references American Academy of Pediatrics. (1997). Guidelines for Health Supervision III, 3rd edition. Anderson, G.F. (1997). In search of value: An international comparison of cost, access, and outcomes. Health Affairs, 16(6), 163–71. Benson, V., & Marano, M.A. (1994). Current Estimates from the National Health Interview Survey, 1993. Vital and health statistics. Series 10, no. 190. (DHHS publication no. PHS 95–1518.) Washington, D.C.: Government Printing Office. Boyle, C.A., Decoufl´e, P., & Yeargin-Allsopp, M. (1994). Prevalence and health impact of developmental disabilities in US children. Pediatrics, 93(3), 399–403. Carnegie Task Force on Meeting the Needs of Young Children. (1994). Starting Points: Meeting the Needs of Our Youngest Children. New York, NY: Carnegie Corporation of New York. Ettner, S.L. (1996). The timing of preventive services for women and children: The effect of having a usual source of care. American Journal of Public Health, 86(12), 1748–54. Fronstin, P. (1997). Expanding health insurance for children: examining the alternatives. EBRI Issue Brief no. 187. Washington, DC: Employee Benefits Research Institute. General Accounting Office. (1996). Health insurance for children: private insurance coverage continues to deteriorate. (Publication no. GAO/HEHS-96-129.) Washington, D.C Green, M. (ed). (1994). Bright Futures: Guidelines for Health Supervision of Infants, Children and Adolescents, Arlington, VA: National Center for Education in Maternal and Child Health. Hadley, J., Steinberg, E.P., & Feder, J. (1991). Comparison of uninsured and privately insured hospital patients. Condition on admission, resource use, and outcome. Journal of the American Medical Association, 265(3), 374–9. Halfon, N., Wood, D.L., Valdez, R.B., Pereyra, M., & Duan, N. (1997). Medicaid enrollment and health services access by Latino children in inner-city Los Angeles. Journal of the American Medical Association, 277(8), 636–41. Halfon, N., & Hochstein, M. (1997). Developing a system of care for all: What the needs of vulnerable children tell us. Recent trends in the health of U.S. children.

Access to Health Care for Young Children

319

In R. Stein (ed.), Health Care for Children: What’s Right, What’s Wrong, What’s Next. New York, NY: United Hospital Fund. Halfon, N., Inkelas, M., DuPlessis, H., & Newacheck, P.W. (1999). Challenges in securing access to care for children. Health Affairs, 18(2), 48–63. Holahan, J. (1997). Expanding Insurance Coverage for Children. Washington, DC: Urban Institute. Howes, C. (1997). Children’s experiences in center-based child care as a function of teacher background and adult:child ratio. Merrill-Palmer Quarterly, 43(3), 404–25. Klerman, L.V., & Perloff, J.D. (1997). Recent trends in the health of U.S. children. In R. Stein (ed.), Health Care for Children: What’s Right, What’s Wrong, What’s Next. New York, NY: United Hospital Fund. Kogan, M.D., Alexander, G.R., Teitelbaum, M.A., Jack, B.W., Kotelchuck, M., & Pappas, G. (1995). The effect of gaps in health insurance on continuity of a regular source of care among preschool-aged children in the United States. Journal of the American Medical Association, 274(18), 1429–35. Lewit, E.M., & Monheit, A.C. (1992). Expenditures on health care for children and pregnant women. The Future of Children, 2, 95–114 National Center for Health Statistics. (1998). Health, United States, 1998, with Socioeconomic Status and Health Chartbook. Hyattsville, MD. National Center for Health Statistics. (1994). Public Health Service unpublished data. In Child Health USA, ‘96–’97. US Department of Health and Human Services, Public Health Services. Newacheck, P.W., Hughes, D.C., & Cisternas, M. (1995). Children and health insurance: an overview of recent trends. Health Affairs, 14(1), 244–54. Newacheck, P.W., Hughes, D., Pearl, M., & Halfon, N. (1998). Medicaid and children: A decade of change. Journal of the American Medical Association, 280(20), 1789–93. Newacheck, P.W., Hughes, D.C., and Stoddard, J.J. (1996). Children’s access to primary care: Differences by race, income, and insurance status. Pediatrics, 97(1), 26–32. Newacheck, P.W., Stoddard, J.J., Hughes, D.C., & Pearl M. (1997). Children’s access to health care: the role of social and economic factors. In R. Stein (ed.), Health Care for Children: What’s Right, What’s Wrong, What’s Next. New York, NY: United Hospital Fund. Regalado, M., & Halfon N. (2001). Primary care services promoting optimal child development from birth to age 3 years. Review of the literature. Archives of Pediatrics and Adolescent Medicine, 155, 1311–1322. Selden, T.M., Banthin, J.S., & Cohen, J.W. (1998). Medicaid’s problem children: eligible but not enrolled. Results from the 1996 Medical Expenditure Panel Survey explain why the president’s priority to improve Medicaid outreach has come none too soon. Health Affairs, 17(3), 192–200. Shah, B.V., Barnwell, B.G., & Bieler, G.S. (1996). SUDAAN User’s Manual. Release 7.0. Research Triangle Park, N.C.: Research Triangle Institute.

11 Anticipatory Guidance: What Information Do Parents Receive? What Information Do They Want? Mark A. Schuster, Michael Regalado, Naihua Duan, and David J. Klein

Our [medical] curriculum covers a certain amount of study of the anatomy and physiology of the child about which mothers never ask us, but the information which they seek has to do with that which cannot be obtained from books, but rather is that sort of knowledge which has passed from mouth to mouth down through the centuries. Instead of asking mother or grandmother what should be done, the doctor is consulted. If her confidence is to be retained, the physician must be as familiar with the proper manner of bathing a baby as he is with the treatment of pneumonia, and he may render the baby as notable a service in one instance as in the other. The doctor is taking the place more and more of the ‘advice offering neighbor,’ and it behooves him to be able to advise the mother correctly. B. R. Hoobler, “The Desirability of Teaching Students Details Concerning the Care of the Normal Infant,” Transactions of the Association of American Teachers of Diseases of Children, 1917.

Physicians who take care of children are often called on to provide education and counseling about child rearing. Their advice may be especially important for the parents of young children, who grow and develop rapidly and sometimes seem to change overnight. Although physicians have been playing this role for at least the last century, the medical profession has given it greater attention in recent years in response to several factors: technological advances in medicine, a growing understanding of neuroscience and child development, and evolving systems of health care delivery. The nature of parenting has also changed, and parents in turn have placed new demands on their children’s physicians. Knowledge about child development has grown rapidly over the past several decades. Reports in the popular media, along with the Carnegie Corporation’s influential Starting Points report (1994), have brought greater attention to the importance of early childhood, the needs of parents, and the role of the health care system in meeting those needs. Even so, concerns 320

Anticipatory Parental Guidance

321

have been raised about the ability of the health care system to provide counseling and address problems, given current time and financial constraints. In this chapter, we examine the role of the physician as child-rearing counselor. We start with a brief review of the historical evolution of that role and a summary of more recent research. We then analyze data from the Commonwealth Fund Survey of Parents with Young Children. The chapter concludes with a discussion of implications for health care practice and physician training. Like much of the relevant literature, our discussion focuses on pediatricians, who care for the vast majority of children ages 0–3 (Hickson et al. 1988), yet the lessons are relevant to family practitioners and other clinicians who take care of families with young children.

the evolution of the pediatrician as educator and counselor The Child Welfare Movement and the Rise of Pediatrics In the early 1900s, child mortality approached 15 percent (Cone 1979). Diarrhea was responsible for many of those deaths, and contaminated milk was thought to be a primary cause of the problem. Therefore, milk stations were established to dispense sterilized milk to the poor (Cone 1979; Meckel 1990; Halpern 1988). Public health departments found, however, that making safe milk available was insufficient; families also needed education and guidance about proper hygiene and safe health habits. Milk stations therefore evolved into infant welfare clinics, where multidisciplinary teams of nurses, social workers, and physicians worked together to provide comprehensive advice about hygiene and feeding techniques (Meckel 1990; Halpern 1988; Parmelee 1995). During the 1910s, the child welfare movement in the United States shifted its focus to child growth and development, prompted in large part by the emerging field of child psychology (Halpern 1988). A principal goal of the child welfare movement was to educate parents about normal development and child-rearing techniques, and educational activities became the central feature of services provided by infant welfare clinics. Physicians at those clinics learned to address aspects of infant care not typically within the purview of general practitioners of the time. Middle-class mothers were eventually attracted to the clinics by the opportunity to get advice on infant care (Halpern 1988; Parmelee 1995). As the demand for such advice grew, the field of primary care pediatrics expanded beyond the infant welfare clinics. Thus, at its inception, the field of pediatrics emphasized the general health of the child, prevention of illness, promotion of normal development, and concern about the impact of the child’s social context (Halpern 1988).

322

Schuster, Regalado, Duan, Klein

Between the two world wars, pediatric practice evolved rapidly. The profession standardized the well-baby visit, which emphasized advice to parents on prevention issues and child development (Halpern 1988). Pediatricians looked to the growing body of knowledge from child development research institutes that had been established across the country. Arnold Gesell, a psychologist and pediatrician who devoted his career to the scientific study of normal infants and preschool children, was particularly influential in calling attention to the behavioral aspects of child health supervision (Ames 1989). After World War II, subspecialization emerged within pediatrics in response to technological advances and the evolution of academic departments at schools of medicine. Pediatric training in academic centers began to place greater emphasis on the study of specific diseases, and a growing disparity emerged between hospital-based pediatric training, which tended to be disease-oriented, and office-based general practice, which maintained its focus on prevention and the needs of well children. The final decades of the twentieth century saw a resurgence of the pediatric generalist in academic settings (Halpern 1988) and a renewed interest in addressing psychosocial issues (Richmond 1985).

Anticipatory Guidance and the Child-Rearing Concerns of Parents Services addressing the preventive and psychosocial aspects of health are important components of comprehensive care (Haggerty 1971; Breslau et al. 1975). To achieve the goal of providing such care, the profession has routinized the provision of advice about the broader aspects of health and development into what has been called “anticipatory guidance.” Following a set of age-appropriate preventive counseling topics routinely discussed at well-child care visits, the pediatrician tries to prepare parents for the child’s changing needs by supplying appropriate information before the child reaches specific developmental milestones. For example, physicians are expected to begin discussing toilet training before the child is developmentally ready so that parents have a strategy and reasonable expectations when the time comes. The American Academy of Pediatrics (AAP) emphasizes the role of the pediatrician in the prevention and early detection of various behavioral, developmental, and social functioning problems (AAP 1993). Both the AAP and the Maternal and Child Health Bureau (MCHB) have published manuals that list age-related topics for pediatricians and other clinicians to discuss with parents as part of well-child care. Over recent years, the lists have grown substantially longer. The third edition of the AAP’s Guidelines for Health Supervision, published in 1997, devotes several times more space to anticipatory guidance than does the second edition, published in 1988. MCHB’s Bright Futures (Green 1994) presents a comprehensive treatment

Anticipatory Parental Guidance

323

of anticipatory guidance appropriate for different ages. It is unlikely that these organizations expect clinicians to address all components of anticipatory guidance with every patient, yet the message is clear that physicians should be knowledgeable about all recommended topics and able to tailor their advice to a family’s particular needs. Studies assessing what parents expect from pediatric health supervision confirm the primacy of the physician’s biomedical role in ensuring children’s physical health (Breslau et al. 1975; Blum 1950; Stine 1962; Deisher et al. 1965; Korsch et al. 1971; Hickson et al. 1983; McCune et al. 1984). Parents’ expectations for how physicians should address psychosocial issues, however, are less clear. Over the last few decades, several studies have suggested that most parents identify the pediatrician as an important source of child-rearing information (Deisher et al. 1965; Hickson et al. 1983; Cone 1979; McCune et al. 1984; Cheng et al. 1996b), although they do not typically rank psychosocial concerns as their highest priority for the health supervision visit. Even parents who consider psychosocial concerns to be important may view the pediatrician primarily as a biomedical professional, and so not seek advice about psychosocial matters. In a study exploring reasons why parents are unlikely to bring up psychosocial concerns with their doctors, Hickson and colleagues (1983) found that 39 percent were not aware that pediatricians could help, 16 percent thought the pediatrician was too busy, 15 percent felt they did not need help, 12 percent thought the pediatrician was not qualified to help, 10 percent thought the pediatrician was not willing to help, and 8 percent were too embarrassed to ask for help. On the other hand, most pediatricians in the study believed that parents would bring psychosocial concerns to their attention if those concerns were important. Physician interest in addressing psychosocial concerns was the strongest predictor of communication about such a concern during a health visit, suggesting that both parents’ and physicians’ interests and expectations may influence whether nonmedical issues are discussed during well-child visits. Although pediatricians report that they generally engage in many aspects of anticipatory guidance (Stickler and Simmons 1995; Cheng et al. 1999), prior studies provide only limited information about what components are being provided to parents and whether parents feel they have enough information to help them with common child-rearing concerns. The few available studies suggest large variations in the delivery of anticipatory guidance, depending on the population and clinical setting. The quality of documentation and the research methodology may also account for much of the apparent variation. One study found that 39 percent of children in rural North Carolina receiving well-child care visits through the Medicaid-affiliated Early and Periodic Screening, Diagnosis, and Treatment Program had no documentation of having received anticipatory guidance (Richardson et al. 1994). By contrast, a study of a Denver practice that used age-specific anticipatory guidance checklists found 99 percent of parents had received anticipatory guidance

324

Schuster, Regalado, Duan, Klein

on safety, nutrition, and family concerns (Brown et al. 1993). A study of parents in a Massachusetts health maintenance organization found a wide range of rates of discussion, from a low of 6 percent for gunrelated issues to a high of 98 percent for both growth/nutrition and language development (Cheng et al. 1996a). The limitations of the current literature make it difficult to determine how adequately child health care providers are able to meet parents’ needs for child-rearing information. Concern about the effectiveness of comprehensive pediatric care has been growing for some time (Haggerty 1971; Lewis 1971), particularly as market pressures have emerged to place constraints on both time and money. A report published by the Office of Technology Assessment (1988) concluded that there is little evidence showing whether or not well-child care has an influence on developmental and social outcomes. The report cited only one evaluation, which showed a positive relationship between developmental outcome and extent of physician teaching with parents (Chamberlin and Szumowski 1980). Similarly, it cited very little evidence to support the use of developmental screening during preschool years. Using a different approach, Regalado and Halfon (2001) examined evidence regarding the effectiveness of developmental and behavioral services in pediatric care of infants and toddlers. The AAP and Bright Futures guidelines for anticipatory guidance were used to construct a typology and a comprehensive list of anticipatory guidance activities. Regalado and Halfon found that most studies are limited to evaluations of efficacy; the evidence base remains sparse for how well pediatricians can implement anticipatory guidance activities.

analysis of the commonwealth survey The Commonwealth Fund Survey of Parents with Young Children sheds light on parents’ views about how their needs for child-reading guidance are currently being met. The survey data indicate whether parents of infants and toddlers have discussed each of six recommended anticipatory guidance topics with a clinician, for which topics parents feel they could use more information, what other useful services they have received, and how receipt of anticipatory guidance relates to their impressions of the quality of service they receive.

Outcome Variables The Commonwealth Survey asked parents if they had ever personally discussed a list of specific topics with a health professional. The topics,

Anticipatory Parental Guidance

325

recommended in both AAP and Bright Futures guidelines, were how to care for a newborn (asked only of parents with children under 3 months of age), how to deal with the child’s sleeping patterns, what to do when the child cries, how to help and encourage the child to learn, how to discipline the child (asked of parents with children ages 6–36 months), and how and when to toilet train the child (asked of parents with children ages 18–36 months) (AAP 1997, 1988; Green 1994). The survey went on to ask whether or not parents felt they could use more information on each topic. Among those who answered affirmatively, some had already discussed the topic with a health professional, while others had never discussed the topic. For this analysis, we have classified the latter group of parents – those who felt they could use anticipatory guidance information but had not received it at all – as an “unaddressed need group” for the topic. We believe that these parents could be especially receptive to guidance from their pediatric providers. Parents were also asked about receipt and usefulness of several special services: a packet of information about how to care for a newborn, a home visit by a nurse or other health care professional to help them learn how to care for a newborn, a special telephone line staffed by nurses and doctors 24 hours a day, a service to remind parents by telephone or mail to bring the child in for a check-up or immunization, and a booklet to keep track of the child’s health information. Respondents were asked, as well, if they would be willing to pay an additional $10 a month to receive anticipatory guidance and other additional services, along with several questions about sources of child-rearing information. Each parent was asked to rate the health care its child received as poor, fair, good, or excellent (this skewed distribution is typical of such items, as found in the 1994 National Health Interview Survey). Few respondents rated clinicians as less than good; we therefore further divided their responses into two groups: excellent and not excellent. We also derived a dichotomous summary rating, set to 1 if the respondent gave at least one nonexcellent rating and 0 if the respondent gave only excellent ratings. Age-Normed Percentiles Depending on the child’s age, discussion of an anticipatory guidance topic varies in content (for example, encouraging learning is different for parents of a 3-month-old infant and a 2-year-old toddler) and importance (discipline, for example, may be a more compelling concern for parents of older children). Some topics apply only to certain age groups: of the six topics addressed in the survey, only four applied to children less than 3 months old, three to children ages 3–5 months, four to children ages 6–17 months, and five to children ages 18–36 months. The distribution of responses differs substantially across topics (Table 11.1). Therefore, summary measures (such as the proportion of topics not discussed or the percentage

Schuster, Regalado, Duan, Klein

326

table 11.1. Percentage of Parents Who Had Not Discussed Each Topic, Who Could Use More Information, and Who Were in “Unaddressed Need” Groupsa

Discussion topica Care for newborn Crying Sleeping patterns Encouraging learning Discipline Toilet training

Percentage who could use more information

Percentage of all parents who had not discussed topic

Among parents who had discussed topic

38

18

22

9

65 59

22 33

22 28∗

15 17

77

53

55

42

75 66

52 39

41∗∗ 37

30 24

Among parents Percentage of all who had not parents in discussed “unaddressed topic need” groupb

a Care

for newborn includes only parents whose index child was younger than 3 months old (n = 170); discipline includes parents of children ages 6–36 months (n = 1,645); toilet training includes parents of children ages 18–36 months (n = 1,011). The n for the rest of the items is 2,009–2,012. b“Unaddressed need” group includes parents who reported both that they did not discuss the topic and that they could use more information on the topic. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001 for t test for comparison of columns 2 and 3.

of parents in unaddressed need groups) cannot be compared across age groups. To provide a common metric across age groups, we normed the summary measures by age group using percentiles similar to those used in standardized growth curves (Barone 1996). Children were divided into four age groups (less than 3 months, 3–5 months, 6–17 months, and 18–36 months), and the summary measures were converted into percentiles using the distribution in each age group. As with percentiles in standardized growth curves, age-normed percentiles for topics not discussed have the same meaning across age groups and measure how well each child’s status compares with that of other children in the same age group. The median number of missed topics for each age group was set at the 50th percentile. A percentile higher than 50 indicates that parents missed more topics than parents of children in the same age group. Such an interpretation is not available in unnormed measures. Therefore, when we report that the

Child’s characteristics Gender Male (1036) Female (981) Firstborn Yes (812) No (1146) Insurance Employer (1236) Government (495) Other (146) Uninsured (127) Usual provider of care Private physician’s office (1369) Health maintenance organization (148) Community clinic (227) Emergency room/other (263) Usually sees same clinician Yes (1615) No (276)

Characteristics (n)

50 50 51 49 48 51 51 59∗∗∗ 49 48 55 51∗ 49 55∗∗

48 52∗ 47 53 54 63∗∗∗ 47 51 59 58∗∗∗ 48 56∗∗∗

(continued)

Number of “unaddressed need” groups (percentile)

49 51

Number of topics not discussed (percentile)

table 11.2. Parents’ Mean Percentile Score for Number of Topics They Did Not Discuss and for Number of “Unaddressed Need” Groups They Are In

Parent’s (respondent’s) characteristics Gender Male (881) Female (1136) Age < 25 (468) 25–29 (365) 30–34 (539) 35 + (432) Education College graduate (521) Some college (549) High school graduate (665) Did not complete high school (276) Employment (by one or two adult households) One adult: full time (56) One adult: part time (22) One adult: unemployed (105) Two adult: full time/full time (586) Two adult: full time/part time (348) Two adult: full time/unemployed (689) Two adult: part time/part time or unemployed (42) Two adult: unemployed / unemployed (51)

Characteristics (n)

table 11.2. (continued)

49 51 51 52 50 48 46 49 52 53∗∗ 54 49 53 49 50 50 57 53

51 51 48 49 45 44 55 59∗∗ 51 52 56 49 47 49 54 68∗∗∗

(continued)

Number of “unaddressed need” groups (percentile)

52 48∗

Number of topics not discussed (percentile)

< .01. ∗∗∗ p < .001.

∗∗ p

∗ Percentiles

52 59 48 50∗∗∗ 56 53 49 45∗∗∗ 53 50 50 49 50

59 53 48 43∗∗∗ 54 50∗∗ 46 52 51∗∗

Number of “unaddressed need” groups (percentile)

54 59 48 51∗∗∗

Number of topics not discussed (percentile)

differ across the subgroups at p < .05 using an analysis of variance test.

Race/Ethnicity African American (213) Hispanic (254) White (1458) Other (82) Household income ($1,000s) < 10 (198) 10–30 (556) 30–60 (685) > 60 (426) Number of adults in household 1 (183) 2 or more (1834) Provision of daily care for child (in two-parent homes) Respondent provides more care (630) Daily care shared equally (591) Spouse/partner provides more daily care (498)

Characteristics (n)

table 11.2. (continued)

330

Schuster, Regalado, Duan, Klein

uninsured were at the 63rd percentile for the number of missed topics, for example, as we do in Table 11.2, we are indicating that a typical parent whose child was uninsured had missed more topics than 63 percent of parents in the overall population, controlling for age. This finding does not mean that the uninsured missed 63 percent of the topics or that 63 percent of the uninsured had missed the topics.

Independent Variables Independent variables appear in Table 11.2. In addition to standard demographic variables (child’s gender; parent’s gender, age, education, and race/ethnicity; and household’s number of adults, employment, and income), we included whether the child was firstborn as an indicator of the parent’s potential need for information (the expectation being that parents of firstborn children would have a greater need for anticipatory guidance). We also included a variable indicating whether respondents in two-parent households provided more or less of the daily care of the child than their spouse or partner. We believe respondents who provide more of the care would typically (although not always) be more likely to see the clinician and thus have the opportunity to receive anticipatory guidance; therefore, we repeated some analyses with the subsample of parents who provided more of the care.

Analyses We conducted univariate, bivariate, and multivariate analyses. Weighted analyses were used to account for unequal sampling probabilities to represent the overall distribution of parents with children younger than 3 years of age in the United States. For all statistical inference, standard errors and test statistics were adjusted for the design effect due to unequal sampling probabilities and stratification, using the sandwich variance estimator (Stata Corporation 1997). For bivariate analysis, we specified independent variables as dichotomous or categorical. For continuous outcome variables, we tested for the association using the one-way analysis of variance F test. For dichotomous outcomes, we used a chi-squared test. A multiple linear (logistic) regression was performed to examine the relationship between each continuous (dichotomous) outcome variable and the independent variables from Table 11.2, although we excluded sharing child care responsibilities (because it does not apply to single-parent households) and employment (because of colinearity with number of adults in the household). Income and parental age were entered as continuous variables.

Anticipatory Parental Guidance

331

findings Aspects of Child Rearing Discussed by Parents and Clinicians The six anticipatory guidance or child-rearing topics addressed in the survey were selected to cover issues considered important for healthy child growth and development and of particular interest to parents during a child’s first three years of life. Caring for a newborn, sleep patterns, crying, learning, discipline, and toilet training are topics recommended by AAP’s Guidelines for Health Supervision (1997, 1988) and MCHB’s Bright Futures (Green 1994). In Table 11.1, column 1 shows the percentage of parents reporting that they had never discussed these topics with their children’s health care provider. For five of the six topics (the exception being newborn care), over half the parents reported that they had never discussed the topic. Thirty-six percent of respondents had discussed no topics with a clinician. When this set of analyses was repeated only with respondents who provided more of the daily care for the child (and were thus more likely to have contact with the clinician), a higher percentage of parents reported discussions of some topics, but the overall findings remained the same: for all topics except newborn care, over half of parents still reported that they had never had a discussion with a clinician. Columns 2 and 3 in Table 11.1 divide the sample into parents who had and had not discussed each topic, showing the percentage of each group who could use more information on each topic. For two topics (sleeping patterns and discipline), parents who had discussions were significantly more likely than other parents to report that they could use more information. For the other topics, there was no significant difference. When these analyses were limited to parents who provide more of the daily care for the child, the changes were small. Column 4 shows the percentage of all parents who are in unaddressed need groups. These are parents who had not discussed the topic with a clinician but who said they could use more information about the topic. They may have the greatest unmet need and most receptivity to discussions. The percentage of parents in unaddressed need groups varied from a low of 9 percent for newborn care to a high of 42 percent for encouraging learning. When the sample was limited to parents who provide more of the daily care for the child, the changes in percentages were small.

Factors Influencing Receipt of Anticipatory Guidance We compared rates of discussion by demographic and other characteristics. To adjust for age-related differences in content and importance of some of the anticipatory guidance topics, we used age-normed percentiles as the

332

Schuster, Regalado, Duan, Klein

outcome variables for the analysis of variance (Table 11.2 ). Column 1 in table 11.2 shows percentiles for the number of missed discussion topics. Parents of children with no insurance were much more likely to have missed discussions than parents whose children had various forms of insurance. As shown in column 2, the same pattern was found in the number of topics for which parents were in an unaddressed need group. We also found, notably, that Hispanic parents reported the most missed discussions and were in the most unaddressed need groups, perhaps owing in part to language and cultural barriers. Higher income was associated with fewer missed discussions and fewer unaddressed need groups. Respondents who provide more of the daily child care had fewer missed discussions. Parents of children who usually see the same clinician also had fewer missed discussions and were in fewer unaddressed need groups, a finding that speaks to the value of continuity of health care in facilitating productive discussions about child rearing. Knowing the child’s physician seems to make it easier to discuss sensitive topics and allows the physician the benefit of accumulated knowledge of the child and family over time. Seeing the same clinician may also serve as a proxy for receiving health care in full service settings, as opposed to emergency rooms and other settings designed to provide episodic care. A multiple linear regression predicting the percentiles of missed discussions showed fewer significant predictors than the bivariate analyses. Groups or characteristics that showed significant positive associations with missed discussions included uninsured children compared with children with employer-based insurance (p = .002), male respondents (p = .024), Hispanic respondents compared with white respondents (p = .035), and care received at a community clinic compared with a private office (p = .012). Significant negative associations were found for children who usually see the same clinician (p = .023), college graduates (p = .023), and parents with some college education compared with high school graduates (p < .001). A multiple linear regression was also conducted for the percentiles of unaddressed need groups. Hispanic parents had significant positive associations with more unaddressed need groups than did white parents (p = .001). A significant negative association was found for household income (p < .01) and forchildrenwhousuallyseethe same clinician (p = .01). Other Sources of Information about Child Rearing Clinicians are not the only sources of information for parents. Seventy-four percent of parents reported that they use “books, magazines, television, or videos to get information about how to raise” their children. Thirty-five percent had attended a class or discussion group about child rearing or parenting. When asked to list up to two people with whom they are most comfortable discussing child rearing, 52 percent named their spouse or partner

Anticipatory Parental Guidance

333

and 46 percent named their mother or mother-in-law. The next highest response was a doctor, nurse, or other health professional, at 15 percent. Given that respondents were invited to give only two responses, we do not know how many would have mentioned clinicians after listing key family members or friends.

Other Services Offered by Clinicians The Commonwealth Survey also asked whether parents had received special services from their clinicians that supplement and enhance office-based health care. Only 10 percent had not received a packet of information about how to care for a newborn, and 33 percent had not received a booklet to keep track of their children’s health information. Over half of parents had not received a home visit or appointment reminder, and over half did not have a clinician with a 24-hour information line (Table 11.3). Almost all parents who received one of these services found it useful, and most who had not received a service expected that they would find it useful. For four of the five services, parents who had received the service were more likely to consider it useful than parents who had not received the service. This finding could indicate that those who want such services seek out clinicians who provide them, or that the value of services becomes more apparent when they are made available. There is large variation in the percentage of people in unaddressed need groups for these services. table 11.3. Receipt of Services and Usefulness of Services (%) Percentage who found service usefula

Topic Information packet Home visit 24 hour phone line Appointment Reminder Child health book a Useful

Percentage who did not receive service

Among parents received service

Among parents who did not receive service

Percentage of all parents in “unaddressed need” groupb

10 80 68 58

55 67 88 85

77∗∗∗ 58∗∗ 74∗∗∗ 57∗∗∗

8 47 50 33

33

84

68∗∗∗

23

indicates that the parent responded “very” or “somewhat” useful. not discuss topic and could use more information. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001 for t test for comparison of columns 2 and 3. b Did

334

Schuster, Regalado, Duan, Klein

table 11.4. Ratings of Clinicians among Parents Whose Child Usually Sees the Same Health Professional How would you rate your child’s regular doctor or nurse on . . . Providing good health care? Giving you guidance on how to care for your child? Helping you to understand how your child is growing and developing? Listening to you carefully and answering your questions?

Percentage of Respondents Excellent

Good

Fair / poor

69 56

28 37

3 7

58

35

6

66

29

5

Parents’ Ratings of the Quality of Service Delivered For each aspect of care rated (using a four-point scale from poor to excellent), more than half the respondents rated their clinician as excellent and over 90 percent rated their clinician excellent or good (Table 11.4). Fifty-seven percent of respondents gave a rating less than excellent for at least one aspect of care. There were significant variations in ratings for only a few characteristics. Male respondents were more likely than females to give at least one nonexcellent rating (60 percent vs. 54 percent, p < .05). There were also significant differences by race or ethnicity, with Hispanic (68 percent) and other (69 percent) respondents more likely than African American (59 percent) or white (54 percent) respondents to give at least one nonexcellent rating (p < .001). The likelihood of parents’ giving a less than excellent rating varied by regular source of care: private physician’s office (54 percent), health maintenance organization (59 percent), emergency room/other (62 percent), and community health center (71 percent) (p < .001). A logistic regression showed that Hispanic parents were more likely than white parents to give at least one nonexcellent rating (p < .05), as were male respondents (p < .05) and people whose children usually received care at a community health center compared with a private physician’s office (p < .05). Parents who had discussed more topics with a clinician reported receiving excellent care for more topics. Parents who were in fewer unaddressed need groups were also likely to report receiving excellent care for more topics. The pattern was not as strong for receipt of services, but there was a significant relationship between those who had received more services and better ratings (Table 11.5). Parents’ Willingness to Pay Extra Sixty-four percent of respondents would be willing to pay an extra $10 per month to discuss the child-rearing topics and receive the services listed in

Anticipatory Parental Guidance

335

table 11.5. Percentage of Parents Who Gave at Least One Nonexcellent Rating and Percentage Willing to Pay an Extra $10 per Month for Anticipatory Guidance Discussions and Special Services Gave one or more nonexcellent rating Total sample discussion topicsa Number not discussed Low percentile Medium percentile High percentile Number not discussed and could use more informationb Low percentile Medium percentile High percentile services Number not received Low percentile Medium percentile High percentile Number not received and perceived very useful Low percentile Medium percentile High percentile

Willing to pay extra $10 per month

57

64

46 59 64∗∗∗

67 63 62

48 55 70∗∗∗

55 66 73∗∗∗

52 56 65∗

65 62 64

52 57 61

49 65 77∗∗∗

a Cutoffs for percentiles are less than 34th percentile (low), less than 67th percentile (medium),

and 67th percentile or higher (high). need” group. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001 using chi-squared test. b “Unaddressed

Tables 11.1 and 11.3. Parents who were in more unaddressed need groups for discussion topics or who had not received special services that they thought would be useful were much more likely to be willing to pay for them (Table 11.5). However, caution is necessary in interpreting this finding. A more comprehensive survey would be needed to determine the actual amount people would pay and how they would be willing to pay for it – as an additional fee for an add-on option within a particular health plan, for example, or as part of a higher fee for a plan providing a higher level of service. Nonetheless, the findings on willingness to pay lend further credence to the notion that parents would like physicians to provide anticipatory guidance and additional services.

336

Schuster, Regalado, Duan, Klein

discussion Physicians can play an important role as child-rearing advisors for parents of infants and toddlers. The importance of this aspect of the doctorparent relationship is underscored by evidence that the first three years of life are crucial to the developing brain and the social and emotional development of children (Carnegie Corporation 1994; Karoly et al. 1998). The Commonwealth Survey allows us to draw several conclusions about parents’ receipt of child-rearing guidance from their children’s pediatric providers, and their potential receptivity to such guidance. Most Parents Are Not Discussing Child Rearing with Their Doctors Because of their regular contact with parents, physicians have a unique opportunity to help parents manage typical developmental challenges and optimize children’s learning and social experiences during the first three years of life. Yet for each age-appropriate child-rearing topic except newborn care, over half the parents reported that they had never discussed the topic with a clinician, and about one-third had not discussed any of the topics. These findings appear to indicate a major unmet need for parents. Families Most Likely to Benefit from Discussions about Child Rearing Were Less Likely to Have Those Discussions at Routine Health Visits Lack of insurance and continuity of care, receiving care at a community clinic, and low education were associated with a lower probability of discussing child-rearing topics at health visits. These factors are associated with lower socioeconomic status and access barriers to health care in general (St. Peter et al. 1992; Wood et al. 1990). African American and Hispanic children have been found to have more limited access to health care than white children (Newacheck et al 1996). Our study also found that Hispanic respondents were less likely to report discussions about child rearing at health visits. This may reflect language or other barriers to communication. Regardless, families with limited access to health care, child-rearing information, and parenting supports, and families with greater parenting challenges, are less likely to receive developmental services. Parents Value Child-Rearing Information from Their Physician Other studies show that parents value the receipt of child-rearing information from pediatricians (Deisher et al. 1965; Hickson et al. 1983; McCune et al. 1984; Cheng et al. 1999). In the Commonwealth Study, 64 percent of

Anticipatory Parental Guidance

337

parents reported that they would be willing to pay an extra $10 per month to discuss child-rearing topics and receive special services. Parents who had discussed child-rearing topics were more likely to give an excellent rating to their clinicians. Parents Turn to Other Sources for Help with Child Rearing Parents often get child-rearing information from other sources. Seventyfour percent of parents reported that they use “books, magazines, television, or videos to get information about how to raise” their children. Thirty-five percent had attended a class or discussion group about child rearing or parenting. Just over half reported that they were most comfortable discussing child rearing with their spouse or partner, while just under half named their mother or mother-in-law. Health professionals were a distant third at 15 percent. The difference in findings between this study and others may in part reflect the use of the word “comfortable” in the Commonwealth Survey. Parents may feel more comfortable talking with family members, yet trust the information more when it comes from a physician. Parents Are Not Getting the Help They Need for Several Reasons Both physician and parent factors have been identified as contributing to a failure to discuss child rearing (Hickson et al. 1983). Parents’ expectation of receiving this service from doctors is low, even though they consider their physicians to be a resource for child-rearing information. Biomedical concerns still rank as the top priority for both parents and physicians. Likewise, parents are not encouraged to express psychosocial concerns (Hickson et al. 1983; Cheng et al. 1996). When they do express those concerns, they are frequently ignored or inadequately addressed (Korsch et al. 1971; Werner et al. 1979; Ferris et al. 1998). Another concern is that market forces may be limiting the delivery of anticipatory guidance services. Constraints on time and lack of reimbursement for anticipatory guidance services create a negative incentive for physicians (Hirsch 1995), although they may not be the primary factors affecting preventive counseling for psychosocial concerns (Cheng et al. 1999). The lack of an evidence base for these services creates a negative incentive for third-party payers. Also, physicians may be hesitant to screen for developmental and behavioral problems if they fear they lack the time or training to address them (Perrin 1998). Another factor is the limited training pediatricians receive in developmental and behavioral pediatrics. Clinicians may not know how to discuss many aspects of anticipatory guidance because of limited training, especially in child development (Cassel et al. 1998; Freed et al. 1995; Goldstein et al. 1987; Kenney et al. 1988; Adler and Korsch 1985).

338

Schuster, Regalado, Duan, Klein

Several Helpful Steps Have Already Been Taken The latest AAP and Bright Futures guidelines acknowledge the complexity of child development and emphasize the importance of contextual influences. The development of The Classification of Child and Adolescent Mental Diagnoses in Primary Care: Diagnostic and Statistical Manual for Primary Care provides a framework for categorizing the typical developmental issues that arise in routine well-child care (AAP 1996; Kelleher et al. 1996). Also, a typology for categorizing developmental and behavioral services in primary care has been proposed to organize approaches to evaluation and reimbursement for services (Regalado and Halfon 1999).

recommendations Based upon our review of the literature and analysis of the Commonwealth data, we make the following recommendations. Expand Evaluation of Anticipatory Guidance Services Market forces do not appear to encourage thorough delivery of anticipatory guidance and developmental services as part of well-child care. Evidencebased outcomes may increasingly determine what services are delivered, yet there has been little evaluation of the influence of anticipatory guidance services on parent/patient behavior and health. Indeed, findings have generally been mixed (Office of Technology Assessment 1988; Casey et al. 1979). There is some evidence that anticipatory guidance on injury prevention (Bass et al. 1993), violence reduction (Sege et al. 1997), and infant sleep patterns (Adair et al. 1992) can have a positive impact. A study of counseling in prenatal pediatric visits showed some impact on breastfeeding but not on other topics (Serwint et al. 1996). Other studies have also had mixed results (Chamberlin and Szumowski 1980; Casey and Whitt 1980; Chamberlain et al. 1979). One study that aimed to improve the impact of anticipatory guidance by presenting it in the context of the child’s developmental capacities showed no improvement over simple presentation of the anticipatory guidance material (Dworkin et al. 1987). Anticipatory guidance may be more effective when tailored to the educational level of the parent, and when advice is specific rather than general (Glascoe et al. 1998; Korsch 1984; Korsch and Negrete 1972). It also may have more impact when supplemented with additional materials, such as handouts, or educational modalities, such as videotapes (Glascoe et al. 1998). Further research is needed on how to provide anticipatory guidance in an effective manner. Research must also address how best to deliver anticipatory guidance. One approach that may be effective is prescreening parents about their

Anticipatory Parental Guidance

339

developmental concerns (Glascoe et al. 1998; Triggs and Perrin 1989). Further study of the process of communication during well-child visits is critical, particularly in today’s high-volume settings. Doctor-patient communication problems have been identified as an important area for improvement for several decades (Korsch 1971). Solutions, particularly communication approaches to help parents effect behavioral change, are needed to help physicians move from information resources to counselors. Make It Easier for Physicians to Provide Anticipatory Guidance We may also be able to improve delivery of anticipatory guidance by incorporating it into standardized medical record forms that prompt the clinician to offer age-appropriate advice (Schiffman et al. 1997). Incentives for clinicians and health care delivery organizations could encourage the incorporation of anticipatory guidance into quality assessment systems (Bauchner 1995; National Committee for Quality Assurance 1997; Schuster et al. 1997). Explore Alternative Models of Service Delivery There has been an ongoing debate over whether or not pediatricians should accept responsibility for helping parents with child-rearing issues. Limitations in pediatric training, time constraints, and barriers to reimbursement for services raise questions about what alternative approaches might work. Examples of alternative approaches to preventive and supportive care for parents of infants and toddlers include the Commonwealth Fund’s Healthy Steps for Young Children Program, Boston’s Pathways Program, and Zero to Three’s Developmental Specialist program (Eggbeer et al. 1997; KaplanSanoff et al. 1997; Zuckerman et al. 1997). These programs incorporate a specialist in child development into the routine of the pediatric office to provide help with parents’ questions and concerns. These approaches are reminiscent of the infant welfare clinics that employed social workers, nurses, and physicians to provide comprehensive teaching and support to mothers. Improve the Training of Physicians In preliminary releases of The Future of Pediatric Education II, the American Academy of Pediatrics (2000) has affirmed the emphasis of its 1978 task force on training in child development and behavioral pediatrics. Some progress has been made on the training front since the 1978 report was released. Medical schools have been adapting education programs to help students become more empathic and patient-focused (Moore et al. 1994; Wilkes et al. 1998). Surveys of recent residency graduates suggest that there has been improvement in developmental

340

Schuster, Regalado, Duan, Klein

and behavioral pediatrics training (Wender et al. 1992; Roberts et al. 1997). The Commonwealth Survey results suggest, however, that the goals of improved training from the parent’s perspective are still far off. The training of physicians who take care of children should incorporate an understanding of how to provide advice to parents. Physicians need to understand normal and abnormal development and child rearing, and they should understand the role that contextual influences play in child development. Parents bring varying degrees of experience and resources to the task of caring for and rearing their children and face different personal life issues in their own development as adults. The need for information, advice, support, and reassurance from physicians is of particular importance to families whose life circumstances limit their resources and present extraordinary challenges to caring for and raising children. These challenges include managing work and parenthood, social isolation, single parenthood, poverty, limited education, and becoming parents while still teenagers. Cultural factors, particularly those associated with ethnic minority parents, require not only an understanding of their role in shaping the parent’s perspective but also an ability to accommodate those issues into the clinical care and advice given to parents (Garcia-Coll and Meyer 2000).

summary Although the profession of pediatrics was established with the goal of optimizing parents’ preparation for and success with child rearing, the priorities of training and clinical practice have gradually drifted away from the psychosocial aspects of children’s health. Data from the Commonwealth Survey suggest that many parents of infants and toddlers are not receiving anticipatory guidance about common child-rearing concerns. We need to determine how best to deliver such guidance to parents, then assist health care professionals and the health care system in providing it.

references Adair, R., Zuckerman, B., Bauchner, H., Philipp, B., & Levenson, S. (1992). Reducing night waking in infancy: a primary care intervention. Pediatrics, 89, 585–8. Adler, R., & Korsch, B.M. (1985). Pediatric generalists in an academic setting. Pediatrics, 76, 307–8. American Academy of Pediatrics Committee on Psychological Aspects of Child and Family Health. (1988). Guidelines for Health Supervision II. Elk Grove, IL: American Academy of Pediatrics. American Academy of Pediatrics Committee on Psychological Aspects of Child and Family Health. (1997). Guidelines for Health Supervision III, Third Edition. Elk Grove, IL: American Academy of Pediatrics.

Anticipatory Parental Guidance

341

American Academy of Pediatrics Committee on Psychosocial Aspects of Child and Family Health. (1993). The pediatrician and the “new morbidity.” Pediatrics, 92, 731–733. American Academy of Pediatrics. (1996). The Classification of Child and Adolescent Mental Diagnoses in Primary Care: Diagnosis and Statistical Manual for Primary Care (DSM-PC). EIK Grove, IL: American Academy of Pediatrics. American Academy of Pediatrics. (2000). The Future of Pediatric Education II. Organizing pediatric education to meet the needs of infants, children, adolescents, and young adults in the 21st century. A collaborative project of the pediatric community. Pediatrics, 105, 157–212. Ames, L.B. (1989). Arnold Gesell: Themes of His Work. New York, NY: Human Sciences Press. Barone, M.A., & The Johns Hopkins Hospital. (1996). The Harriet Lane Handbook: A Manual for Pediatric House Officers. St. Louis, MO: Mosby-YearBook. Bass, J.L., Christoffel, K.K., Widome, M., Boyle, W., Scheidt, P., Stanwick, R., & Roberts, K. (1993). Childhood injury prevention counseling in primary care settings: a critical review of the literature. Pediatrics, 92, 544–50. Bauchner, H. (1995). The pediatric report card for preventive services. Pediatrics, 95, 930–4. Blum, L. (1950). Some psychological and educational aspects of pediatric practice: A study of well baby clinics. Genetic, Social, and General Psychology Monographs, 41, 3–97. Breslau, N., Haug, M.R., Burns, A.E., McClelland, C.Q., Reeb, K.G., & Staples, W.I. (1975). Comprehensive pediatric care: the patient viewpoint. Medical Care, 13, 562–9. Brown, J., Melinkovich, P., Gitterman, B., & Ricketts, S. (1993). Missed opportunities in preventive pediatric health care. Immunizations or well-child care visits? American Journal of Diseases of Children, 147, 1081–4. Carnegie Corporation of New York. (1994). Starting Points: Meeting the Needs of Our Youngest Children: The Report of the Carnegie Task Force on Meeting the Needs of Young Children. New York, NY: Carnegie Corporation. Casey, P., Sharp, M., & Loda, F. (1979). Child-health supervision for children under 2 years of age: A review of its content and effectiveness. Journal of Pediatrics, 95, 1–9. Casey, P.H., & Whitt, J.K. (1980). Effect of the pediatrician on the mother-infant relationship. Pediatrics, 65, 815–20. Cassel, C.K., Nelson, E.A., Smith, T.W., Schwab, C.W., Barlow, B., & Gary, N.E. (1998). Internists’ and surgeons’ attitudes toward guns and firearm injury prevention. Annals of Internal Medicine, 128, 224–30. Chamberlin, R.W., & Szumowski, E.K. (1980). A follow-up study of parent education in pediatric office practices: impact at age two and a half. American Journal of Public Health, 70, 1180–8. Chamberlin, R.W., Szumowski, E.K., & Zastowny, T.R. (1979). An evaluation of efforts to educate mothers about child development in pediatric office practices. American Journal of Public Health, 69, 875–86. Cheng, T.L., DeWitt, T.G., Savageau, J.A., & O’Connor, K.G. (1999). Determinants of counseling in primary care pediatric practice: Physician attitudes about time, money, and health issues. Archives of Pediatrics and Adolescent Medicine, 153, 629–35.

342

Schuster, Regalado, Duan, Klein

Cheng, T.L., Savageau, J.A., Bigelow, C., Charney, E., Kumar, S., & DeWitt, T.G. (1996a). Assessing mothers’ attitudes about the physician’s role in child health promotion. American Journal of Public Health, 86, 1809–12. Cheng, T.L., Savageau, J.A., DeWitt, T.G., Bigelow, C., & Charney, E. (1996b). Expectations, goals, and perceived effectiveness of child health supervision: A study of mothers in a pediatric practice. Clinical Pediatrics, 35, 129–37. Cone, T.E. (1979). History of American Pediatrics. Boston, MA: Little, Brown. Deisher, R., Engel, W., Spielholz, R., & Standfast, S. (1965). Mothers’ opinions of their pediatric care. Pediatrics, 35, 82–90. Dworkin, P.H., Allen, D., Geertsma, A., Solkoske, L., & Cullina, J. (1987). Does developmental content influence the effectiveness of anticipatory guidance? Pediatrics, 80, 196–202. Eggbeer, L., Littman, C.L., & Jones, M. (1997). Zero to Three’s Developmental Specialist Program in pediatric practice project: An important support for parents and young children. Zero to Three, 17, 3–8. Ferris, T.G., Saglam, D., Stafford, R.S., Causino, N., Starfield, B., Culpepper, L., & Blumenthal, D. (1998). Changes in the daily practice of primary care for children. Archives of Pediatrics and Adolescent Medicine, 152, 227–33. Freed, G.L., Clark, S.J., Lohr, J.A., & Sorenson, J.R. (1995). Pediatrician involvement in breast-feeding promotion: A national study of residents and practitioners. Pediatrics, 96, 490–4. Garcia-Coll, C., & Meyer, E. (2000). The sociocultural context of infant development. In: C. Zeanah (ed.), Handbook of Infant Mental Health. New York, NY: Guilford. Glascoe, F.P., Oberklaid, F., Dworkin, P.H., & Trimm, F. (1998). Brief approaches to educating patients and parents in primary care. Pediatrics, 101, E10. Goldstein, B., Fischer, P.M., Richards, J.W., Jr., Goldstein, A., & Shank, J.C. (1987). Smoking counseling practices of recently trained family physicians. Journal of Family Practice, 24, 195–7. Green, M. (ed.), (1994). Bright Futures: Guidelines for Health Supervision of Infants, Children, and Adolescents. Arlington, VA: National Center for Education in Maternal and Child Health. Haggerty, R. (1971). Does comprehensive care make a difference? Introduction. Historical perspectives. American Journal of Diseases of Children, 122, 467–8. Halpern, S.A. (1988). American Pediatrics: The Social Dynamics of Professionalism, 1880–1980. Berkeley, CA: University of California Press. Hickson, G.B., Altemeier, W.A., & O’Connor, S. (1983). Concerns of mothers seeking care in private pediatric offices: Opportunities for expanding services. Pediatrics, 72, 619–24. Hickson, G.B., Stewart, D.W., Altemeier, W.A., & Perrin, J.M. (1988). First step in obtaining child health care: selecting a physician. Pediatrics, 81, 333–8. Hirsch, A.T. (1995). The economic survival of pediatric practice. Pediatrics, 96, 825–829. Hoobler, B. (1917). The desirability of teaching students details concerning the care of the normal infant. Transactions of the Association of American Teachers of Diseases of Children, 11, 43. Kaplan-Sanoff, M., Brown, T.W., & Zuckerman, B.S. (1997). Enhancing pediatric primary care for low-income families: Cost lessons learned from Pediatric Pathways to Success. Zero to Three, 17, 34–36.

Anticipatory Parental Guidance

343

Karoly, L.A., Greenwood, P.W., Eveeringham, S.S., Hoube, J., Kilburn, M.R., Rydell, C.P., Sanders, M., & Chiesa, J. (1998). Investing In Our Children. Santa Monica, CA: RAND. Kelleher, K.J., & Wolraich, M.L. (1996). Diagnosing psychosocial problems. Pediatrics, 97, 899–901. Kenney, R.D., Lyles, M.F., Turner, R.C., White, S.T., Gonzalez, J.J, Irons, T.G., Sanchez, C.J., Rogers, C.S., Campbell, E.E., & Villagra, V.G. (1988). Smoking cessation counseling by resident physicians in internal medicine, family practice, and pediatrics. Archives of Internal Medicine, 148, 2469–73. Korsch, B.M., Negrete, F., Mercer, A.S., & Freemon, B. (1971). How comprehensive are well child visits? American Journal of Diseases of Children, 122, 483–8. Korsch, B.M., & Negrete, V.F. (1972). Doctor-patient communication. Scientific American, 227, 66–74. Korsch, B.M. (1984). What do patients and parents want to know? What do they need to know? Pediatrics, 74, 917–9. Lewis, C. (1971). What is the evidence? American Journal of Diseases of Children, 122, 469–74. McCune, Y.D., Richardson, M.M., & Powell, J.A. (1984). Psychosocial health issues in pediatric practices: Parents’ knowledge and concerns. Pediatrics, 74, 183–90. Meckel, R.A. (1990). “Save the babies”: American public health reform and the prevention of infant mortality, 1850–1929. The Henry E. Sigerist Series in the History of Medicine. Baltimore, MD: Johns Hopkins University Press. Moore, G.T., Block, S.D., Style, C.B., & Mitchell, R. (1994). The influence of the New Pathway curriculum on Harvard medical students. Academic Medicine, 69, 983–9. National Center for Health Statistics. (1995). Public Use Data Tape Documentation Part I, National Health Interview Survey, 1994 (Machine readable data file and documentation). Hyattsville, MD: National Center for Health Statistics. National Committee for Quality Assurance. (1997). HEDIS 3.0: Narrative: What’s In It and Why It Matters. Washington, DC: National Committee for Quality Assurance. Newacheck, P.W., Hughes, D.C., & Stoddard, J.J. (1996). Children’s access to primary care: Differences by race, income, and insurance status. Pediatrics, 97, 26–32. Office of Technology Assessment. (1988). Healthy Children: Investing in the Future. Washington, DC: US Congress. Parmelee, A.H. (1995). The early history of pediatrics. SRCD Newsletter, 5, 8. Perrin, E.C. (1998). Ethical questions about screening. Journal of Developmental & Behavioral Pediatrics, 19, 350–2. Regalado, M., & Halfon, N. (2001). Primary Care Services Promoting Optimal Child Development From Birth to Three Years: A Review of the Literature. Archives of Pediatrics and Adolescent Medicine, 155, 1311–1322. Richardson, L.A., Selby-Harrington, M.L., Krowchuk, H.V., Cross, A.W., & Williams, D. (1994).Comprehensiveness of well child checkups for children receiving Medicaid: A pilot study. Journal of Pediatric Health Care, 8, 212–20. Richmond, J. B. (1985). Coming of age: Developmental pediatrics in the late twentieth century. Journal of Developmental and Behavioral Pediatrics, 6, 181–7. Roberts, K.B., Starr, S., & DeWitt, T.G. (1997). The University of Massachusetts Medical Center office-based continuity experience: Are we preparing pediatrics residents for primary care practice? Pediatrics, 100, E2.

344

Schuster, Regalado, Duan, Klein

Schuster, M.A., Asch, S.M., McGlynn, E.A., Kerr, E.A., Hardy, A.M., & Gifford, D.S. (1997). Development of a quality of care measurement system for children and adolescents. Methodological considerations and comparisons with a system for adult women. Archives of Pediatrics and Adolescent Medicine, 151, 1085–92. Sege, R.D., Perry, C., Stigol, L., Cohen, L., Griffith, J., Cohn, M., & Spivak, H. (1997). Short-term effectiveness of anticipatory guidance to reduce early childhood risks for subsequent violence. Archives of Pediatrics and Adolescent Medicine, 151, 392–7. Serwint, J.R., Wilson, M.E., Vogelhut, J.W., Repke, J.T., & Seidel, H.M. (1996). A randomized controlled trial of prenatal pediatric visits for urban, low-income families. Pediatrics, 98, 1069–75. Shiffman, R.N., Brandt, C.A., & Freeman, B.G. (1997). Transition to a computerbased record using scannable, structured encounter forms. Archives of Pediatrics and Adolescent Medicine, 151, 1247–53. St. Peter, R.F., Newacheck, P.W., & Halfon, N. (1992). Access to care for poor children. Separate and unequal? JAMA, 267, 2760–4. Stata Corporation. 1997. Stata Statistical Software: Release 5.0. College Station, TX: Stata Corporation. Stickler, G.B., & Simmons, P.S. (1995). Pediatricians’ preferences for anticipatory guidance topics compared with parental anxieties. Clinical Pediatrics, 34, 384–7. Stine, O. (1962). Content and method of health supervision by physicians in child health conferences in Baltimore, 1959. American Journal of Public Health, 52, 1858– 1865. Triggs, E.G., & Perrin, E.C. (1989). Listening carefully. Improving communication about behavior and development. Recognizing parental concerns. Clinical Pediatrics, 28, 185–92. Wender, E., Bijur, P., & Boyce, W. (1992). Pediatric residency training: Ten years after the task force report. Pediatrics, 90, 876–80. Werner, E., Adler, E., Robinson, R., & Korsch, B. (1979). Attitudes and interpersonal skills during pediatric internship. Pediatrics, 63, 491–9. Wilkes, M.S., Usatine, R., Slavin, S., & Hoffman, J.R. (1998). Doctoring: University of California, Los Angeles. Academic Medicine, 73, 32–40. Wood, D.L., Hayward, R.A., Corey, C.R., Freeman, H.E., & Shapiro, M.F. (1990). Access to medical care for children and adolescents in the United States. Pediatrics, 86, 666–73. Zuckerman, B., Kaplan-Sanoff, M., Parker, S., & Young, K.T. (1997). The Healthy Steps for Young Children Program. Zero to Three, 17, 20–25.

This chapter was adapted in part from Schuster M.A., Duan, N., Regalado, M., & Klein, D.J. (2000). Anticipatory guidance: What information do parents receive? What information do they want? Archives of Pediatrics and Adolescent Medicine, c 2000 American Medical Association. December 2000:154, 1191–98. Copyright


part

IV

FUTURE DIRECTIONS AND POLICY IMPLICATIONS

12 New Models of Pediatric Care Barry Zuckerman and Steven Parker

The landscape for families is changing at a rapid pace. Managed care has left its stamp, perhaps indelibly, on the way pediatrics is practiced. Income inequality is increasing, and racial disparity in health remains. The media has saturated families with more information than they can process and spawned a culture for children that seems to change almost daily. At the same time, children are exposed at earlier and earlier ages to ideas and behaviors most of us have difficulty making sense of, even as adults. The world facing families with young children offers great promise, but also significant challenges. The dizzying pace of change may affect the content and duration of pediatric primary care visits, but the structure of those visits will probably change very little. The pediatric clinician will likely continue to see a child on at least eleven occasions for well-child care in the first 3 years of life and at least several other times for sick visits. Pediatric clinicians will almost surely remain the professionals who see families with young children most consistently. Indeed, for many families, pediatricians are the only professionals who see them together, in a family context. These visits come at a time when parents are extremely receptive to professional support, information, andadvice. Findings fromTheCommonwealthFund Survey of Parents with Young Children indicate, for example, that mothers are more likely to breastfeed if a physician discusses the advantages of the practice with her. In the past, the pediatric clinician took a narrow view of child health and focused largely on accident prevention, infectious diseases, and general health maintenance. This has changed somewhat in the past two decades as “the new morbidity,” or psychosocial issues, has entered the pediatric clinician’s job description. Over the next decade, still newer concerns – how to promote school success, resist the negative influences of the mass media, childhood antecedants to adult disease, or live in an increasingly culturally diverse society, for example – may force pediatric clinicians to adopt an even broader view of their role 347

348

Zuckerman, Parker

in supporting families and children. Given this shifting landscape, how can pediatric clinicians best serve the needs of families and children in the twenty-first century?

why should health care for young children change? Findings from the Commonwealth Survey, discussed throughout this book, make the concerns of parents clear. Four out of five parents surveyed expressed the need for more information and support on common childrearing issues, such as sleep issues, responding to crying, toilet training, discipline, and how to encourage their child to learn. The vast majority expressed satisfaction with their pediatric clinician, yet most felt that additional services such as a home visit by a nurse or a special telephone advice line would be useful. Approximately half of parents said they were having difficulty coping, and 40 percent said they felt frustrated by their child’s behavior or that their child got “on their nerves” at least once a day. These parents were also more likely to report negative disciplinary practices such as spanking, hitting, and yelling. These are not new needs, but the stakes seem higher today than in the past. In part, these concerns have acquired more urgency because of the convergence of four powerful sociocultural forces: the pace of social change, the growing disparity between the haves and the have-nots, new understanding of early brain development, and the link between parental health and child health and development. The Pace of Social Change The pace of social change continues to escalate. As a result, long-held cultural values have disintegrated and extended family and neighborhood support around child rearing has declined. As illustrated in the Commonwealth Study, “common sense” has lost its currency as the guiding principle of raising children, leading many parents to seek more information and support from pediatricians and other child experts in parenting magazines, television and, most recently, the internet. Beyond basic child rearing, many parents are insecure and troubled about just how to raise their children to cope with the vicissitudes of growing up in a world that little resembles the world of their own childhood. Should they try to raise a child who is focused on making friends and being cooperative, or would it be more advantageous to raise a child who is fiercely competitive and who can thereby secure a better life in an increasingly competitive workplace? Should they foster their child’s cognitive development, or is emotional development more important? Additionally, the list of potential new threats to children’s well-being seems to be expanding : parents’ returning to work, media sex and violence,

New Models of Pediatric Care

349

growing up in nontraditional families, dealing with community and family violence, and the quality of child care are just a few. As society grapples with new values and new challenges to children, and the media and internet present an increasing array of “experts,” each with a different message and a different agenda, to advise parents, it is no wonder that many mothers and fathers have become confused and look to their pediatric clinicians to help them make sense of the cacophony of mixed messages. Growing Income and Social Disparity While young children remain poor at an unacceptably high rate, the gulf between impoverished and advantaged families has significantly widened over the past 30 years (Wolff 1995). Approximately one child in five is living below the poverty line. Children of the “working poor” may be even more disadvantaged because they may not qualify for Medicaid, food stamps, supplemental income, subsidies for child care, and other income-based entitlements. For many of the working poor, high-quality child care alone can consume a disproportionate amount of their income. Overall, poorer children are more likely to experience double jeopardy than their more advantaged peers (Parker et al. 1988). First, they are more likely to be exposed to health risks (such as prematurity and malnutrition) and psychosocial risks (such as abuse and neglect). Second, they are likely to suffer greater consequences from those problems. For example, rates of iron deficiency anemia among 12–36-month-old children are approximately double in poor children and in black children (Looker et al. 1997; Centers for Disease Control and Prevention 1998). The current rates of high blood lead levels (lead > 10 mcg/dl) are also substantially higher in low income children (8 percent) than in higher income children (1 percent) and in black children (11 percent) than in white children (2 percent) (Centers for Disease Control and Prevention 1997). The sequelae of early iron deficiency anemia include persistent developmental delays, decreased attention to tasks, and poorer social interaction (Lozoff et al. 1991; Booth and Aukett 1997). Even mildly elevated lead levels have been associated with lower IQ, reading disability, worse coordination, and dropping out of school (Bellinger et al. 1992, 1991; Needleman et al. 1990). The deepening separation of the rich and the poor provides a challenge to pediatric medicine and its traditional role of devising strategies to disrupt the link between poverty and poor child health and development. Better Understanding of Brain Development and Early Learning New technology provides an understanding of brain development in children that emphasizes, to a degree not previously recognized, the importance of the first decade of life. To summarize: To a significant extent, the

350

Zuckerman, Parker

structure, or architecture, of the adult brain is permanently etched by early experiences. Using such methods as PET scans, MRIs, and counting synaptic density in pathological specimens, studies have demonstrated that the full complement of neurons (about 100 billion) is formed before the third trimester of gestation. The synapses between neurons, on the other hand, form largely after birth. For example, at birth the human brain has about 50 trillion synapses; by 3 years of age, the number is 1,000 trillion. But this is far too many to allow for coherent information processing; by age 15, the number of synapses has declined to 500 trillion (Huttenlocher and Dabholkar 1997). Scientists believe these initial excess synapses allow the organism to adapt to any evolutionary niche, from a solitary primitive hunter-gatherer to a sophisticated on-line member of modern society. The initial exuberant synaptogenesis is largely under genetic control. The loss of half of the synapses between ages 3 and 15 occurs through the “pruning” of superfluous neural connections and the selective strengthening of those that foster environmental adaptation (Shore 1997). This pruning is directed, to a significant degree, by the child’s experiences. In a sort of “neural Darwinism,” synapses that are not utilized through the child’s experience tend to wither, like plants that have not been watered, while those that are consistently used are cemented, perhaps for life. In this way, the connections of the brain are, to a great extent, socially constructed. Biology and experience are reciprocal and interdependent aspects of development. Unfortunately, the popular press and others have taken this information about brain development to mean that parents should relentlessly “stimulate” their infants and children, thereby improving their IQ scores and other intellectual functions. A cottage industry of “brain stimulating toys” has arisen. This interpretation is misguided by focusing too narrowly on cognitive development and undervaluing the importance of brain development for emotional functioning. Emotional and other types of circuitry are also being formed; the social and emotional experiences in the first few years of life provide the structural underpinnings for later emotional development and learning. Emerging research has also outlined the role of the hormone cortisol in shaping brain development and, in particular, setting patterns of response to stresses and modifying the neural connections needed for learning (Gunnar et al. 1997; Carlson and Earls 1997; Liu et al. 1997; Meaney et al. 1996; de Quervain et al. 1998). Cortisol is a central mediator of the body’s adaptive response to stress and in normal circumstances is released in the presence of a perceived environmental challenge. Receptors for cortisol in the brain function like a thermostat to monitor and maintain control of the cortisol-based stress response. The early postnatal environment, and in particular early exposure to nurturance and stimulation, appears to modify the brain’s ability to control the stress response (Gunnar et al. 1997). For

New Models of Pediatric Care

351

example, in a population of extremely deprived children in a Romanian orphanage, cortisol levels failed to turn off after a mild stress (Carlson and Earls 1997). On the other hand, children who had sensitive caregivers and were securely attached to the mother secreted less cortisol in the face of a minor stress (Liu et al. 1997). Laboratory work with animals confirms that early environment actually “programs” the nervous system’s response to stressors. Early maternal grooming of rat pups was found to increase gene expression of the cortisol receptors necessary to prevent excessive cortisol release in response to mild stress, even when the pups became adults (Meaney et al. 1996). Another animal study links cortisol to active learning. Rats taught to swim to a platform lost that capacity when subjected to stress-induced cortisol release. The stress did not interfere with learning when cortisol release was chemically blocked. Conversely, when cortisol was directly injected, learning was again impaired (de Quervain et al. 1998). Although we must be very cautious in extending those findings to human beings, the sum of this early work on cortisol helps to highlight one mechanism by which early experience shapes the brain’s role in behavior and learning. One way to operationalize this understanding of brain development is to view the goal of pediatric practice as preparing the young child for success – and there is little doubt that the ability to read is a vital foundation for that success. The new understanding of how children learn reading and writing, called emergent literacy, emphasizes the importance of skills acquired while children are still very young (Whitehurst and Lonigan 1998). Reading and writing were previously believed to be separate cognitive processes that required specific forms of tutoring to develop. As a result, formal reading and writing instruction began at school age, when children were seen as ready, and was thought to be more important in the development of literacy than exposure to books and reading aloud by parents. It is now understood that the development of literacy is a continuous process that begins early in life and is dependent on environmental influences (Whitehurst et al. 1998). When young children are exposed to books and reading, they begin to experience books as fun, learn to interpret pictures, follow the plot of a story, and attach meaning to the printed word. These emergent skills provide the foundation of later literacy. Books also provide opportunities for focused, pleasurable interactions between parent and child, supporting not only their literacy development but also social and emotional development, memory, curiosity, and fine motor skills. Early failure in learning to read affects from 20 percent to 35 percent of American children (Fitzgerald et al. 1991). Children growing up in poverty are at especially high risk. As a group, children from low income families have fewer books in the home and are exposed to relatively little

352

Zuckerman, Parker

reading aloud. Reasons for this include parents’ lack of discretionary income to purchase books, less understanding that young children should be read to, fewer children’s book stores in most urban or rural settings, and parental illiteracy or difficulty in school that precludes enjoyment in reading. For these reasons, helping parents promote literacy skills in their children may be the most effective way to address their desire, expressed in the Commonwealth Survey, for more information on helping their children to learn, promoting brain development, and enhancing school readiness. It is important to emphasize that reading should be mutually pleasurable. Indeed, children will learn to love books if they share books happily with a parent or other caregiver. Tape-recorded books for young children miss the important social and emotional contact of book sharing with parents. Many parents need information about book activities with children at different ages. Book sharing must not be too boring or too advanced. Six-month-old children like to mouth books and look at pictures of faces, 18-month-olds like to point to pictures, and 24-month-olds like to name pictures and complete sentences for parents. The Importance of Parental Health on Children’s Health and Development Advances in scientific knowledge have highlighted the importance of maternal health, not only during pregnancy, to children’s health (Wise et al. in press; Zuckerman and Brazelton 1994). As morbidity from infectious diseases declines, pediatric clinicians have the opportunity to turn their attention to the morbidities caused by parent-child interactions and parental health issues, including mental health, reproductive health, and other health behaviors. For this reason, advances in child health and development over the next decade will require much greater attention to women’s health and health services. Unsafe parental sex, for example, may lead to HIV or other sexually transmitted diseases in newborns. Unplanned, unwanted pregnancy can be associated with child abuse and behavior problems (Katon et al. 1995). Parental cigarette smoking (prenatally or after birth) is associated with respiratory, learning, and behavior problems and with increased likelihood of smoking during adolescence (Hurt et al. 1997). Parental mental health problems, especially alcoholism and depression, are associated with a myriad of behavior and learning problems and childhood injuries. The development of effective treatments for many of these parental problems requires new strategies and a redoubling of efforts to ensure that parents have access to health care for the benefit of their own and their children’s health (Zuckerman and Brazelton 1994; Katon et al. 1995; Hurt et al. 1997; Glasier 1997). Substantial barriers, however, currently limit access to comprehensive and continuous health care for women of

New Models of Pediatric Care

353

reproductive age. The range of uninsured women varies from 3 percent of those with incomes above 400 percent of the poverty line to 35 percent of those below the poverty line (Farley Short 1996). Among women receiving public insurance, there is often significant discontinuity. More than a quarter of Medicaid recipients under age 65 years disenroll within two years (Farley Short 1996) ; among recipients 16–34 years old, the rates of disenrollment are far higher (Carrasquillo et al. 1998). Approximately 15 percent of disenrollees are women who qualified only because of a pregnancy and lose coverage soon after delivery (Carrasquillo et al. 1998). According to one study, 40 percent of women bringing a young child (18 months or less) to pediatric care report at least one barrier to receiving health care (Kahn et al. 1999).

can expanded pediatric care be effective? Expanded and enhanced child development services within primary care settings would build on previous early intervention efforts to improve parenting and promote children’s development. What has been learned by such efforts? The results of many published interventions show positive changes in child development, including (and perhaps most important) an altered developmental pathway that leads to fewer arrests, episodes of running away, sex partners, and alcohol consumption (Kardy et al. 1998 ; BrooksGunn et al. 1990; Olds et al. 1986, 1994). Yet intervention strategies are difficult to generalize, given differences in location of services (home, center-based, or doctor’s office), intensity and duration of services, content, subjects, and measured outcomes. The positive effects of many interventions are most pronounced among subgroups of women or infants at risk for poor outcomes. Nevertheless, these clinical interventions comprise the best available evidence to inform clinical and programmatic activities to enhance and expand pediatric care. Interventions with more limited goals may be applicable to implement as part of pediatric practices. For example, one successful intervention provided parents with information and training to help them cope with irritable infants by identifying and responding to their infants’ cues (VandenBoom 1994). Best practices in pediatrics use child assessments as an intervention (Zuckerman and Parker 1997; Parker and Zuckerman 1990). The Brazelton Neonatal Behavioral Assessment Scale (BNBAS), the prototype assessment used as a therapeutic intervention, evaluates infant neurobehavioral functioning in such areas as interactive behaviors, selfregulatory abilities, physiological vigor, and neurological integrity (Adams et al. 1985). Conclusions from a review of studies examining the administration of the BNBAS in the presence of parents concluded that it can enhance their understanding of their babies’ behavior, especially if they

354

Zuckerman, Parker

are at greater risk for parenting dysfunction, and thereby improve parentchild outcomes (Parker and Zuckerman 1990). The use of the BNBAS as an intervention is based on the following guidelines: r Assessments must be performed sensitively and expertly if they are to be therapeutic. r Assessments are most beneficial for parents who are less knowledgeable regarding their child’s behavior and development and more at risk for suboptimal interaction. r Active parental involvement in the assessment process is necessary to achieve therapeutic results. r Ongoing assessments can have a greater positive impact than any single evaluation. r Assessments are best utilized as part of a comprehensive therapeutic program to enhance family functioning. In a study that combined use of the BNBAS with home visiting, positive outcomes continued to be seen when the children were five years of age, indicating a halo effect (Rauh et al. 1988). The low birthweight infants from the intervention group out-performed infants from the control group on measures of cognitive development and academic achievement, including higher standardized parent and teacher ratings of school functioning. These studies suggest that using the BNBAS (and other child assessments) in the pediatric setting can be an effective strategy to help enhance behavioral and developmental outcomes. Some studies have more directly used counseling strategies in pediatric offices as a base for promoting young children’s development. One involved a 20–30-minute postpartum group meeting and written instruction. This intervention was successful in achieving the goal of delaying the introduction of solid foods and juices until the infants were at least four months of age (Adams et al. 1985). A pediatric-based intervention for mothers of young infants involving specific discussions with the pediatrician (in addition to routine well-child care) enhanced the interaction between mothers and infants in the first months of life (Whitt and Casey 1982). Mothers were more responsive to interactional cues, had improved understanding of infant development and individuality, had enhanced feelings of confidence and competence to advance their infants’ outcomes, and had more appropriate play interactions. A similar approach not conducted within a pediatric practice produced similar findings (Metzel 1980). Findings from these studies represent important progress in establishing the effectiveness of early intervention. Adapting or linking effective interventions to the pediatric setting may result in a synergistic effect, producing a whole that is greater than the sum of its parts.

New Models of Pediatric Care

355

a new approach: a system of care focusing on child development A pediatric agenda that emphasizes child development, parental health, and helping parents promote their children’s health and development represents a significant change in the priorities of pediatric practice. The pediatric system cannot effect such significant changes alone. These measures need to be accompanied by simultaneous programmatic changes in public policy, health policy, public health, and business sectors and closer linkages among those systems. In the face of the economic pressures on pediatric clinicians to streamline care and increase productivity, how can those challenges be met? The present model of pediatric care has changed little in the past 50 years and cannot, we contend, adequately address this agenda. A single pediatrician (often in conjunction with a nurse) takes a history, conducts a physical examination, and provides health-related anticipatory guidance during a 5 to 20-minute visit. While solo pediatric practices are rare, most pediatricians still practice in a relatively independent fashion. In order to provide appropriate care, pediatric practices of the twenty-first century will need to be reconceptualized, enhanced, and expanded. Any new system of pediatric care will be based on the earlier work of Haggerty and colleagues (1975), Starfield (1992), and Alpert and colleagues (1976) and on health supervision guidelines developed by the American Academy of Pediatrics (Green et al. 1985), the Bright Futures initiative (MCHB 1995), and research on early intervention. A recent report on a study commissioned by Kaiser Permanente identified thirteen strategies being used by pediatric sites to create an early childhood development infrastructure and provide a seamless system of early childhood development services (Children NOW 1998). Those efforts, as well as our own experience at Boston Medical Center, lead us to propose a system of care that expands the range of services offered in the pediatric office through five inter-related strategies.

Expand Child Development Services Promote Child Development. A new system of pediatric care needs to expand and enhance efforts to optimize children’s cognitive, emotional, and physical development. This effort would supplement but not substitute for community-based family support efforts, such as child care resources and referrals, drop-in child care services, early Head Start, adult basic skills classes, and other information and support services to achieve similar goals. Parents often bring their anxieties, questions, concerns, vulnerabilities, and fears about their children, their families, and themselves to the

356

Zuckerman, Parker

pediatric office. Such concerns often result in a heightened sensitivity and receptivity to input from the clinician, making the pediatric visit a special opportunity to help parents feel more confident in their child-rearing roles. Pediatricians can make effective use of office visits via “teachable moments” (Metzl 1980) by promoting literacy, offering a greater variety of written materials or videos, and connecting parents to community resources or parenting groups. Because parents’ reading to and talking with a young child are crucial factors in children’s readiness to learn and language development, promoting early reading activities can be an integral part of pediatric primary care. As previously discussed, looking at books gives children a pleasurable and positive way to elicit parental attention. Although many parents know that reading aloud to children is important, some are not sure how to proceed and do not realize that they can support their child’s learning in infancy. In the Reach Out and Read (ROR) program, pediatricians introduce ageappropriate children’s books into the well-child visit, starting at 6 months. The pediatric clinician gives the child a book, comments on the child’s response to the book in the examination room, and offers information on the child’s cognitive development – for example, by pointing out that the child can turn the pages of a cardboard book or point to a named object in the book. The child is then invited to pick out a book to take home. For many parents, the very fact that the book comes from the pediatric clinician communicates the importance of reading more effectively than any amount of explanation could. An evaluation of the Reach Out and Read (ROR) program showed that mothers participating in the program were four times more likely and mothers on welfare eight times more likely to read to their children than were similar mothers who did not participate (Needleman et al. 1991). Approximately 1100 practices in the United States include ROR. Identify Problems of Child Development and Behavior. Developmental and behavioral surveillance provide opportunities for teachable moments and identification of potential problems. Developmental delays of any significance are likely to become apparent in the first years of life. A vital component of enhancing infant development is detecting such delays at an early stage and instituting prompt, appropriate diagnostic and therapeutic measures. Glascoe and Dworkin (1993) and others have stressed the importance, efficacy, and implementation of developmental surveillance (as opposed to screening) in young infants. Simple questions to the parents about developmental milestones and whether they have concerns about various aspects of their child’s development and behavior provide a reasonably sensitive and specific first screen for developmental problems. If parents are concerned or milestones are delayed, then the chances of a significant

New Models of Pediatric Care

357

developmental problem are much greater. If parents are not concerned, on the other hand, the odds are quite low. When concerns are raised, the pediatric clinician should use a standard developmental screening test or refer the child for an outside evaluation. Observation of the interactions and developing relationship between parent and child may detect possible emotional and attachment issues in a time-effective way (Greenspan 1995). Are warmth and affection evident throughout the exam? Are the parents reassuring and soothing? Do the parents respond appropriately to the child’s behavioral cues? Is there mutual eye contact, smiling, and regard? As the only professional seeing both parent and child, the pediatric clinician is best poised to detect an early poor fit, for example, between a child with a difficult temperament and an insecure parent. If such a mismatch is detected, early targeted interventions, either by the pediatric clinician or through referral for counseling, can prevent problems later on. The identification of problematic parentchild interactions is another important way to support children’s long-term development. Create a Two-Generation Approach to Child Health Our clinical experience suggests that the best way to help young children is to help their parents, and the best way to reach parents is through their children. New findings that have received inadequate attention suggest the need for a two-generation approach to child health, one that uses the pediatric visit to enhance parental health (especially health behaviors and mental health) by identifying problems and making referrals. The responsibility lies with the pediatric clinician, in part, because the pediatric primary care system may represent the only health system in which parents are consistently involved. Almost by default, it affords a rare window of opportunity to provide health promotion interventions for parents, thereby improving child outcomes. Family planning is an example of an important health behavior that could be addressed as part of an enhanced system of pediatric care. Unplanned, unwanted pregnancies can result in child abuse and behavior problems (Zurarin 1987; David 1986), and a short interpregnancy interval that can lead to low birthweight (Rawlings et al. 1995). Pediatric clinicians must continue to address unintended pregnancies among teenagers, yet they also should be aware that older women (including those who already have children) give birth to three times as many unplanned children as do teenagers despite new, highly effective contraceptives. The pediatric setting provides a forum to address child spacing, provide family planning information, and make referrals to family planning services. Women’s health before conception is receiving needed attention. Vitamin supplementation (especially with folic acid) before conception, for

358

Zuckerman, Parker

example, prevents birth defects such as spina bifida. Counseling regarding safe sex and maternal use of excessive alcohol, psychoactive drugs, and cigarettes during pregnancy can be addressed during the pediatric visit. Smoking cessation programs can be part of or linked to pediatric care. (This is especially important because of the well-documented adverse effect of cigarette smoking, both prenatally and within the first two years of life.) Nicotine gum and nicotine patches can be important aids to parents who try to stop smoking after the birth of their child. Mental health problems, especially depression, among mothers occur commonly and have well-documented adverse effects on children’s health, including increased risk of behavior problems, accidents, learning difficulties, attention deficient disorders, and other problems (Zuckerman and Beardslee 1987). Although drug addiction garners special attention as a mental health problem, alcoholism is the most common parental addiction. Approximately one in eight children in the United States has an alcoholic parent. While the prenatal effects of excessive alcohol use are important, the health and development implications of being parented by an active alcoholic are equally worrisome. Children of alcoholics have increased risk of injuries, hospitalizations, depression, and heavy drinking (Bijur et al. 1992; Olson et al. 1998). Pediatricians could play a number of roles in identifying and managing depression among mothers. Surveys have shown that mothers would welcome inquiries and assistance with depression and psychosocial issues (Kahn et al. 1999). Over half of pediatricians in a recent survey said they have responsibility for recognizing (but not treating) maternal depression (Kaplan-Sanoff 1995). A reported lack of knowledge and skills by the majority of pediatricians likely limits their effectiveness. This finding suggests that physicians in the primary care setting need training in identifying parental depression and initiating effective intervention. Establish Links to Other Services While pediatric clinicians cannot be expected to solve all of society’s ills, they can serve as a nexus for services and information that connect families to other needed clinical or community services and information. Links between the pediatric office and public health, family support, early intervention, parent support groups, and child care have special potential to address families’ needs. Formal mechanisms for linking pediatric offices with community-based child development services are needed. In the past, it has been difficult for pediatric offices to link up with community-based organizations because of communication roadblocks, interdisciplinary “cultural” differences, and lack of understanding of the services each does (and does not) offer. For example, physicians and public health-oriented providers may have separate conceptual frameworks for

New Models of Pediatric Care

359

prevention and treatment of childhood problems. Each system typically addresses the same issues independently. This was not always so. In the 1950s, public health providers supported well-baby clinics that went beyond traditional pediatric practices by offering home visiting services and education about diet, child care, and living patterns. Nurses were posted in schools to identify children with vision and hearing problems or other physical impairments. In this way, the public health sector ensured that all women and children received appropriate health education and that all children were screened for health problems and linked with a medical provider. New incentives based in managed care may catalyze the reinstatement of this kind of collaboration. Expansion of Medicaid and extended coverage for clinical preventive services such as immunizations and counseling serve as prototypes. As managed care shifts financial risk to pediatric clinicians and their organizations, resources available through the public health sector and elsewhere may look increasingly attractive for addressing child development and preventable problems (through, for example, immunization, bicycle helmet, and smoking cessation programs). Combining early Head Start and pediatric practices would create an important synergy for helping parents help their children. Provide Support to Parents Many parents of young children, especially first-time parents, feel insecure in the quality and appropriateness of their parenting. Although the pediatric clinician rarely has magic answers for the worried parent, the support and trust engendered during well-child visits often can suffice. Many clinicians underestimate the power of simply providing emotional support to parents who are experiencing difficult times, even if they lack the definitive answers to resolve their fears. Advocate for the Basics It is difficult and often impossible for families to use health or child development information if they lack the bare necessities of life: adequate housing, health insurance, food, and heat. For many disadvantaged families, the pediatric clinician may be the only professional who is aware of their plight and, more poignantly, its effects on their children. Understanding their situation may mean asking questions that do not come easily to many clinicians: Do you ever run out of food? How many people are living in your house? Do you have enough money to pay your bills? Serving as an advocate for parents in helping to provide those basics is the first step toward promoting infant health and development for many families (Kaplan-Sanoff 1997). For some pediatric practices, this means

360

Zuckerman, Parker

establishing easy access to a social worker or family advocate, plus developing an exhaustive understanding of local and governmental resources and entitlements. As a professional with some clout, a pediatrician can be extremely compelling in forcing an agency to respond to a family’s need. Public health efforts should be directed toward women to urge them to improve their attitudes and social norms regarding their own health and its implications, especially for their children. The basic message of such a campaign should be “You have taken great care of your child, now take care of yourself.” Health promotion efforts have gone a long way toward reducing cigarette smoking and drinking and driving, yet efforts aimed at women have primarily focused on post-menopausal women. Media messages should stress the mutual benefits for mother and child of positive health behaviors, such as breastfeeding, folic acid supplementation, and protection and safety against domestic violence, while also highlighting problems such as depression and cigarette smoking among mothers with young children.

healthy steps: a potential model Our experiences at Boston City Hospital (now Boston Medical Center) with a program called Pediatric Pathways to Success (Kaplan-Sanoff et al. 1997, 1995) and a national initiative called Healthy Steps (Zuckerman et al. 1997) provide one approach to enhance pediatric care. These services create an early childhood development infrastructure and community linkages that enhance and expand pediatric care to better meet the needs of parents. In these programs, the traditional pediatric primary care visit has been expanded to address goals in three areas: promoting child development, enhancing parental health, and ensuring “the basics.” A New Team Member To accomplish those goals, a new member of the pediatric team was added: the Healthy Steps specialist in Healthy Steps and the family advocate in Pediatric Pathways to Success. This new team member typically has a background in pediatric nursing, child development, or community experience and serves as the primary resource for the new model of pediatric care. This resource specialist’s responsibilities include: r Conducting office visits, either jointly with the pediatric clinician or alone during regular office hours or by appointment. r Conducting home visits to support and enhance parent and child interactions and to promote home safety. r Assessing children’s developmental progress as part of a health visit, as a separately scheduled office visit, or as part of a home visit.

New Models of Pediatric Care

361

r Helping parents manage common behavioral concerns related to early learning, fussiness, sleep, feeding, discipline, toilet training, and other issues. r Providing referrals and follow-up as appropriate to help families make connections within the community. r Facilitating parent groups. r Staffing a child development information telephone line. r Coordinating literacy promotion activities. For practices that serve primarily poor, culturally diverse populations, an individual (with or without a degree) knowledgeable in community resources and the community’s culture has proven valuable, backed up by a supervisor with a professional degree. The resource specialist’s first job has been to ensure that families are able to meet their basic needs, a necessary precondition if parents are to address their own health needs and meet their child’s developmental-behavioral needs. The resource specialist participates with the pediatrician and nurse in delivering many enhanced services. Costs and Benefits In an era of cost containment, it might be considered foolish to recommend new initiatives that add staff and potentially cost additional money unless benefits and/or value can be demonstrated. In Pediatric Pathways to Success, children who participated had fewer emergency room visits and hospitalizations. Specifically, during the first year of the program, infants who had a Pathways family advocate averaged 1.5 emergency room visits, compared with 5.3 visits among a historical comparison group; 18 Pathways children were hospitalized, compared with 33 in the comparison group. The charges for these services was $149,000 in the Pathways group and $328,00 in the comparison group (Kaplan-Sanoff 1997). Thus, although the findings need to be replicated with a more methodologically sound study design, the suggestion is clear that use of the family advocate paid for itself by reducing emergency room and hospital utilization in a group of relatively poor, culturally diverse children and their parents. A methodologically rigorous evaluation is currently examining the effects of the Healthy Steps program, composed of twentyone sites nationally. Many measures of outcome (including injuries, language development, utilization of services, and cost, as well as cost effectiveness) are being measured. Data from the study should inform payers, public and private, about the benefits and cost effectiveness of such a program. On an actuarial basis, any program that is effective in reducing unplanned or unwanted pregnancies and/or cigarette smoking would generate significant cost savings. Preventing unwanted or unplanned pregnancy,

362

Zuckerman, Parker

for example, would reduce costs associated with prenatal care, newborn care, and well-child care, as well as other costs associated with potential stresses on families and children. If the percentage of mothers smoking were reduced, the prevalence of low birthweight infants would likely decline, as would the number of children’s respiratory infections and learning and behavior problems, and the children would be less likely to smoke in adolescence. Finally, health care costs should be lower for the mother throughout her life. Savings in the utilization of medical care and perceived value to parents must be factored into any cost/benefit ratio. Value to parents could be especially important for nonprofit organizations, whose status depends on providing community benefits.

conclusion If Healthy Steps or other developmentally oriented primary care interventions prove to be cost effective and contribute to the value of pediatric care, employers may soon advocate for their implementation by HMOs and other insurers. Expanded services could also be made available if they are seen as conferring a competitive advantage in the marketplace or if parents are willing to make copayments to cover part or all of the costs. Finally, some services could be extended to lower income families if early Head Start and home visiting programs would locate publicly employed staff to primary care settings. To create a synergy of effort, staff from those programs would need to become members of the pediatric team. Our experience suggests that a new approach to pediatrics will be implemented only through strategic but straightforward reengineering efforts ; it will not occur spontaneously. As shown in the preliminary evaluation of the Pediatric Pathways to Success program, the team approach reduces hospitalizations and emergency room visits, outcomes that can likely be produced only from within the health care delivery system. Over and above these benefits, however, are the voices of the parents of the Commonwealth Survey, who set a clear mandate for pediatric clinicians to give them the support and information they desire. It is they who challenge the health care system (and their employers) to devise new systems of delivering and financing care in an era of “the bottom line.” They will support a pediatric system that fulfills their needs. Pediatricians need to support parents by advocating for a system that incorporates some or all of the necessary changes.

references Adams, H.M., Stern, E., & Stein, R. (1985). Anticipatory guidance: a modest intervention in the nursery. Pediatrics, 76(5), 781–786.

New Models of Pediatric Care

363

Alpert, J.J., Robertson, L.S., Kosa, J., Heagerty, M.C., & Haggerty, R.J. (1976). Delivery of health care for children: report of an experiment. Pediatrics, 57, 917–30. Bellinger, D.C., Stiles, K.M., & Needleman, H.L. (1992). Low-level lead exposure, intelligence and academic achievement: A long-term follow-up study. Pediatrics, 90, 855–61. Bellinger, D., Sloman, J., Leviton, A., Rabinowitz, M., Needleman, H.L., & Waternaux, C. (1991). Low-level lead exposure and children’s cognitive function in the preschool years. Pediatrics, 87, 219–27. Bijur, P.E., Kurzon, M., Overpeck, M.D., & Scheidt, P.C. (1992). Parental alcohol use, problem drinking and children’s injuries. Journal of the American Medical Association, 367, 3166–72. Booth, I.W., & Aukett, M.A. (1997). Iron deficiency anemia in infancy and early childhood. Archives of Disease in Childhood, 76, 549–54. Brooks-Gunn, J., Berlin, L., & Fulign, A.S. (1990). Early childhood intervention programs: what about the family? In J. Shonkoff & S. Meisels (eds.), Handbook of Early Childhood Intervention: Theory, Research and Treatment. New York: Cambridge University Press. Carrasquillo, O., Himmelstein, D., Woolhandler, S., & Bor, D. (1998). Can Medicaid managed care provide continuity of care to new Medicaid enrollees? An analysis of tenure on Medicaid. American Journal of Public Health, 88, 464–66. Carlson, M., & Earls, F. (1997). Psychological and neuro-endocrinological sequelae of early social deprivation in institutionalized children in Romania. Annals of the New York Academy of Sciences, 807, 409–28. Centers for Disease Control and Prevention. (1998). Recommendations to prevent and control iron deficiency in the United States. Morbidity and Mortality Weekly Report, 47 (no. RR-3), 1–30. Centers for Disease Control and Prevention. (1997). Update: Blood lead levels – United States, 1991–1994. Morbidity and Mortality Weekly Report, 46, 141–6. Children NOW. (1998). Right Place, Right Time: Managed Care and Early Childhood Development. Oakland, CA. David, H.P. (1986). Unwanted children : a follow-up from Prague. Family Planning Perspectives, 18, 143–44. de Quervain, D., Roozendaal, B., & McGaugh, J. (1998). Stress and glucocorticoids impair retrieval of long-term spatial memory. Nature, 394, 787–90. Farley Short, P. (1996). Medicaid’s Role in Insuring Low-Income Women. Washington, DC: RAND. Fitzgerald, J., Speigal, D., & Cunningham, J. (1991). The relationship between parental literacy level and perceptions of emergent literacy. Journal of Reading Behavior, 23, 191. Glascoe, F.P., & Dworkin, P.H. (1993). Obstacles to developmental surveillance. Journal of Developmental and Behavioral Pediatrics, 14, 344–349. Glasier, A. (1997). Emergency post-coital contraception (comments). New England Journal of Medicine, 337, 1058–64. Green, M., Brazelton, T.B., Fine, L., Korsch, B., Nelson, K., Willis, D., & Zuckerman, B. (1985). Guidelines for Health Supervision. Evanston, IL: American Academy of Pediatrics. Greenspan, S. (1995). Monitoring Social and Emotional Development in Young Children: Behavioral and Developmental Pediatrics. Boston: Little, Brown.

364

Zuckerman, Parker

Gunnar, M.R., Tout, K., deHaan, M., Pierce, S., & Stansbury, K. (1997). Temperament, social competence, and adrenocortical activity in preschoolers. Developmental Psychobiology, 31, 65–85. Haggerty, R.J., Roggham, K.J., & Pless, I.B. (1975). Child Health and the Community. New York: John Wiley. Hurt, R., Sachs, D., Glover, E., Offord, K.P., Johnston, J.A., Dale, L.C., Khayrallah, M.A., Schroeder, D.R., Glover, P.N., Sullivan, C.R., Croghan, I.T., & Sullivan, P.M. (1997). Comparison of sustained-release buproprion and placebo for smoking cessation. New England Journal of Medicine, 337, 1195–202. Huttenlocher, P.R., & Dabholkar, A.S. (1997). Regional differences in synaptogenesis in human cerebral cortex. Journal of Comparative Neurology, 387(2), 167–78. Kahn, R., Wise, P., Finkelstein, J., Bernstein, H., Lowe, J., & Homer, C. (1999). The scope of unmet maternal needs in pediatric settings. Pediatrics, 103, 576–81. Kaplan-Sanoff M. (1995). Pediatric pathways to success: the power of pediatric practice to support families. Zero to Three, 16(2), 12–17. Kaplan-Sanoff M, Brown T, & Zuckerman B. (1997). Enhancing pediatric primary care for low-income families : cost lessons learned from pediatric pathways to success. Zero to Three, 17(6), 34–36. Karoly, L., Greenwood, P., Everingham, S., Hoube, J., Kilburn, M., Rydell, P., Sanders, M., & Chiesa, J. (1998). Investing in Our Children: What We Know and Don’t Know About the Costs and Benefits of Early Childhood Interventions. Santa Monica, CA: RAND. Katon, W., VonKorff, M., Lin, E., Walker, E., Simon, G. E., Bush, T., Robinson, P., & Russo, J. (1995). Collaborative management to achieve treatment guidelines. Impact on depression in primary care. Journal of the American Medical Association, 273, 1026–31. Liu, D., Diorio, J., Tannenbaum, B., Caldji, C., Francis, D., Freedman, A., Sharma, S., Pearson, D., Plotsky, P.M., & Meaney, M.J. (1997). Maternal care, hippocampal glucocorticoid receptors, and hypothalamic-pituitary-adrenal responses to stress. Science, 722, 1659–62. Looker, A., Dallman, P., Carroll, M., Gunter, E., & Johnson, C. (1997). Prevalence of iron deficiency in the United States. Journal of the American Medical Association, 277, 973–6. Lozoff, B., Jimenez, E., & Wolf, A.W. (1991). Long-term developmental outcome of infants with iron deficiency. New England Journal of Medicine, 325, 687–94. Maternal and Child Health Bureau. (1995). Bright Futures. National Center for Education in Maternal and Child Health, Arlington, VA. Meaney, M.J., Diorio, J., Francis, D., Widdoweon, J., LaPlante, P., Caldji, C., Sharma, S., Seckl, J.R., & Plotsky, P.M. (1996). Early environmental regulation of forebrain glucocorticoid receptor gene expression: Implications for adrenocortical responses to stress. Developmental Neuroscience, 18, 49–72. Metzl, M. (1980). Teaching Parents a Strategy for Enhancing Infant Development, vol. 51. Needleman, R., Fried, L.E., Morley, D.S., Taylor, S., & Zuckerman, B. (1991). Clinicbased intervention to promote literacy. American Journal of Diseases in Children, 145, 696–698. Needleman, H.L., Schell, A., Bellinger, D., Leviton, A., & Allred, E.N. (1990). The long-term effects of exposure to low doses of lead in childhood. An 11-year follow-up report. New England Journal of Medicine, 322, 83–88.

New Models of Pediatric Care

365

O’Brien, S., Parker, S., Greenberg, J., & Zuckerman, B. (1997). Putting children first: the pediatrician as advocate. Contemporary Pediatrics, 103–18. Olds, D.L., Henderson, Jr., C.R., Chamberlin, R., & Tatelbaum, R. (1986). Preventing child abuse and neglect: A randomized trial of nurse home visitation. Pediatrics, 78(1). Olds D.L., Henderson, Jr., C.R., & Kitzman, H. (1994). Does prenatal and infancy nurse home visitation have enduring effects on qualities of parental caregiving and child health at 25 to 50 months of life? Pediatrics, 93(1), 89–98. Olson, A., Kelleher, K., Kemper, K., Zuckerman, B., Hammond, C., & Dietrich, A. (1998). Primary care pediatricians’ roles and perceived responsibilities in the management of depression in children, adolescents, and mothers. Unpublished manuscript. Parker, S., Greer, S., & Zuckerman, B. (1988). Double jeopardy: the impact of poverty on early child development. Pediatric Clinics of North America, 35, 1227–40. Parker, S., & Zuckerman, B. (1990). Therapeutic aspects of developmental assessment. In J. Shonkoff, & S. Meisels (eds.), Handbook of Early Childhood Intervention: Theory, Research and Treatment. New York: Cambridge University Press. Rauh, V.A., Achenbach, T.M., Nurcombe, B., Howell, C.T., & Teti, D.M. (1988). Minimizing adverse effects of low birthweight: four-year results of an early intervention program. Child Development, 59(3), 544–53. Rawlings, J.S., Rawlings, V.B., & Read, J.A. (1995). Prevalence of low birthweight and preterm delivery in relation to the interval between pregnancies among white and black women. New England Journal of Medicine, 332, 69–74. Shore, R. (1997). Rethinking the Brain: New Insights into Early Development. New York: Families and Work Institute. Starfield, B. (1992). Primary Care. New York: Oxford University Press. VandenBoom, D.C. (1994). The influence of temperament and mothering on attachment and exploration: An experimental manipulation of sensitive responsiveness among lower class mothers with irritable infants. Child Development, 65. Whitehurst, G., & Lonigan, C. (1998). Child development and emergent literacy. Child Development, 69, 848–72. Whitehurst, G., Falco, F., Lonigan, C., et al. (1988). Accelerating language development through picture book reading. Developmental Psychology, 24, 552–9. Whitt, J.K., & Casey, P.H. (1982). The mother-infant relationship and infant development: the effect of pediatric intervention. Child Development, 53. Wise, P., Kahn, R., & Zuckerman, B. (In press). Enhancing the Health of Women and Their Children. Wolff, E. (1995). How the pie is sliced: America’s growing concentration of wealth. American Prospect, 22, 58–64. Zuckerman, B, & Beardslee, W. (1987). Maternal depression: an issue for pediatricians. Pediatrics, 79, 110–117. Zuckerman, B, & Brazelton, T.B. (1994). Strategies for a family supportive child health system. putting families first. In Kagan, S., & Weissbourd, B. (eds.), Putting Families First: America’s Family Support Movement and the Challenge of Change. pp. 73–92.

366

Zuckerman, Parker

Zuckerman, B., Kaplan-Sanoff, M., Parker, S., & Taaffe Young, K. (1997). The Healthy Steps for Young Children program. Zero to Three, 17(6), 20–25. Zuckerman, B., & Parker, S. (1997). Teachable moments : assessment as intervention. Contemporary Pediatrics (December), 41–53. Zurarin, S.J. (1987). Unplanned pregnancies, family planning problems and child maltreatment. Famly Relations, 36, 135–39.

13 Families with Children Under 3: What We Know and Implications for Results and Policy Neal Halfon and Kathryn Taaffe McLearn

This volume offers a rich and detailed picture of the lives of families with very young children. Using data from The Commonwealth Fund Survey of Parents with Young Children as a starting point, each contributor has provided important insights about the context of early child rearing and the challenges young families face, the supports they receive, especially from the health system, and how they are performing the important tasks of rearing young children. Such a broad portrait, based on nationally representative data, has not been attempted before. What emerges is a picture of the complex forces that influence families and their child-rearing behaviors during a period when much is happening and changing in the life of a young child. In this concluding chapter, we synthesize major themes and consider how the research findings, analysis, and interpretations presented by the contributors could inform public policies targeted at improving the lives of young children and their families. We begin by considering the family, community, and policy context of early childhood in order to frame and interpret the results. We then summarize the major findings and consider their implications for the development of specific and general policies in three distinct areas: the health care system, particularly the provision of pediatric health services; community-level efforts to support families with young children; and wider federal and state policies, including funding for family support and coordination of early childhood programs.

the changing context of child rearing in american families As explained in the Introduction to this book, childhood and the context for rearing young children are changing in numerous ways. First, changes in family demographics and structure appear to be having a profound influence on how families are rearing young children. Changes in 367

368

Halfon, McLearn

family demographics are also transforming public policies that target the needs of families with young children. For example, results presented in this volume document findings that a majority of young children are being reared in families where both mothers and fathers are working, where income and earning potential are often less than what is needed to support even basic needs, and where time demands and relationships with extended families create new pressures and undercut traditional supports. Second, family relationships and trends in child rearing are also changing in significant ways. For example, data presented by Fuligni suggest that more young children are being raised in families where fathers have an important role, and indeed where basic child-rearing responsibilities are increasingly shared by both parents. This important trend is a product not only of changing employment patterns related to the increased number of women with young children in the workforce, but also of more fundamental changes in social roles and normative behaviors. These changes have profound implications for how families develop, function, and perform their child-rearing duties. Simultaneously and third, the cultural and ethnic diversity of families with young children has increased as well. In states like California and Texas, over half of the babies born in 1997 were born to Hispanic women, and a large proportion were born to mothers who were themselves born outside the United States (Texas Department of Health 1997; California Department of Finance 1997). Future demographic projections indicate that this trend will continue and be extended to many more states. While immigrants have always represented a significant proportion of new U.S. families, understanding the impact of different and diverse cultural traditions on child rearing has become important for planning and implementing service and support programs for children and families. Consider, for example, the culturally influenced practice of where a young child sleeps. The dominant cultural expectation in the United States is that a child sleeps alone in a crib, often in a separate room, as early as possible. In fact, both lay and professional monographs present the necessary steps by which parents accomplish this difficult task (Ferber 1990). In contrast, for many cultures, the norm is for the child to sleep with parents or other family members. Some have argued that this difference reflects divergent cultural values relating to individualization and independence, on the one hand, and the value of interdependence and harmony, on the other (McKenna 2000; Morrelli 1992). In any case, it raises real issues for the pediatrician, visiting nurse, or child care provider who interacts with a family about health, behavior, or development issues (Anderson 2000). In this volume, various contributors highlight how different patterns of parental behaviors, including reading, breastfeeding, or utilizing time-outs as significant discipline strategies, vary according to ethnicity.

Families with Children Under 3: Policy Implications

369

Appreciating the role of different cultural traditions is likely to promote better understanding of differences and disparities in children’s development, as well as assisting with the process of planning and customizing family-based interventions and services targeted at children under three and their families (Garcia-Coll and Magnusson 2000). For example, as the chapter by Fuligni and colleagues shows, even as fathers are playing a significantly greater role in child-rearing activities and more shared caregiving is taking place in families with young children, these trends vary dramatically among different cultural and ethnic groups. Fourth, there have also been large changes in the normative values and social expectations that are transmitted to parents and the public about what it means to be a parent and how a parent should behave. The prominence of research on brain development is bringing about changes in how America thinks about the early years – so much so that that the topic has received considerable attention in the media and through foundation initiatives and activities of nonprofit research and policy organizations. In the late winter and early spring of 1997, both Newsweek and Time magazines issued special issues or cover stories on children from birth to age 3 and early brain development. The explosion of information represented in magazines, books, television programs, and the internet about child development and child rearing provides new parents with a range of (often untested) new information and suggestions about skill-building activities for their children and themselves (see www.expertparent.com). The growing market of early childhood specialty stores (such as Smart Kids, Right Start, or Buy Buy Kids) popping up in shopping malls across the United States and the ubiquitous availability of baby Mozart CDs and video tapes are perhaps the best indicators of changing social norms and expectations being placed on parents. These and other marketing trends suggest ways in which the “early childhood market” is evolving and the role that new economic incentives are playing in redefining acceptable and desired behaviors and norms. Computer programs that allow even an infant to execute mouse-driven manipulations of cute cartoon characters on a television monitor are increasingly marketed to parents and prominently displayed in toy stores alongside more traditional alphabet blocks. In an iformation technology economy where strategies for success are being redefined, it appears that the acquisition of information processing skills cannot start too early. Echoing the newfound importance of early childhood, major foundations have sponsored task force reports and worked with nonprofit organizations to call attention to the needs of young children and their families. In 1994, Starting Points: Meeting the Needs of Our Youngest Children (Carnegie Corporation 1994) called for action in four interrelated areas – parenting, child care, health, and community mobilization – and became the basis for community and state-level early childhood initiatives across the United

370

Halfon, McLearn

States. The I Am Your Child Foundation, the brainchild of actor-producerdirector Rob Reiner, produced a major prime time television special in the spring of 1997 and is creating educational videos on parenting that will be made available to every new parent in California. Reiner also sponsored a successful 1998 ballot initiative in California to dramatically increase funding for community-based early childhood services and supports for families with young children. This and similar efforts in North Carolina, Vermont, Minnesota, and other states are focused on building comprehensive, publicly funded early childhood education, development, and health systems (Cauthen et al. 2000). The Commonwealth Fund, in partnership with over seventy national and local funders, launched the Healthy Steps for Young Children Program, which links primary health care for infants and toddlers with parent education and family support in over twenty five medical practices nationwide. Numerous nonprofit research and policy organizations have focused their work on furthering the agenda for families with very young children. For example, national advocacy and educational organizations like Zero To Three have played a major role in translating technical research on child development and has developed key messages on what is at stake in fostering a child’s development in the early years. The Families and Work Institute has been a central player in the Early Childhood Public Engagement Campaign, a national public awareness multi-media campaign that focuses on the most effective ways that families and communities can support the healthy development and school readiness of young children. In addition, the Institute has conducted research to document the way employment and workforce changes have challenged many families and influenced their child-rearing capacities. For nearly two decades, the National Center for Children in Poverty has documented the extent of early childhood poverty in the United States and the impact of poverty and current social policies on the lives of young children. The Center’s recent Map and Track reports collect and analyze state data on child well-being and investments in young children and families (Cauthen et al. 2000). Fifth, early childhood was highlighted in the national policy arena as a result of two White House conferences in 1996. The first White House Conference on Early Childhood Development and Learning focused on the revolution in neurobiology and developmental psychology and the importance of children’s earliest experiences in helping them get off to a strong and healthy start and reach their full potential. The conference provided an opportunity for the White House to make a series of policy announcements to demonstrate commitment to young children. These policies included improving child care in the military, providing health coverage for children, expanding Early Head Start, giving parents and caregivers early childhood tool kits and establishing a “Safe Start.” The second White House Conference on Child Care signaled recognition of the importance of child care on the national agenda. Despite occasional

Families with Children Under 3: Policy Implications

371

bursts of attention, historically child care had been addressed only sporadically by public policies. It has never been clear who has been responsible for child care: parents who need the service, corporations that would like to offer it but deem themselves constrained by fiscal realities, or the government, which is uncertain of how much support to provide to whom (Kagan and McLearn 2000). The passage of welfare reform as the Personal Responsibility and Work Act (PRWA) of 1996, with new mandatory work requirements for mothers and the resulting expansions in child care, put pressure on the federal government to plan and fund efforts to expand the availability and improve the quality of child care. Convened after the passage of PRWA, the White House Conference on Child Care provided a national platform for child care. The conference highlighted research documenting the critical role of high-quality child care in the healthy development of young children, coupled with the many factors that conspire to make the care of young children so weak: inadequate regulatory provisions, inadequate provider training, and inadequate compensation and funding. The conference also suggested new initiatives to expand availability and improve quality. Sixth, scientific revolutions in neurobiology, genetics, and developmental biology are transforming our understanding of early childhood and the long reach that childhood has on adult health and social outcomes. Several distinct research traditions have emphasized that adult health and functional status is not merely the result of cumulative experiences that add up over the lifespan, but also of biological, psychological, and social programming during critical and sensitive periods in human development (Wadsworth 1999; Sylva 1997). New research evidence has increasingly demonstrated that environmental and social influences and experiences are embedded into biological pathways during these sensitive periods, defining patterns of autonomic, immunological, and endocrinological reactivity that persist from childhood to adulthood (Brunner and Marmot 2000; Kandel 1998). New understandings of these influences and patterns are illuminating the ways in which developmental trajectories are established early in life and can be influenced by changes in a child’s experiential environment (Halfon and Hochstein, in press). In the fall of 2000, the National Research Council and Institute of Medicine released From Neurons to Neighborhoods: The Science of Early Childhood Development. A product of a two and a half year study by the Committee on Integrating the Science of Early Childhood Development, the report reviewed what is known about the nature of early development and the influence of early experiences on children’s health and well-being. It suggested the following conclusions, as supported by the science of early childhood: r The traditional “nature versus nurture” debate is simplistic and scientifically obsolete.

372

Halfon, McLearn

r Early experiences clearly influence brain development, but a disproportionate focus on “birth to three” begins too late and ends too soon. r Early intervention programs can improve the odds for vulnerable young children, but those that work are rarely simple, inexpensive, or easy to implement. r How young children feel is as important as how they think, particularly with regard to school readiness. r Healthy early development depends on nurturing and dependable relationships. r Culture influences all aspects of early development through childrearing beliefs and practices. r There is little scientific evidence that special “stimulation” activities above and beyond normal growth-promoting experiences lead to “advanced” brain development in infancy. r Substantial scientific evidence indicates that poor nutrition, specific infections, environmental neurotoxins, drug exposures, and chronic stress can harm the developing brain. r Significant parent mental health problems (particularly maternal depression), substance abuse, and family violence impose heavy developmental burdens on young children (Shonkoff and Phillips 2000). In addition to these conclusions, Neurons to Neighborhoods called for a new public dialogue about the shared responsibility for children in the United States and suggested a need to: r Focus greater attention on the social and emotional development and mental health needs of young children. r Recognize the significance and important role of early childhood caregivers and educators. r Enhance supports for working families with young children. r Integrate child development research, neuroscience, and molecular genetics. r Integrate the basic science of human development and the applied science of early childhood intervention. r Improve evaluations of early childhood interventions in order to improve the quality of programs available to young children and their families (Shonkoff and Phillips 2000). The ability of early experience to program later behaviors has been illustrated in several animal and human experimental models (Suomi 1999; Boyce et al. 2000). Models of development of depression in childhood and antisocial behavior in adolescence show how risk and protective factors influence a child’s trajectory and pattern of resilience and reactivity (Aguilar et al. 2000; Graham and Easterbrooks 2000).

Families with Children Under 3: Policy Implications

373

Seventh, the macro- and micro-economics of early childhood has become a focus of analysis (Danziger and Waldfogel 2000). International economic summits have called attention to national strategies to invest in human capital and the effect of those strategies on social, health, and developmental outcomes in different nations (Keating and Hertzman 1999). Many advanced industrial countries have approached the needs of young children and their families more explicitly than the United States has done. Several recent reports have highlighted early childhood policies in France, Sweden, Italy, and other advanced industrialized countries, and how those differences could potentially manifest in short- and long-term child outcomes (Kamerman and Kahn 1995; Keating and Hertzman 1999). Such arguments have become even more salient as differences in educational achievement are translated in economic terms into differences in production, innovation, and competitive advantage. In the United Kingdom, the newly minted Sure Start Program is helping local communities to provide comprehensive early childhood services in order to improve early childhood outcomes. What is of interest in this program is that it is being managed by the Department of the Treasury as an investment in longterm human capital. Recent studies of the economics of early childhood have highlighted the importance of investing in the early years, and the potential return on investment that is attainable by a society that pursues such a strategy (Smith 1999; Heckman 1999). As the emerging research on educational achievement clearly points to the impact of early childhood programs and family-centered services on children’s school readiness and educational outcomes, it has affected recent public policy decisions in the United States as well. For instance, public policy aimed at creating universal preschool programs are being justified not only because they might be beneficial for children in their own right and could minimize potential disparities in educational achievement, but also as a means of guaranteeing global economic competitiveness of the U.S. work force. Recently, Keating, Hertzman, and colleagues from the Canadian Center for Advanced Studies have promoted a greater understanding of the implications of optimizing development in early childhood on the health, educational, and social achievement and social capital of nations. Arguing that early childhood is the crucible in which major social disparities are born, they use cross-national data depicting the social gradients of functional literacy within different nations to illustrate how disparities relate to socioeconomic status (Keating and Hertzman 1999). They show that functional literacy follows a much steeper gradient in the United States than in nations such as Canada and Sweden, where differences in functional literacy vary less across socioeconomic strata. Keating and Hertzman suggest that the slope and magnitude of these gradients are related to differences that originate early in life and are compounded over the lifespan. They

374

Halfon, McLearn

also suggest that differential outcomes observable early in a child’s life are magnified as populations of children grow to become the next generation of citizens and workers, and that these outcomes are the fundamental determinants of the economic and social attainments of different nation states. Given the impact of the social, psychological, economic, cultural, and informational trends on young children and their families, the analyses put forth in this volume are all the more important. What these trends highlight is the importance of early childhood as the focus of multiple complex and competing influences, and the launching point for a number of developmental processes that have life-long influences for individuals, families, and the larger society. The developmental trajectories initiated early in life set in motion life-long learning and functional trajectories. As illustrated in Figure 13.1, risk reduction or health promotion strategies can serve as protective factors to mitigate the influence of risk factors on a child’s developmental trajectory. The policy implications that we draw must respond to a multifaceted context and its impact on these developing trajectories, as well as to the current policy environment for families with young children.

social policy for young children and families since 1960 Although no coherent national policy or social strategy currently exists for infants and toddlers, the dramatic contextual changes affecting young children and their families that we have just outlined are beginning to change the policy environment. Historically, infants and toddlers have remained remote from the specific allocation of durable services, and policy attention to their needs has been well circumscribed. Until the 1980s, most mothers of young children, across the income spectrum, were not in the workforce. Health care was less costly, and in the decades after World War II, most working families received health insurance coverage from their employers. Family support was provided by less mobile and more geographically proximate family members. Parents, extended families, and local private, faith-based, or grassroots services seem to have filled the need. The demographics, family relationships, and mobility patterns of the past era did not compel policy attention or action. Beyond demographics, prevailing attitudes also prevented the coalescence of a public will to enact policy for infants, toddlers, and their families. American child and family policy is predicated on a noninterventionist approach that presumes families are able to care for themselves and nurture their young (Grubb and Lazerson 1990). Americans have been reluctant to involve government in the private lives of families, and especially families with young children. The architects of American democracy promoted a fundamental belief that families should bear the greatest responsibility for

0

HP

HP

20

Protective factors

HP

RR

RR

RR

Risk factors

40 Age (Years)

60

Health promotion strategies

HP

80

Optimal Health Development Trajectory

Health Development Trajectory Without RR and HP strategies

Risk reduction strategies

RR

figure 13.1 Influence of Risk Reduction and Health Promotion Strategies on Health Development Trajectories. Source: Halfon, N.,Inkelas, M.,Hochstein,M. (2000). The health development organization: An organizational approach to achieving child health development. The Milbank Quarterly© 78(3): 447–497.

Health Development

376

Halfon, McLearn

nurturing their children, with government playing a secondary role, if any (Kagan and McLearn 2000). Designing public policy that supports families’ caregiving role and focuses on the internal life of the family is fraught with age-old controversies that have pitted right against left on the political spectrum (Lakoff 1996). Until recently, family policy focused primarily on what Kamerman and Kahn have called the bread-winning roles and served to support the family as a viable economic unit (1978). Government tended to intervene in the privacy of the family only in those “exceptional” cases when families failed economically, parents died or became completely disabled, or more recently when children were significantly abused or neglected. The provision of services to children has been a target of criticism from more conservative policymakers because of concerns about family self-sufficiency and the assumption that government programs can foster long-term dependence. Such critics argue that building more supports for families with young children will not benefit poor, largely minority families, and will encourage more out-of-wedlock births and dependency on public systems (Kaus 1992). A related concern of individuals with this viewpoint is the belief that any social policy designed to support families should be directed toward “traditional” families and not used to encourage or sustain other forms of family organization. In addition to doctrinal and ideological arguments about the proper role and desirability of family policy, there are other practical considerations, such as the heterogeneity of the U.S. population, the presence of varying microenvironments within local communities, and barriers related to families’ racial, ethnic, religious, and cultural differences (McLloyd 1990; Garcia Coll and Magnusson 2000). Nonetheless, with the exception of the early 1980s, the past four decades in the United States have witnessed increasing policy activism focused on the needs of families, and especially those with young children (Shonkoff and Meisels 2000). The 1960s and early 1970s witnessed the passage of “War on Poverty” programs such as Head Start, Medicaid, and the Women, Infants and Children Food Program (WIC). These programs have played a significant role in providing health care, nutrition, and early child development services for economically disadvantaged young children throughout the United States (Currie and Thomas 1995). Inspired by his family’s own experience with mental retardation, President Kennedy stimulated a decade-long focus on the treatment and rehabilitation of individuals with mental retardation and other developmental delays. Public Law 88-156 provided new federal funding for projects for children with mental retardation in 1963, and in 1968 Public Law 90-538, the Handicapped Children’s Early Education Assistance Act, provided funds for model early intervention programs. The 1970s witnessed additional funding for children with disabilities. With the passage of Public Law 94142, the Education for All Handicapped Act in 1976, the right to a full and

Families with Children Under 3: Policy Implications

377

appropriate public education was established for all children, regardless of the presence of a disability. The new law brought with it the provision of screening for disabilities in all public schools and greater recognition of the early origins of learning disabilities. While early intervention programs for developmentally disabled children took shape, Head Start focused on the developmental competence of children from disadvantaged environments, using a bold, multidisciplinary preventive approach that has continued to expand until the present day. During the early 1980s, many of these programs were threatened with significant cutbacks in funding as a result of the Omnibus Budget Reconciliation Act of 1981. This law consolidated many existing programs into block grants, which significantly reduced appropriations. However, a determined Congress later in that decade responded by expanding funding and passing several significant pieces of legislation. In 1978, the U.S. House of Representatives Select Committee on Children, Youth and Families had been established under the chairmanship of Representative George Miller and with broad representation across the political spectrum. Congressman Miller succeeded in creating a pro-family umbrella under which members from both sides of the aisle could participate. During the mid 1980s, Congress was also able to expand the Medicaid program and decouple it from the Aid to Families with Dependent Children (AFCD) program, and dramatically expanded coverage to pregnant women and young children by mandating coverage well above the federal poverty thresholds (Hutchins 1997). The 1989 Omnibus Budget Reconciliation Act further expanded Medicaid health insurance coverage to children ages 1 to 6 years, with a plan to cover all children below the federal poverty level, using a phase-in formula by year 2002. Among the major laws passed during this period were Public Law 99-457, which extended the Education for All Handicapped Act to include infants, toddlers, and their families (1986); the Comprehensive Child Development Act (1987); Public Law 100-385, the Family Support Act (1988), an initial attempt at welfare reform; and the Child Care and Development Block Grant (1990). This activity continued into the 1990s, with the passage of the Family and Medical Leave Act (1993), expansions of the Earned Income Tax Credit, and the passage of the State Child Health Insurance Program (1997). In June 1991, the bipartisan National Commission on Children, chaired by Senator John D. Rockefeller, released its final report, Beyond Rhetoric (National Commission on Children 1991), with recommendations for a series of new policy initiatives to promote income security by expanding the Earned Income Tax Credit, increasing educational achievement by expanding Head Start, strengthening and supporting families, and protecting vulnerable children. What emerged from this and similar bipartisan efforts was a reaffirmation of the policy goal of enabling families to achieve

378

Halfon, McLearn

financial independence. There was also recognition that independence was not likely to be achieved or sustained without adequate child care, health care, and workplace supports to enable parents to maintain their financial independence and also perform their roles as parents (Davies and Jacobs 1991, Carnegie Corporation 1994). The economic status of young families during the 1970s and 1980s was also changing. As opposed to the prosperity that marked the 1950s and 1960s, the 1970s and 1980s witnessed rather flat or decreasing earning capacity for entry-level workers, and large increases in childhood poverty rates. The economic recession of the early 1980s resulted in childhood poverty rates of approximately 23 percent, with even higher rates for families with children less than six years of age (National Center for Children in Poverty 1996). Also significant was the growth during these years of the poverty rate among two-parent families, largely attributable to low wages (Bane and Ellwood 1989). During one of the most significant peacetime economic recoveries of the twentieth century, following the recession in the 1980s, childhood poverty rates remained above 20 percent nationally and over 30 percent in many urban areas. As the United States entered the 1990s, nearly a quarter of all children under 6 were still living in poverty. Inflationadjusted wages dropped 13 percent between 1979 and 1989 for workers in the poorest tenth of the workforce. Only with significant changes in tax policy, and specifically expansions of the Earned Income Tax Credit in the early 1990s, did childhood poverty rates begin to drop in 1995. The impact of family poverty and the social ecology of childhood was also being transformed during this time period. With more single mothers, more women in the workforce, and greater family mobility, traditional child-rearing supports were also changing. Available non-parental child care increasingly had to be paid for. The need for families to pay for child care and other traditional supports put even more financial demands on young families (Gabarino and Ganzel 2000). During this period, the role that poverty and the twin influences of economic and social deprivation on the social ecology of early risks became more fully appreciated (Bronfenbrenner 1986). While studies had documented the relationship of poverty and infant mortality, new studies were linking poverty and child maltreatment, and the cumulative risk profiles that were measurable as lower preschool intelligence (Sameroff 1987; Gabarino and Ganzel 2000). As Davies and Jacobs have argued in their analysis of this period (1991), several other tensions were resolving themselves and formulating the basis for child development and family policy at this time. Ecological theories of child development, focusing on the role of transactions between the child’s environment and individual and genetic characteristics, were becoming broadly accepted by academic and practicing child development specialists (Sameroff 1987; Bronfenbrenner 1979). This model suggests that

Families with Children Under 3: Policy Implications

379

the family, rather than the individual child, should be the focus of analysis and the target of therapeutic options (Bronfenbrenner and Morris 1998; Bronfenbrenner 1979, 1986). Family therapy and family systems theory were also shifting the assessment and treatment of individual problems into a family context. Both models supported family-centered rather than individually focused interventions and approaches to improve child development and family functioning. A family-centered and community-based approach to preventing disability and promoting optimal child development outcomes emerged and was codified in federal laws such as PL-99-457 (Part C of IDEA), passed in 1986. Significantly, this law required states to develop comprehensive systems of early intervention for infants and toddlers with developmental delays or disabilities, making all children from birth to age two with developmental delays eligible for services. To receive funds, states must demonstrate a coordinated approach, with a lead agency and documentation of coordination between health and education providers and agencies (Shonkoff and Meisels 2000). By explicitly acknowledging that families, rather than the potentially disabled child, are the central focus of professional services, this law brought child disability funding into a closer connection with contemporary notions of the importance of families in influencing a child’s development. While child and family policy in the United States has traditionally been targeted rather than universal and has focused on the “exceptional” child and those who are most needy, several policies developed in the 1990s had a more universal focus. Notions emerging from life course developmental psychology suggested that the development of all children could be improved by more universal programs. In many European nations, such as France, universal programs had become the norm (Kamerman and Kahn 1995). Support for broad-based universal programs emerged during the 1990s through programs such as California’s Proposition 10, which was passed by California voters in 1998 and targets a new $700 million yearly allocation toward improving the development of all young children in California. Over the past five years, a number of states have focused greater policy attention on young children and families (Cauthen et al. 2000). As of 2000, thirty one states were funding one or more child development and family support programs for infants and toddlers, and state funding levels had more than doubled since 1998. The programs for infants and toddlers take various forms, including supplementing the federal Early Head Start program, providing home visit programs to families with newborns, and paying special attention to families with multiple risks. Even so, nineteen states were not funding programs that specifically targeted infants and toddlers. State investments in prekindergarten programs continue to grow, with forty-three states now supporting pre-kindergarten services that include a wide range of child development supports, such as Head

380

Halfon, McLearn

Start, child care, and full and extended day prekindergarten programs for working parents. Another important tension that has driven the evolution of child and family policy in the United States is the role of the private sector in improving investments in children. Over the past two decades, the Committee for Economic Development (CED), representing business leaders and educators, has issued a number of reports on the importance of private and public investment in America’s children. The CED reports make the business case for why investing in the lives of young children is important to the nation’s bottom line. The first report, Investing in Our Children: Business and the Public Schools (1985), identified quality education as the most important investment a nation can make in its future productivity and competitiveness. This was the first report issued by a business organization to target preschool education for poor children as a superior investment and one of the most effective for preventing school dropout. Subsequent reports – The Unfinished Agenda (1991) and Why Child Care Matters (1993) – based their analyses on the notion of the instrumental value of children for the future of the American workforce. Their arguments have considered declining fertility rates, the improved economic status and longevity of the elderly, and the importance of each worker’s contribution to future social security. As these and other studies suggest, the importance of work and employment-related issues for families with young children is high, and employers have an important role to play in ensuring the quality of early childhood, in coordination with public policy. From this somewhat uniquely American perspective, effective investment in human capital is seen to require both public and private investment strategies to ensure children’s optimal development.

major findings: a portrait of families with young children Given the current and ever-changing social policy context, this book has addressed several important questions about families with young children. Who are the parents who are rearing children in the United States today? How are they faring in meeting the challenges of caring for our nation’s youngest children? What are their concerns? What kind of instrumental support are they receiving from the health care system? Who Are the Parents? The national picture of parents with young children that emerges from the Commonwealth Survey of Parents with Young Children is consistent with and supports other national surveys of families (Moore 1996). The majority of these families (73 percent) identified themselves as white;

Families with Children Under 3: Policy Implications

381

13 percent identified themselves as Latino and represent a large and growing population in many regions in the United States. Parents are older, with nearly half of mothers reporting they were age 30 or older at the time of the child’s birth. Parental education is also increasing, with over half of parents (53 percent) reporting some college or college graduation. And similar to national data, 86 percent of these infants and toddlers are currently growing up in two parent families, while 14 percent are being reared in single parent households. A great number of new parents in the United States are having trouble making ends meet. The Commonwealth data found 24 percent of families had incomes less than $20,000, and 54 percent had incomes less than $40,000. Of real importance is the fact that 47 percent of parents report that they received some form of government assistance since the time of their child’s birth. This is also reflected in the high proportion of parents – many of them working – who report having trouble paying for everything from prenatal care, medical care, and child care to basic supplies like formula, food, diapers, and other necessities (14–24 percent). The reliance on government support programs (such as Medicaid, food stamps, and WIC) for families with very young children is related less to the generosity of public benefits and policies that support families with young children than to the large number of families with incomes below 200 percent of the poverty level, making them eligible for government programs. Results presented indicated that many of the instrumental conditions for parenting are improving. Most mothers (95 percent) receive prenatal care in the first trimester, an important indicator of access to comprehensive risk reduction strategies during pregnancy. Chapter 3 on preparing for parenthood concludes that most new parents are reasonably well prepared, with 95 percent of parents reporting that they wanted the child and 75 percent having attended prenatal birth or parenting classes. However, the contributors do point out that the overall confidence of new parents is actually much lower, with only 40 percent expressing confidence in their abilities and coping well. As has been reported in the past, education and age are associated with planned pregnancy and preparation for parenthood, with AfricanAmericans and Hispanics reporting less support. While McLanahan and colleagues address the important role that parenting classes might play in supporting parenting readiness, competence, and skills, they also suggest that the research literature on parenting classes and their impact is neither overwhelming nor clear cut. Nonetheless, the Commonwealth Survey suggests that participation in parenting classes was associated with increased readiness for parenting and much higher rates of breastfeeding, regular routines, and less coercive forms of discipline. As might be expected, a number of risk factors for parenting readiness emerged, including low education, minority ethnic status, single parenthood, negative family experiences including reporting of past abuse, and

382

Halfon, McLearn

young maternal age. Particularly interesting was the finding that confidence and ability to parent were inversely related to education. As the contributors suggest, this may in fact reflect an association between more education and greater recognition of limitations. The chapter also pointed to the effect of parents’ childhood experiences on both the timing of the decision to become parents and their own parenting behaviors. While much has been made over the last 15 years of the increased need for new mothers to re-enter the workforce soon after their child’s birth, data from the survey demonstrates a more complex and nuanced picture. Parental employment patterns are highly associated with the age of the child. Among two parent households, at least one parent is home in 44 percent of households where the child is less than 6 months of age. As the child’s age increases this number falls by age two to 34 percent. Two parent families report both parents working full-time at a rate similar (34 percent) to that of single parents reporting full-time work (30 percent). Of interest is the fact that having one parent who is not employed and at home providing child care is less common among two parent families (40 percent) than among single parent families (57 percent). Given the fact that welfare reform was being implemented at the time that this data was collected, we would imagine that over the intervening years an even larger proportion of parents has been engaged in the workforce. Given workplace time commitments, the desire for more time with their children was high among both mothers and fathers. Only 37 percent of parents felt they were spending about the right amount of time with their children, and 57 percent reported they would like to spend more time with their children. Overall, these findings suggest a strong trend in two parent families toward desiring more time with their children, even among those who are primary care givers and not burdened with full-time employment. These findings suggest that parents rearing young children have a number of competing responsibilities. The study also suggests that wealthier families express greater satisfaction with the time spent with their children, perhaps because they are able to afford a range of other services that enable them to have more quality time with their children. These analyses also suggest that employers need to recognize that employees with young children, including fathers, are often going home to substantial child care responsibilities. Over the past two decades, paternal engagement with children has increased, and a father’s involvement with his children usually increases as children age (Furstenburg 1988). The literature also suggests that a father’s involvement can have a positive impact on a child’s cognitive and social emotional development (Parke 1995). Increased paternal parenting involvement can also be very beneficial to the quality of a marital relationship and to the reported stability of commitment by both parents (Parke 1995; Lamb et al. 1985; Belsky et al. 1984). The Commonwealth data show that

Families with Children Under 3: Policy Implications

383

today’s fathers are participating in the care giving of their children even in the very early years. Over a third of parents (36 percent) reported that they shared responsibility equally with their spouse or partner. Importantly, employment patterns affected care giving patterns. When parents were both working outside the home, higher amounts of shared care giving was reported. This analysis also suggests that shared care giving arrangements are more common in families with lower socioeconomic status, where the ability to hire outside help or pay for outside care is more limited. Even so, and despite the growth in shared child-rearing responsibilities, it is still the norm for the mother to be the primary care giver, regardless of employment status. How parents are using their available time with their children also varies. In examining a number of specific behaviors, contributors found that 39 percent of parents reported reading or looking at a book with their child on a daily basis, 59 percent reported singing and playing music, and 84 percent reported actively engaging in daily play with their child. In addition to these unquestionably positive behaviors, analysis of parental discipline practices indicated that although routine spanking is relatively uncommon in children less than 3 years of age, about 10 percent of children ages 6 to 11 months have reportedly been spanked; by 18 to 23 months, about 59 percent of parents said they had spanked their child at least once. As Wissow explains in his chapter, while normative views in the professional pediatric community are mixed concerning the use of spanking for older children, few if any pediatricians or child psychologists would recommend spanking for children this young, and there is no research suggesting that children under 3 are capable of deriving any cognitive, emotional, or behavioral advantage or “lessons” from physical discipline. Consequently, this finding in itself is powerfully indicative of a significant lack of knowledge among parents about how children learn, what they are capable of learning at different ages, and perhaps their unfamiliarity with effective alternative approaches. Doubtless, it also indicates that parents themselves are under a great deal of stress, as will be discussed later. The survey provided unique information about family and other support structures available to young families. Surprisingly, nearly 80 percent of parents report grandparents living within an hour’s drive from their home, and 70 percent report having many friends and relatives they can count on. These findings seem to refute popular accounts of high geographical mobility and concomitant disconnection from family and friends. Reports about available resources and information about parenting indicate that parents are more likely to seek advice on parenting from family members than from health professionals, and a substantial proportion use books, magazines, newspapers, and videos as sources of parenting advice. There were clear ethnic distinctions, with African Americans and Hispanics less likely to report using media resources. Only about a third of

384

Halfon, McLearn

parents report having attended a class or discussion group about parenting. The analysis by Britto and colleagues that examines the relationship between attending a parenting class and parental behaviors such as reading and organizing routines indicates that parenting classes seem to mediate some of the negative effects of lower educational status. Again, attending parenting classes is much more common among married, higher income, and higher educated families. These analyses suggest significant gaps and substantial opportunities to provide additional child-rearing and parenting support for young families, especially those with fewer resources. The lack of reliance on health professionals for information and support indicates that health professionals could be doing more to engender and encourage trust and to cultivate their roles as advisors on parenting and child rearing. The fact that higher income and more educated parents are availing themselves of additional parenting supports in the form of parenting classes suggests that such classes are likely to be useful to families with fewer resources if made available. How Are Parents Rearing Their Young Children? While there is a large popular and scientific literature on the role that different parenting styles, behaviors, and practices play in the lives of young children, there are few data to document what parents actually do with their children. One unique aspect of the data is the insight provided into certain “sentinel parenting behaviors and practices” that are potential indicators of the quality of family life and functioning during the early years of a child’s life. We sought, as sentinel indicators, behaviors that are symbolic or representative of parenting styles, approaches, and values. These sentinel behaviors and practices include the time spent by parents reading, singing, and engaging in regular routines with their children, breastfeeding, and disciplinary practices that reflect parenting approaches to the social and emotional development of their child. Each of these behaviors is indicative of many other decisions and behaviors. For example, in order for a mother to breastfeed successfully, a range of conditions must be met, including a supportive spouse, a supportive workplace, and a strong desire to use this method of feeding in spite of other barriers from different quarters. In addition to these behaviors, the psychological status of parents based on their reporting on the prevalence of depressive symptoms is also included as an important indicator of parenting capacity and quality of family life. In describing each of these behaviors and child-rearing practices, it becomes clear that other risk factors, including income, family structure, parental maturity, employment status, and time availability, have an effect on each. In the chapter by Britto, Fuligni, and Brooks-Gunn, the authors suggest that parental activities such as reading or book sharing, rhymes (playing, hugging, and cuddling), and routines are the three “R”s of early education

Families with Children Under 3: Policy Implications

385

and are fundamental to early childhood development. These parenting behaviors provide important opportunities to stimulate cognitive, social, physical, and emotional development in children, yet the authors report that many parents miss opportunities to read, play, and stimulate their child’s development. Only 37 percent of parents reported reading or early book sharing with their infant or toddler on a daily basis. The authors report that parents with toddlers (1 to 3 years of age) are more likely to engage in book sharing (45 percent), but that very few parents (22 percent) report reading to their infants. The survey also suggests that physicians can make a difference in the sentinel behavior of parental reading or book sharing with an infant or toddler. The findings reveal that parents who speak with their pediatric clinician about encouraging their child to learn are more likely to read to their child on a daily basis than are parents who do not discuss learning with their child’s physician. Also, as stated in the Zuckerman and Parker chapter, an evaluation of the Reach Out and Read program shows that when physicians give a young child an age appropriate book during a primary care visit, they communicate the importance of early book sharing and increase the likelihood that parents will read to their children (Golova et al. 1999; High et al. 1998; High et al. 2000). Interestingly, parents follow the opposite pattern in playing with their children, with parents reporting that they were more likely to play during a child’s infancy (85 percent) than during toddlerhood (78 percent). The fact that more parents enter the workforce or return to work as their child reaches 2 years may indicate less time availability and less ability to play. Nonetheless, parents seem to be reading less to their infants and playing less with their toddlers than might be expected or desired. As has been reported in other studies, the Commonwealth Survey confirmed that older parents, parents who are more educated, and those with more resources report higher rates of reading, singing, and hugging (Bornstein 1995). The importance of these observations needs to be considered in light of ongoing longitudinal research that has examined the potential role of early childhood behaviors and routines, such as reading, on longer term social, emotional, and cognitive functioning (Whitehurst and Lonigan 1998). Researchers examining disparities in high school achievement tests have noted that a significant proportion of the variance observed can be attributed to differences that already exist when a child enters school (Phillips et al. 1998). Using data from the National Longitudinal Study of Youth, researchers have documented the observation that these differences in achievement are associated with the amount of time parents spend reading to their child (Phillips et al. 1998). The second sentinel behavior – providing predictable daily routines for the young child – was also examined in this volume. Parents who are able to provide consistent routines must make significant efforts to structure other family, work, and social obligations around a core and consistent set

386

Halfon, McLearn

of child focused care giving activities. Unfortunately, only slightly more than 51 percent of parents reported having daily routines for naps, meals, and bedtime for their 1- and 2-year-olds. The authors found that parents were more likely to have daily routines with their 1-year olds than their 2-year olds and that two-parent families and parents with more than a high school education were more likely than single parents and those with less than a high school education to establish daily routines. Two important factors appear to contribute to these positive, prodevelopmental behaviors. The first is positive mental health, and the second is parents’ participation in parenting classes. The role of parental mental health was captured by parents’ reports of depressive symptoms. Parents reporting high levels of depressive symptoms were much less likely to engage in pro-developmental behaviors and routines. Similarly, parents who report having attended parenting classes reported much higher levels of these behaviors. What cannot be determined from this cross-sectional data is whether parents who attend such classes were predisposed to engage in these behaviors. Parents in a more optimal parenting trajectory may be more likely to take parenting classes, to have enough time and be sufficiently organized to participate in a parenting class, and to possess the necessary time and organization skills to create more explicit child rearing routines. The chapter by Slusser and Lange reports that breastfeeding has been associated with a range of positive short- and long-term health, emotional, and cognitive outcomes in the child and that breastfeeding trends are on the rise. This study confirmed previous findings that despite increased breastfeeding rates, single mothers, less educated mothers, and mothers from low income and minority families are much less likely to breastfeed. Slusser and Lange also indicate that, although initiation rates are high, the duration of breastfeeding falls far short of the established recommendation that breastfeeding continue through the first year of life (Gartner et al. 1997). Moreover, the chapter indicates a very strong association between parents receiving either childbirth education classes or health care provider encouragement to breastfeed with initiation rates. These positive effects on the initiation of breastfeeding do not seem to influence the duration of breastfeeding. The contributors raise an important issue regarding the impact of employment and returning to work on mothers’ ability to continue to breastfeed. If duration of breastfeeding is to be increased, a more sustained set of reinforcers, potentially provided by health care providers, must be coupled with reducing the barriers to breastfeeding that current employment related practices and lack of support represent. At present, most workplaces are not supportive of the breastfeeding mother (Moore and Jansa 1987; Danyliw 1997). Discipline is the fourth sentinel behavior analyzed in this volume. While the Commonwealth Survey documents routine spanking as relatively

Families with Children Under 3: Policy Implications

387

uncommon among parents with infants and toddlers, physical punishment of very young children by mothers and fathers nonetheless does exist. The prevalence of having ever spanked increased rapidly from 6 to 11 months, when about 10 percent of children were reportedly spanked, to 18 to 23 months, when 59 percent of children had been spanked. Parental frustration was the factor most associated with spanking, and this association was stronger than income or education. Nonetheless, spanking was more common among single, low-income parents and especially those parents with more depressive symptoms. These same parents are less likely to read to their child and use non-physical forms of punishment. Patterns of spanking were related to other forms of punishment and interaction depending on the characteristics of the parent. For example, women who typically spank their child also had higher levels of other measured forms of interaction, except hitting. Men who typically spank their child have a high prevalence of non-physical punishment (such as timeout and taking things away) but are not as likely to be engaged in reading, music, and play with their child. Interestingly, those who were least likely to report spanking their child were also least likely to report either positive or negative interactions with their child. This suggests that efforts to encourage active development by promoting interaction between parents and children must include a strong focus on managing discipline and setting limits, so that the level and intensity of negative interactions do not increase along with positive ones. Wissow also highlighted the role that health professionals currently play and the potential role they could play in providing information about and support for appropriate discipline practices. While only slightly more than half of parents in this survey (58 percent) reported that they felt they had enough information about discipline, many parents also reported that health professionals were not a common source of information about child discipline. However, parents’ ratings of satisfaction with their child’s physician was positively associated with having sought information on discipline. Parents who reported that their child’s physician did an “excellent” job of listening carefully to questions and helping them understand their child’s growth and development were more likely to receive advice from their provider about discipline as well. Nonetheless, these results suggest that there is a significant opportunity for pediatric clinicians to provide additional information about discipline in a sensitive, respectful, and family-centered manner. For each of the sentinel behaviors, the behavioral output of parents is related to age of the child or developmental level and context of family life. Infants are more likely to be played with, less likely to be read to, more likely to have regular routines, and less likely to experience negative discipline. Toddlers elicit their own pattern of parental behaviors. Moreover, the internal state of the parent measured by self reporting

388

Halfon, McLearn

of frustration and depression – as discussed below – interacts with and accentuates these patterns. Income, education, employment, and other family relationships are important predictors of these sentinel behaviors and practices. In almost all cases, there is a strong association between the receipt of additional support or guidance through formal mechanisms (such as parenting classes or anticipatory guidance by a health provider) and a more positive behavior or parenting practice (such as breastfeeding or book sharing). Even so, the cross-sectional nature of the data makes it impossible to determine the true role of parenting classes and other instrumental supports in determining these behaviors. How Are Parents Faring? As the discussion of sentinel behaviors suggests, an important area of parental function is the presence of depressive symptoms. Epidemiologic data documenting increased rates of clinical depression and studies about parental depression impact on the development of young children have highlighted the importance of depression (Lyons-Ruth et al. 1990; Orr and Miller 1995; Field 1995). As contributors in this volume suggest, depressive symptoms become important indicators of the degree of imbalance between parents’ psychological coping resources and the sources of psychological stress and their ability to respond appropriately to their child’s needs. Even though a limited inventory of depressive symptoms was used for this survey, Lyons-Ruth and colleagues explain that the results are comparable to those of other assessments of depressive symptoms in large populations. Depressive symptoms in this analysis are shown to be associated with less positive and more negative child-rearing behaviors. As the contributors point out, the popular stereotype of depression and the associated picture of a parent who might be sad and withdrawn does not accurately reflect current understanding of the manifestation of depression in parents (Field 1995). The Commonwealth Survey documented in a national sample the observation that depressive symptoms in both mothers and fathers can affect their ability to parent effectively in two ways. Parents reporting more depressive symptoms are not only sad, withdrawn, and less likely to engage in the sentinel childrearing behaviors of reading, playing with, and hugging a child and establishing daily routines, but are also more likely to be irritable and critical with their children and to use coercive or intrusive methods of discipline. These parents reported greater feelings of aggravation and frustration with their young child’s behavior, resulting in more yelling, spanking, and hitting than among parents who did not report depressive symptoms. Other researchers have highlighted the link between

Families with Children Under 3: Policy Implications

389

maternal depression and maternal-childaggression (Zuravin 1989; Osofsky and Thompson 2000). An important and unusual finding was that primary child care responsibilities could constitute a risk factor for parental depression. Taking care of young children is a “high demand-low control” job that seems to be associated with increased rates of depression regardless of gender. While reported depression rates in mothers are often higher than fathers’ rates, reports from primary caregiving fathers suggest that their depression rates are similar to those of mothers with primary child-rearing responsibility. Expressions of frustration with parenting and greater difficulty in setting up and following through with child care routines are often manifestations of depressive symptoms and important indicators that a parent is having a problem. As this chapter points out, a number of other studies have suggested that depressive symptoms and other parental self-reported personality characteristics are less powerful predictors of parenting behaviors than parents’ descriptions of the parenting they received as a child. Several contributors (such as Lyons-Ruth et al. and Zuckerman and Parker) in this volume suggest a role for obstetric and pediatric clinicians in addressing maternal mental health. These clinicians are in constant contact with young parents and are natural points for increased delivery of services to parents or referrals to other health services or parenting supports. As Lyons-Ruth suggests, “simple screening procedures for depression among expectant mothers and parents of young children constitute an easy-toimplement first level intervention to identify depressed parents and provide or refer for further services.” In a related analysis of coping, confidence, and frustration experienced by parents, the contributors find that white parents report higher levels of frustration but lower levels of depression than African American or Hispanic parents. They also suggest that parents with higher incomes report feeling much more satisfied with the amount of time they spend with their children. Parental expressions of frustration appear to be greatest with boys and increase as the child ages from birth to five years. Higher levels of frustration are reported by parents who also report higher numbers of depressive symptoms. Parenting information and other supports are not equally distributed among families. Low-income parents are less likely to receive parenting information from hospitals, and parental education was associated with obtaining information, attending parenting classes, and consulting with other parenting related resources. What Role Is the Health Care System Playing? The health care system, which provides obstetrical and pediatric care, could play a unique and important role in supporting child and parental development. Under current professional standards, parents should have access

390

Halfon, McLearn

to a comprehensive set of continuous services capable of supporting the family planning, pregnancy, and birth processes and of providing adequate preparation and assistance to optimize parenting behaviors and child development (Green and Palfrey 2000). Nonetheless, as we know, the American health care system has not achieved this norm. Some women in the United States continue to receive late prenatal care owing to lack of health insurance and other factors, while many American children are either uninsured or have limited overall access to appropriate care (Newacheck et al. 2000). It must also be noted that access to health care does not guarantee high-quality pediatric care or a benefits package that includes the comprehensive services parents say they want, such as home visiting, developmental screening, and extended time for counseling and anticipatory guidance. The chapter on prenatal care by Paul Wise confirms what similar analyses have indicated: Most women are receiving prenatal care in the first trimester, but disparities between ethnic and income groups continue to exist. Wise also shows that health insurance is associated with the early initiation of prenatal care and that lack of health insurance is associated with a three-fold increase in late initiation of care. Women who are poor, less educated, and from African American and Hispanic ethnic groups are much more likely to report late initiation of prenatal care. The same social demographic variables are associated with rates of readmission to hospital for newborns and breastfeeding initiation. Wise concludes that the experience of childbirth in the United States reflects the collective social distinctions of daily life in America today. While the vast majority of American women start prenatal care early, social disparities in utilization are associated with the relatively large number of premature, low birthweight infants born to low income and African American families. Newacheck and colleagues, using both the Commonwealth data and data from the 1995 National Health Interview Survey, demonstrate the role that health insurance plays in promoting and providing access to pediatric care for young children and describe social disparities in the receipt of care based on income, minority status, and insurance status. Most children under age 3 are insured and have reasonable access to basic pediatric care. Indeed, nearly 22 percent of children in this age group are insured by the Medicaid program. While this analysis also indicates that most parents of young children are able to report a usual source of care and relatively good access and satisfaction, parents of uninsured children from minority families and low-income families report an inability to get needed medical care, dental care, child development services, and mental health care. The disparities in access reported by Newacheck suggest that barriers to access to health care remain for these children: in some cases, there are twoand three-fold differences in relative access when compared with white, non-poor, insured children. The contributors suggest two mechanisms to

Families with Children Under 3: Policy Implications

391

overcome some of these disparities. First, they propose changes in reimbursement and the further expansion of health insurance to ensure that all children have adequate coverage. Second, they suggest greater use of specific performance measures to guarantee that children are receiving recommended services as set forth in current guidelines. Performance measurement is increasingly used by employers and state Medicaid agencies to measure health care quality and to ensure that their enrollees are receiving appropriate health care. In their chapter on the doctor-patient relationship and the role of anticipatory guidance, Schuster, Regalado, and colleagues highlight the potential for pediatricians to optimize development by helping parents manage typical child-rearing challenges and by helping to identify a range of potential developmental and emotional problems in a timely fashion. Their analysis and Newacheck’s, however, report that most parents are not discussing child rearing with their doctors despite a stated interest among a majority of parents in receiving child development and parenting information. While nearly three-quarters of parents have not discussed of learning, discipline, or toilet training with their doctors, nearly a third to a half of all parents state they want more information about these nonmedical topics. Of particular note is the unmet need for information on how to promote children’s learning, as shown by strong parental interest in this topic. This chapter also explains that parents who might have benefited most from discussions about child rearing were least likely to have had those discussions, since lower socioeconomic status and less access to care were associated with fewer child-rearing information transactions. Of special interest is the fact that parents who received parenting information and comprehensive pediatric services such as home visits or a telephone advice line reported significantly higher levels of satisfaction with their child’s physician than did parents who did not receive these enhanced services but said they would be willing to pay for such services. Specifically, more than twothirds of parents surveyed expressed a willingness to pay an additional $10 per visit to have the opportunity to discuss child-rearing topics and receive child development services, thus indicating the value they attach to these topics. The authors suggest a variety of reasons why parents are not receiving the help they need, focusing principally on the current structure and organization of pediatric practice, reimbursement and managed care trends, and physicians’ lack of training in how to provide developmental and psychosocial services to families with young children. In their chapter, Zuckerman and Parker expand on some of those criticisms and concerns about the current state of pediatrics. The potential role of pediatric services in the lives of families with young children is explored, drawing on the personal experiences of the authors and on their involvement in a national initiative to alter the delivery of early childhood pediatric care. Reflecting their appreciation for the impact of social and

392

Halfon, McLearn

family context on child health and development, the authors argue for a comprehensive, family-centered approach to pediatric care – one in which child development services are embedded in a two-generation approach and that focuses attention on parents’ health and developmental needs as well as those of the child. Highlighting the growing recognition that maternal health and health behaviors are significantly related to the child’s health, development, and health services utilization, the contributors suggest how care for mother and child might be more closely linked. They also propose the further expansion and adoption of the Healthy Steps model of pediatric care services, which includes a much greater focus on the psychosocial context of the family with infants and toddlers. The Healthy Steps model creates an early child development infrastructure and community linkages that enhance and expand pediatric care to better meet the needs of mothers and fathers. To accomplish this, a new mid-level professional – an expert in child development – is added to the pediatric team. This person works with the physician and serves as the primary resource for families to deliver the type of care parents say they want and pediatricians say they would like to provide.

interpretation and implications Several important issues emerge from the presentation in this volume of new information about the lives of families with young children and the role of the health care system in their lives. It is clear that many families are experiencing significant stresses, owing to their life circumstances and the demands of their roles as parents and workers. Many families with young children have too little time, too many demands, and insufficient resources. Nearly half are relying on government resources to make ends meet, yet some still report having difficulty paying for basic necessities. For families with less income, just one parent, or parents who are young or depressed, the parenting risk increases, although these parents are not the only ones experiencing frustration with their ability to respond to the responsibilities of parenting their young children. Indeed, a significant portion of families with young children report being overwhelmed. Yet it is also the case that their situation is by no means the norm: Many families reported patterns of childrearing that reflect adequate time and resources. From a public health and public policy perspective, survey results regarding patterns of child rearing and the demands of parenting suggest that much more can be done to ease parents’ burden, address some of the problems they face, and provide basic services to many more families. Children’s health providers and programs (such as child care and WIC) that are frequently in contact with young children and parents could be screening parents universally to identify needs and target appropriate interventions, especially regarding their preparation for parenting, their time

Families with Children Under 3: Policy Implications

393

and resources, and competing demands in their lives. Most parents with young children juggle many competing responsibilities, the most significant of which are placed upon them by their employment outside the home. Associated with insufficient resources and unmanageable time demands are feelings of lack of control and confidence, depressive symptoms, and negative child-rearing behaviors and practices. What are the correctives that might permit working families with young children to receive the support they need and want? To what extent can employers be more responsive to the needs of families with young children? How can employer self-interest be harnessed by persuasion and public policy to recognize those issues? Should greater support for families come from public sector programs, subsidized childcare, or expanded health insurance benefits? For example, the Washington Business Group on Health has described how some far-sighted companies have recognized and supported the needs of parents through support for breastfeeding and child care arrangements. Various proposals for extending the period of parental leave following child birth have also been suggested. Large corporations with highly educated workforces represent one important model, but the needs may be greatest among families working for smaller companies, which tend to pay lower wages and lack the resources of Fortune 500 companies. Public policy solutions that rely on employment-related support for families with young children must contend with this reality, too. Parental Mental Health As the results from the Commonwealth Survey and other studies indicate, parental mental health is an important issue for children. Although major mental illness affects relatively few parents, the high prevalence of mild and moderate depressive symptoms, and the association of those symptoms with feelings of frustration, difficulties in parenting, and negative behaviors are well documented here. These problems are significantly greater for parents who are poor and have less support, and they seem more situational than endogenous. For example, the survey results indicate that fathers in a principal caregiving role have depressive symptoms at rates similar to those reported by mothers in comparable roles. Numerous recent studies have highlighted the impact of parental depression on children’s development and the compounding effect that poverty can have in exacerbating the impact of maternal depression in children’s outcomes (Graham and Esterbrooks 2000). Parental mental health problems can lead to disturbances in parent-child interactions, and the strategies that a young child adopts to relate to a mentally distressed parent can become a persistent, resistant, long-term behavioral pattern of response (Field 1995; Field et al. 1990; Radke-Yarrow et al. 1985). Given the prevalence and potential impact of depression on the development of

394

Halfon, McLearn

young children, the findings have profound implications for interventions intended to improve primary, secondary, and tertiary prevention strategies. How preventable depression is, how responsive it is to early identification, and the impact of providing additional support through early intervention services have not been well defined (Field 1995). As suggested in the chapters by Zuckerman and Parker and by Lyons-Ruth and colleagues, a more continuous screening process that was embedded in the provision of prenatal, post-natal, and pediatric health care might be useful in identifying depressive symptoms and providing interventions that could potentially provide support and reduce the demands that place parents in a compromised situation. As the responses of the surveyed parents indicate, parents are not receiving all the support they may need, either from the health care system, their immediate families, or their communities. The chapters on the health care system clearly indicate that there is a significant segment of the population that is either uninsured or underinsured and/or is not receiving the kind of child-rearing guidance that they would find useful. These findings suggest that simplistic policy nostrums involving the recreation of three generation traditional families may not be in accord with the wishes or interests of typical families. The complexity of these findings suggest the need to listen very carefully to what families are really saying about the supports they have, and those they would like to have. Throughout the chapters, the contributors found significant disparities in preparation for parenting, types of parenting behavior, and use and availability of resources. These disparities were associated with traditional sociodemographic risk factors, such as family income, family structure, and race and ethnicity. These findings reaffirm the results of more focused intervention studies documenting the needs of families with relatively few resources. If parenting support and intervention are to be targeted, then these are the families that potentially stand to gain the most (Karoly 1997). Even so, in this case and others, it is vitally important that the provision of intensive services to families in the greatest need be linked to a more limited but similarly structured, financed, and delivered set of services for all families with young children. A more universal approach to the needs of families with young children avoids stigmatization of families receiving intensive services and breaks the disparity-inducing fragmentation of services into “welfare” and “middle class” services. Similar approaches are now being implemented in several other countries. A policy to promote the universal availability of childbirth classes and other parenting supports provides a good example. Childbirth and other parenting classes seem to make a difference in parental function and parents’ tendency to adopt certain behaviors and routines. Families with more education and income were more likely to avail themselves of these classes, thus making it difficult to untangle the role of motivation and other

Families with Children Under 3: Policy Implications

395

behavioral characteristics in choosing to participate in classes and the associated behaviors. Nonetheless, it appears that these interventions hold promise as potentially important mechanisms for providing support, especially to individuals and families not currently receiving these services. Yet several decades of providing early intervention services to high-risk families also seems to indicate that many families with multiple needs require comprehensive, substantial interventions that are embedded in a continuum of local service networks (Olds 1997; Musick 2000; Halpern 1990; Halpern 2000). Enthusiasm for the role that childbirth and other parenting classes might play must also be tempered by an understanding that, for many lower income families, parenting classes may be stigmatized because of their association with remedial services provided to families suspected of child abuse and neglect. Making such classes more widely available is therefore not necessarily just an issue of payment but of addressing other potential barriers that may inhibit their utility in certain communities. Making such services available through private and public mechanisms for all who choose to use them could, like public education, provide all parents with the opportunity to become as knowledgeable in their parenting roles as modern parenting demands. Children’s Long-Term Development The long reach of early childhood is increasingly recognized (Wadsworth 1999; Keating Hertzman 1999; Halfon, Inkelas, and Hochstein 2000). Research on early child and brain development suggests that early childhood experiences are important for determining long-term behavioral and developmental trajectories of children (Shonkoff and Meisels 2000) and suggests why so much is at stake in the first three years of a child’s life. Previous research on parenting and the parenting process has shown that just as children develop, so also learning to be a parent is a developmental process, and that different parents develop different capacities and abilities on the job and in the midst of a complex system of relationships within families (Cowan 1992; Heinicke 1995; Belsky et al. 1984). However, while one might imagine that new parents learn their parenting skills on the job, several lines of research indicate that the parent-child relationship is not simply constructed as it goes along but is significantly preconditioned by aspects of the parent’s previous adaptation and overall developmental trajectory (Lyons-Ruth and Zeanah 1993; Heinicke 1995). Several researchers have shown that the mother’s and father’s psychological adaptation and the quality of their marriage before the birth of the first child predict the quality of their parenting during the early years. Studies on parental satisfaction indicate that marriage, job, and other social roles have a striking influence on the parenting role, experience, and

396

Halfon, McLearn

satisfaction (Cowan and Cowan 1992). Other studies have highlighted the transmission of parental behavior and experience across the generations and, as this study indicates, the presence of clear relationships between parental reporting of childhood abuse and other experiences and the desire to rear their children as they were reared. Studies over the past two decades have also emphasized the role of the attachment relationship and the parent’s internal representations of relationships in how that parent responds to the child (Zeanah 1993). Parents with difficult family histories cannot help but bring those histories to the parenting relationship (Fraiberg 1980). A range of intervention studies over the past two decades have attempted to understand, explain, and alter this set of relationships (Shonkoff and Meisels 2000). The results portrayed in this volume show the variability in resources and capacity that parents bring to the parenting relationship and the ability to respond appropriately to the needs of a young child. Nonetheless, an age-old adage among pediatricians states that parenting is a great leveler: All parents want to know how their child is doing, and all parents want to know how they are doing in relationship to the child. These basic impulses and desires serve as openings to discussion with parents about what their needs might be and what assistance is available. Over the past three decades, a number of programs and interventions have been developed to improve the capacity of parents to respond appropriately to their young children, including home-based and center-based early intervention programs, family resource centers, expanded health care models, and other day care interventions (Shonkoff and Phillips 2000). Consistent conclusions emerge from much of this research. Early intervention programs that target high-risk, multiple-needs families and involve the family as a unit appear to be very effective in enhancing children’s health status, motivation, and general social competence (Karoly et al. 1997; Halpern 2000). A growing number of family-focused and communitybased interventions are built on an understanding of the inter-relatedness of systems that influence child development: families, schools, health care, and child care (Bronfenbrenner and Morris 1998; Lerner et al. 1995; Schor 1995). Much of this research has also suggested effective strategies to support and improve parenting. The potential effect of interventions in early childhood relies on improving the ecological environment in which parenting takes place. A number of interventions launched from community-based or health system-based service platforms have shown important effects in modifying child, family, and parenting environments (Karoly et al. 1997; Halpern 2000; Furan 2000; Brooks-Gunn et al. 2000). Nonetheless, such programs remain rare, not for lack of substantive knowledge but from our apparent inability to translate promising research findings into more universal public policy responses.

Families with Children Under 3: Policy Implications

397

Health Care and Other Social Policies One of the important conclusions of this volume is that the health care system is not providing the range and depth of services that many families with young children say they want. Although the health care system is potentially an integral part of the lives of all families through the provision of prenatal, birth, and pediatric services, many families are not connected with the system and are not receiving the help and support they want and need. Indeed, given the large number of routine contacts with the health care system that many families report, it appears that health care providers are doing far less than they might to support families with young children. As Zuckerman and Parker point out, for example, pediatric practice has defined itself by its focus on anticipatory guidance regarding developmental issues. They further explain that pediatric practice has not changed much in the past thirty to forty years despite rather significant changes in the social ecology of families with young children, the problems and exposures that children experience, and what parents say they need and want from the health system. For pediatric care to be more responsive to this changing set of needs, it would need to be reoriented and reorganized around prevention and health promotion principles consistent with a more ecological approach to health development (Halfon et al. 2000, McLearn et al. 1998). A number of disincentives keep pediatric providers from offering an expanded set of family support and developmental services. These include inadequate and cumbersome reimbursement and administrative mechanisms; lack of familiarity with appropriate assessment and care coordination tools; a dearth of treatment and support services in the community to address psychosocial, behavioral, and developmental needs; and lack of training that would enable pediatricians to assess family issues and provide appropriate referrals (Halfon et al. 2000). At a practice level, programs such as Healthy Steps hold the promise of reorienting and organizing pediatric primary care. Zuckerman and Parker describe the Healthy Steps model in this volume, including its use of an expanded team of personnel and a new provider, the child development specialist, to address the psychosocial and developmental needs of parents with young children. Services provided under Healthy Steps include additional anticipatory guidance, developmental assessments, home visits, referrals to community services, and other services that families need and are not routinely receiving. At present, there are more than twenty-five Healthy Steps practices throughout the United States, and major national impact evaluation is being conducted by the School of Public Health at Johns Hopkins University. Early findings indicate that Healthy Steps is having an impact. Healthy Steps intervention families, as compared to control families, were receiving a higher quality of pediatric care (Minkovitz et al. 2001). Intervention

398

Halfon, McLearn

families were more likely to have received home visits and other child development services and almost twice as likely to have had someone in the practice discuss child-rearing topics with them. Healthy Steps intervention families were more than twice as likely to be satisfied with the care provided at their baby’s pediatric practice, and parents also appear to be developing stronger links with their pediatric providers than were control families. Healthy Steps also seems to be having an effect on parent practices. For example, despite a major media “back to sleep” infant safety campaign to prevent sudden infant death syndrome (SIDS), control group parents were significantly more likely than Healthy Steps parents to place their babies on their stomachs to sleep. These preliminary findings suggest that Healthy Steps may represent an effective set of new strategies to improve the effectiveness of pediatric care to families with young children. At the community level, there is also a need for a mechanism to help pediatric providers refer families to appropriate services after identifying developmental or psychosocial problems during a comprehensive assessment. ChildServ, established in 1997 in Hartford, Connecticut, is a good example of a community-based program that helps private-practice pediatricians connect their patients with appropriate providers and programs. ChildServ also helps pediatricians overcome their own reticence to perform developmental and psychosocial assessments in practice environments where referral services might not be readily available. The program is built on case management and coordination systems that were already in place for children with special health care needs. Some health plans are also attempting to improve the provision of psychosocial and developmental services to families with young children. Children Now, a national child advocacy organization, has identified the provision of those services and the potential role of managed care plans in their Managed Care and Early Childhood Development Initiative. Children Now has worked with several managed care organizations, including the nonprofit Kaiser HMO system, to consider how to offer a fuller menu and a more responsive set of developmental services to families with young children. The Foundation for Accountability (FACCT), a national consumerfocused quality assessment organization, is also working with managed care organizations by developing a survey that enables managed care organizations to assess the quality of services provided by pediatric providers to families with young children. The Promoting Healthy Development Survey designed by FACCT queries parents in seven domains to determine whether they are receiving appropriate anticipatory guidance, family-centered care, essential information on pediatric care, and developmental information and services (Bethell et al. 2001). Advocacy efforts such as the one developed by Children Now and accountability projects

Families with Children Under 3: Policy Implications

399

such as FACCT’s could begin to influence how managed care organizations provide services to families with young children. In order to improve the delivery of developmentally focused services in pediatric practice, it is important for pediatric providers to be appropriately trained, to have the use of effective developmental and psychosocial screening tools, and to have community-based resources available for referrals. The Pediatric Residency Review Committee of the Accreditation Council on Graduate Medical Education requires residency programs to incorporate behavioral and developmental pediatrics into ambulatory experiences and include a one-month block rotation focused on developmental pediatrics. To accomplish those goals, greater emphasis on family development and psychosocial issues could be included as part of required training programs. Using curricula such as Healthy Steps, providers of pediatric continuing education have also begun to revamp their programs and improve their knowledge. Quality improvement processes developed for other types of medical services could be easily adapted to developmental services for families with young children. A dedicated quality improvement initiative focused on promoting healthy development using the Healthy Steps approach is feasible, although it would demand the ingenuity of leaders in the quality improvement field, the political will of managed care organizations, and the combined efforts of advocacy, consumer accountability, and professional organizations to move the agenda forward. Another major barrier to the provision of developmental health services is inadequate reimbursement. Several mechanisms exist to ensure reimbursement for developmental services such as prenatal birth classes, postnatal parenting classes, and certain primary care services, including home visits, developmental assessments, and referrals. Reimbursement for these services can be facilitated for families covered under Medicaid or the State Child Health Insurance Program (SCHIP) through contracts with managed care organizations. Now that 39 percent of all births are covered by Medicaid and 36 percent of young children are covered by Medicaid and SCHIP, a significant portion of the child population is insured under these two federal programs. Several states have already included, as part of their routine Medicaid-contracting mechanisms, provisions to reimburse for parenting classes. In addition to covering these essential services, Medicaid contracts should allow providers to bill for appropriate services. By improving reimbursement, mandating coverage of developmentally focused family services, and improving training, the provision of developmental services in pediatric office-based settings could be dramatically improved. Beyond the pediatric office, in the community health setting, a number of community-wide approaches have been initiated throughout the United

400

Halfon, McLearn

States to improve the availability of a full continuum of early childhood services. One of the best examples is in Monroe County, New York, where the county health department has negotiated a “master contract” with the State of New York under which the state provides what amounts to a block grant for eight different federal funding streams targeted at early childhood services, such as immunization, lead screening, developmental services, and family support services. Its new-found fiscal flexibility has permitted Monroe County to embark on a total redesign of its community-based early childhood health services. The county now offers a single point of entry for eligible families, a full continuum of early childhood health and developmental services, more integrated case management, and coordination with private pediatric practices. The Monroe County experience is just one example of the ways in which communities can reengineer existing early childhood and health systems into more responsive local networks of child and family health and development services. Community-Level Policies In considering approaches that a local community may wish to embark on in order to provide better support to families with young children, a number of strategies are possible. Several build on community-wide initiatives such as the Carnegie Corporation’s Starting Points, Vermont’s Success by Six, Minnesota’s Early Childhood Family Education, and North Carolina’s Smart Start. Any community-wide initiative must try to improve the awareness and knowledge of all stakeholders – business and community leaders, service providers, educators, parents, employers, and policymakers – about the needs and capacities of families with young children. Much of the information contained in this volume about the importance of early childhood and the challenges families face may be unfamiliar to decision makers and community leaders. One step that communities can take is to assemble their own local portrait of families with young children and report that information in a regular manner. Communities around the country have begun to create community-wide health reports, which could easily incorporate indicators about the functioning of families with young children (Halfon et al. 1998; Fielding et al. 1999). In addition to getting the word out about the situation of families, it would also be important to assess the status of available programs and supports within the community. A comprehensive community needs assessment should include traditional and non-traditional sources of support for families with young children, such as infant and toddler child care, Head Start and Early Head Start services, cooperative nursery schools, health screening, and health promotion and prevention services. The survey should also cover workplace-related programs, such as breastfeeding support projects, child care programs, and

Families with Children Under 3: Policy Implications

401

family friendly employment policies (Sutherland et al. 2001). As has been demonstrated in this volume, many of the stresses that parents currently experience are time and work related. The inclusion of workplace and employment policy issues would be an important innovation in assessing community capacity and willingness to respond to the needs of families. Communities can consider the development of more global approaches to the provision of services to young children. The Carnegie Corporation’s Starting Points program and Proposition 10, the Children and Families First Initiative in California, are two good examples of holistic, community-wide approaches to promoting the development of all children (www.ccfc.ca.gov). Such global approaches serve as a way of integrating a diverse array of services, coordinating efforts to improve effectiveness and efficiency, and creating a set of permanent structures that can sustain children from birth to school (Halfon et al. 1999). The success of Proposition 10 has led a number of states around the country to consider similar broad-based, global initiatives on behalf of families with young children. In California, the strategies that many local Proposition 10 commissions are pursuing involve the development of new, integrated services platforms at birth and community hospitals, schools, WIC sites, and child care centers. At each of these potential family resource centers, community collaborations are being encouraged in order to develop, coordinate, and organize a full set of comprehensive services for families. Statewide initiatives such as California’s Children and Family First or North Carolina’s Smart Start, or the even more far-reaching national initiatives like Sure Start being carried out in the United Kingdom, appear to be part of a growing trend. Many of these broad-based initiatives are designed to optimize the developmental trajectories of young children by providing the appropriate mix of health, parenting, early education, and child development strategies in community-based settings. Underlying these approaches are three basic assumptions. First, the developmental trajectories for children that are initiated early in life can set in motion life-long learning and functional trajectories. The type of trajectory that children experience is partially accounted for by genetic and other biological factors, but also by the relative effect of risk and protective factors, which can either push down or elevate a given trajectory. Figure 13.2 graphically presents this notion by suggesting how different risk factors and protective factors can influence a learning trajectory. The experimental and interventional literature suggests that if young children are to be supported in obtaining necessary developmental competencies, inputs from health, early education, childcare, and other parental support services can play a significant role. Figure 13.3 shows how the attainment of such competencies can be influenced by different continuums of services. What has also emerged from the literature on service delivery and service integration is that, at present, many of the necessary early childhood

6 mo

12 mo

18 mo

Late Infancy

Early Infancy

Birth

Late Preschool 3 yrs 5 yrs Age Early Preschool

Late Toddler Early Toddler

24 mo

Social-emotional, Physical Cognitive, Language

Factors associated with school readiness:

Pre-school Appropriate Discipline

Reading to child

Parent education Emotional literacy

Poverty

Lack of health services

Family Discord

figure 13.2 Strategies to Improve School Readiness Trajectories.

Ready to learn

12 mo

18 mo

24 mo

3 ys

4 ys

Child Care Programs

Center and Home Based Family Resource Programs

Parenting and Family Literacy Programs

Pre-school

Early Intervention Programs (including Head Start and Early Head Start)

Ongoing well child care, acute, developmental and preventive care

6 mo

figure 13.3 Readiness to Learn Trajectory: Service Sectors and Programs That Influence School Readiness.

Social & Family Services:

Education:

Pre/perinatal care

Health

Birth

Social relationships Physiological regulation Entry into verbal milieu Attentional/motoric “control” of object world

Nascent awareness of standards Nascent emotional regulation Nascent selfhood/identity Language Development

Compliance to simple requests Emotional control Functional use of language Exploration of object world Effective peer interactions

5ys

404

Halfon, McLearn

focused services are inadequate in size and intensity, not organized for accessibility, and not configured to produce the type of continuum that would be most optimal. As communities across the nation attempt to build more appropriate and integrated early childhood systems, they are considering how to organize their early childhood services into more coordinated delivery platforms using delivery system innovations such as communitybased family resource centers (Figure 13.4). Provision of community-based family resource and support services is often hampered by financial constraints. These include inadequate funds, or funds that are rigidly administered along strict categorical lines. To better fund services targeted at young families within a community, it is important for communities to assess their current allocation of funds in both public and private sectors and consider strategies to optimize their investment in the lives of their youngest children (The Finance Project 2000). The development of an accurate young children’s budget is one way of beginning this process. In several communities, this has served as a first step in assessing available resources and opening the eyes of policymakers to the need for additional investment. Restrictions on the use of funds imposed by categorical programs is another key financing issue in early childhood services. While a number of programs provide funding for specific services such as child care, Head Start, or WIC, many communities are considering new ways to combine these funds into a more coordinated and flexible pool that could be used toward building more comprehensive and integrated program structures (Hayes 1997; Fisher, Cohen, and Flynn 2000). Several communities have also discovered ways of increasing their revenues by utilizing the Medicaid and EPSDT programs and their inherent flexibility to provide the fiscal mechanisms to develop more integrated approaches. As communities get started with their local funding and systems development strategies, many will identify additional public and private resources and partnering opportunities with business and other interest groups that also have a stake in improving outcomes for young children. For example, some communities are using adult education monies that have been traditionally focused on improving parents’ work skills and education level to provide parenting classes (The Finance Project 2000). In some, businesses have partnered with school and community-based providers to fund parent education and support activities. General Social Policy The scientific research on early childhood, the economic analyses of missed investment opportunities, and the greater recognition of the lifelong trajectories that are developed early in life have transformed child development from a children’s issue to a major issue of social policy. In a sense, these

0

2

3

Family Events

Early Headstart & Headstart

5

6 years

Recreation Programs

Family Day Care

Specialized services

Child oriented

Pediatric Care

4

Play-based problem solving

Parent Support

WIC Programs

Toy and resource libraries

Information & referral services

Pre/Postnatal Health Care

1

figure 13.4 Service Organization for Early Child Development and Parenting.

Libraries

Home-based satellites

Home-visiting network

Conception

Parent oriented

406

Halfon, McLearn

research findings and new theoretical approaches have transformed our view of childhood from something we merely “grow out of” into something we carry with us for the rest of our lives. This subtle shift in perspective means everything for how we develop social policy for children. Because childhood is more difficult to transcend than we once thought, more fundamental to whom each person ultimately becomes than had previously been appreciated, the legitimacy of excuses for ignoring it, or for delaying interventions until strongly negative health or developmental issues are blatantly obvious, is correspondingly reduced. The growing recognition of what is at stake in early childhood, the disparities that emerge from not supporting child and family development adequately, and the implications that these disparities have for educational systems, workplace, and human capital formation are staggering in their importance (Keating and Hertzman 1999). Not only do these have short-term fiscal implications, they have long-term implications for the lives of individuals who do not develop to the fullest extent of their capacity and become long-term dependents on other social programs. As noted here, the Commonwealth Survey has painted a broad picture of the families of children in the United States. Although many distinctions between families have emerged, common themes that cut across class and race boundaries are clear. For example, parental depression is most common among lower income families, but even among upper income families 8 percent of parents struggle with this problem. Difficulty paying for essential child-rearing items such as formula, diapers, and shoes is obviously more burdensome on the poor, but, as this study shows, over 24 percent of families report difficulties of this kind. Similarly, there are interventions of which middle and upper class parents routinely avail themselves, such as parenting classes and breastfeeding support, that are not readily available to all parents. These findings suggest that to some extent, family needs are spread across income and ethnic lines, and that those needs could potentially be addressed, if there is political vision to do so, with an approach that has common elements for all income groups. In short, the needs of families constitute the potential basis for a unifying rather than dividing social policy. The experience of parenting young children offers a common ground on which to re-found a new kind of civic society, predicated on a recognition that disparities are fewer at the beginning of life, and only multiply with time and neglect. Early childhood is a time of relative equality, the greatest equality our society knows. In the most basic sense put forth in the Bill of Rights, “all men are created equal.” Nowhere is the equality of persons more true than in the hospital nursery, despite the subtle influences of the prenatal environment. Realizing the political and ethical imperative of creating a more equal society should by rights begin at the time when equality is closest to being literally true and build forward from

Families with Children Under 3: Policy Implications

407

that foundation. Addressing the needs of families with young children is the lever to make sure that the relative equality with which life begins is not squandered through neglect and inadequate knowledge and support. This ethical and moral imperative has a practical side as well. Study after study has shown that inequality is not just bad for the poor: It drags down both the health and the wealth of society as a whole. The health of the wealthiest citizen is undermined simply by the fact of living in a society with great disparities of wealth and health (Wilkinson 1992, 1996). No less than the poor, the middle class cannot afford to live in a society in which extreme poverty persists. In all these regards, early childhood is more than a narrow sectoral issue for a special interest group. Although each individual family spends relatively little time in the early childhood phase, and although every child emerges from this phase, the experiences of each child individually and in combination with those of other children determine not only individual futures but the future health and wealth of the population as a whole. Early childhood and the family of the young child have the potential to be the focus of public policy that services the economic interests of our entire society and the political goal of making that society more equal and just. The data and analysis presented in this book have provided a glimpse of the family crucible in which the next generation is now being nurtured. It has shown how different forces, from access to resources to patterns of parenting, are already busy working their effects on the next generation. The decisions we make as a society will determine whether the relative equality of each baby in the nursery and each young child in the home will, or will not, become the foundation for a more equitable and prosperous society over the next hundred years.

references Aguilar, B., Sroufe, L.A., Egeland, B., & Carlson, E. (2000). Distinguishing the early-onset/persistent and adolescence-onset antisocial behavior types: from birth to 16 years. Development and Psychopathology, 12(2), 109–132. Anderson, J.E. (2000). Co-sleeping: Can we ever put the issue to rest? Contemporary Pediatrics, 17(6), 98. Bane, M.J., & Ellwood, D.T. (1989). One fifth of the nation’s children: Why are they poor? Science, 245, 1047–1053. Belsky, J., Gilstrap, B., & Rovine, M. (1984). The Pennsylvania Infant and Family Development Project, I: Stability and change in mother-infant and father-infant interaction in a family setting at one, three and nine months. Child Development, 55, 692–705. Bethell, C., Peck, C., Read, D., & Huang, E. (2001). Medicaid parents’ experience with the health care system. Summary of findings from a survey of parents of young children enrolled in Medicaid in three ABCD states. Prepared for the Commonwealth Fund by FACCT. Portland: FACCT.

408

Halfon, McLearn

Bornstein, M.H. (1995). Handbook of Parenting. Vol. 3. Status and Social Conditions of Parenting. Mahwah, NJ: Lawrence Erlbaum. Boyce, W.T., Quas, J., Alkon, A., Smider, N.A., Essex, M.J., Kupfer, D.J., MacArthur Assessment Battery Working Group, & MacArthur Foundation Research Network on Psychopathology and Development (in press). Autonomic reactivity and psychopathology in middle childhood. British Journal of Psychology. Bronfenbrenner, U. (1979). The Ecology of Human Development: Experiments by Nature and Design. Cambridge, MA: Harvard University Press. Bronfenbrenner, U. (1986). Ecology of the family as a context for human development research perspectives. Developmental Psychology, 22, 723–742. Bronfenbrenner, U., & Morris, P.A. (1998). The Ecology of Developmental Psychology. In W. Damon & R.M. Lerner (eds.), Handbook of Child Psychology, 5th Edition. Vol. 1. New York, NY: John Wiley and Sons. Brooks-Gunn, J., Berlin, L.J., & Fuligni, A.S. (2000). Early childhood intervention programs: What about the family? In J.P. Shonkoff & S.J. Meisels (eds.), Handbook of Early Childhood Intervention (549–588). New York, NY: Cambridge University Press. Brunner, E., & Marmot, M. (1999). Social organization, stress, and health. In M. Marmot & R.G. Wilkinson (eds.), Social determinants of health. New York, NY: Oxford University Press. California Department of Finance. (1997) http://www.dhs.ca.gov/hisp/chs/ OHIR/cabirthsbyraceth.htm. Accessed 9/25/00. Carnegie Corporation of New York. (1994). Starting Points: Meeting the Needs of Our Youngest Children. Abridged Version. Waldorf, MD: Carnegie Corporation of New York. Cauthen, N., Knitzer, J., & Ripple, C. (2000). Map and Track: State Initiatives for Young Children and Families. New York, NY: National Center for Children in Poverty. Committee for Economic Development Research and Policy Committee. (1993). Why Child Care Matters. Preparing Young Children for a More Productive America. Washington DC: Committee for Economic Development. Committee for Economic Development. (1991). The Unfinished Agenda: A New Vision for Child Development and Education. Washington DC: Committee for Economic Development. Committee for Economic Development. (1985). Investing in our Children: Business and the Public Schools. Washington DC: Committee for Economic Development. Cowan, C.P., & Cowan, P.A. (1992). When Partners Become Parents. New York, NY: Basic Books. Currie, J., & Thomas, D. (1995). Does Head Start Make a Difference? The American Economic Review, 85(3), 341–364. Danyliw, N.Q. (1997). Got Mother’s Milk? U.S. News and World Report, Dec. 15. Danziger, S., & Waldfogel, J. (2000). Securing the Future: Investing in Children from Birth to College. New York, NY: Russell Sage Foundation. Davies, M., & Jacobs, F. (1991). Rhetoric or reality? Child and family policy in the United States. Social Policy Report, 5(4), 1–25. Ferber, R. (1990). Sleep, schedule-dependent causes of insomnia and sleepiness in middle childhood and adolescence. Pediatrician, 17(1), 13–20. Field, T. (1995). Psychologically depressed parents. In M.H. Bornstein (ed.), Handbook of Parenting. Vol. 4. Mahwah, NJ: Lawrence Erlbaum.

Families with Children Under 3: Policy Implications

409

Field, T., Healy, B., Goldstein, S., & Guthertz, M. (1990). Behavior-state matching and synchrony in mother-infant interactions on nondepressed versus depressed dyads. Developmental Psychology, 26, 7–14. Fielding, J.E., Sutherland, C., & Halfon, N. (1999). Community report cares: Results of a national survey. American Journal of Preventive Medicine, 17(1), 79–86. The Finance Project. (2000). Improving educational achievement: A volume exploring the role of investments in schools and other supports and services for families and communities. Working Paper. Washington, DC: The Finance Project. Fisher, H., Cohen, C., & Flynn, M. (2000). Federal Funding for Early Childhood Supports and Services: A Guide to Sources and Strategies. Washington, DC: The Finance Project. Fraiberg, S. (1980). Clinical Studies in Infant Mental Health. New York: Basic Books. Furan, D.C. (2000). Another decade of intervention for children who are low income or disabled: What do we know now? In J.P. Shonkoff & S.J. Meisels (eds.), Handbook of Early Childhood Intervention. Cambridge: Cambridge University Press. Furstenberg, F.F. (1988). Good dads – bad dads: Two faces of fatherhood. In A.J. Cherlin (ed.), The Changing American Family and Public Policy. Washington, DC: Urban Institute Press. Gabarino, J., & Ganzel, B. (2000). The Human Ecology of Early Risk. In J.P. Shonkoff & S.J. Meisels (eds.), Handbook of Early Childhood Intervention. Cambridge: Cambridge University Press. Garcia Coll, C., & Magnuson, K. (2000). Cultural differences as sources of developmental vulnerabilities and resources. In J.P. Shonkoff & S.J. Meisels (eds.), Handbook of Early Childhood Intervention. Cambridge: Cambridge University Press. Gartner, L., Black, L.S., & Eaton, A.P. (1997). Breastfeeding and the use of human milk. Pediatrics, 100(6), 1035–1039. Golova, N., Alario, A.J., Vivier, P.M., Rodriquez, M., & High, P.C. (1999). Literacy promotion for Hispanic families in a primary care setting: A randomized, controlled trial. Pediatrics, 103:993–997. Graham, C.A., & Easterbrooks, M.A. (2000). School-aged children’s vulnerability to depressive symptomatology: The role of attachment security, maternal depressive symptomatology, and economic risk. Development and Psychopathology, 12(2), 201–213. Green, M., & Palfrey, J.S. (2000). Bright Futures: Guidelines for Health Supervision of Infants, Children, and Adolescents, Second Edition. Arlington, VA: National Center for Education in Maternal and Child Health. Grubb, W.N., & Lazerson, M. (1990). Broken Promises: How Americans Fail Their Children. Chicago, IL: University of Chicago Press. Halfon, N., Gonzalez, R., & Hochstein, M. (1999). Building bridges for California’s young children: A 12-point agenda to enhance proposition 10. Berkeley: California Policy Research Center. Halfon, N., & Hochstein, M. Life Course Health Development: An Integrated Framework for Developing Health and Financing Health Intervention. In Press, The Milbank Quarterly. Halfon, N., Hochstein, M., Wells, C.R., Shoaf, K., Regalado, M., Barthauer, L., & Schuster, M.A. Provision of developmental services to children 0–3 years old. Presented at the annual meeting of the Ambulatory Pediatric Association, May 12–17, 2000.

410

Halfon, McLearn

Halfon, N., Inkelas, M., & Hochstein, M. (2000). The health development organization: An organizational approach to achieving child health development. The Milbank Quarterly, 78(3), 447–497. Halfon, N., Newacheck, P.W., Hughes, D., & Brindis, C. (1998). Community health monitoring: Taking the pulse of America’s children. Maternal and Child Health Journal, 2(2), 95–109. Halpern, R. (1990). Community-based early intervention. In S.J. Meisels & J.P. Shonkoff (eds.), Handbook of Early Childhood Intervention (469–498). New York, NY: Cambridge University Press. Halpern, R. (2000). Early childhood intervention for low-income children and families. In J.P. Shonkoff & S.J. Meisels (eds.), Handbook of Early Childhood Intervention. Cambridge: Cambridge University Press. Hayes, C.D. (1997). Financing Services for Young Children and Their Families: Meeting the Challenges of Welfare Reform. Washington, DC: The Finance Project. Heckman, J.J. (1999). Policies to foster human capital. National Bureau of Economic Research. Working Paper 7288. http//www.nber.org/papers/w7288. Heinicke, C.M. (1995). Determinants of the Transition to Parenting. In M.H. Bornstein (ed.), Handbook of Parenting. Vol. 3. Status and Social Conditions of Parenting. Mahwah, NJ: Lawrence Erlbaum Associates. High, P., Hopmann, M., LaGasse, L., & Linn, H. (1998). Evaluation of a clinic-based program to promote book sharing and bedtime routines among low-income urban families with young children. Archives of Pediatrics and Adolescent Medicine, 152, 459–465. High, P.C., LaGasse, L., Becker, S., Ahlgren, I., & Gardner, A. (2000). Literacy promotion in primary care pediatrics: Can we make a difference? Pediatrics, 105, 927–934. Hutchins, V.L. (1997). A History of Child Health and Pediatrics in the United States. In R.E.K. Stein (ed.), Health Care for Children. What’s Right, What’s Wrong, What’s Next. New York: United Hospital Fund. Kagan, S.L., & McLearn, K.T. (2000). Supporting infants and toddlers: The nascent policy agenda. In D. Cryer & T. Harms (Eds.), Infants and Toddlers in Out-of-Home Care. Baltimore, MD: Paul H. Brooks Publishing Co. Kamerman, S.B., & Kahn, A.J. (1995). Starting Right. New York: Oxford University Press. Kandel, E.R. (1998). A new intellectual framework for psychiatry. American Journal of Psychiatry, 155(4), 457–469. Karoly, L., Greenwood, P.W., Everingham, S.S., Hoube, J., Kilburn, M.R., Rydell, C.P., Sanders, M., & Chiesa, J. (1997). Investing in our children: What We Know and Don’t Know About the Costs and Benefits of Early Childhood Interventions. Santa Monica, CA: RAND. Kaus, M. (1992). The End of Equality. New York, NY: Basic Books. Keating, D.P., & Hertzman, C. (Eds.) (1999). Developmental Health and the Wealth of Nations: Social, Biological and Educational Dynamics. New York, NY: Guilford Press. Lakoff, G. (1996). Moral Politics. Chicago, IL: University of Chicago Press. Lamb, M.E., Plecke, J.H., & Levine, J.A. (1985). The role of the father in child development: The effects of increased paternal involvement. In B. Lahey & E.E. Kazdin (eds.), Advances in Clinical Child Psychology. Vol. 8. New York, NY: Plenum.

Families with Children Under 3: Policy Implications

411

Lerner, R.M., Castellino, D.R., Terry, P.A., Villarruel, F.A., & McKinney, M.H. (1995). Developmental contextualism perspective on parenting. In M.H. Bornstein (ed.), Handbook of Parenting. Vol. 2. Biology and Ecology of Parenting. Mahwah, NJ: Lawrence Erlbaum Associates. Lyons-Ruth, K., Connell, D., Grunebaum, H., & Botein, S. (1990). Infants at social risk: Maternal depression and family support services as mediators of infant development and security of attachment. Child Development, 61, 85–98. Lyons-Ruth, K., & Zeanah, C. (1993). The family context of infant mental health Part I: Affective development in the primary caregiving relationship. In C. Zeanah (ed.), Handbook of Infant Mental Health. New York, NY: Guilford. McKenna, J. (2000). Cultural influences on infant and childhood sleep biology, and the science that studies it: Toward a more inclusive paradigm. Zero to Three, 20(3). McLearn, K.T., Zuckerman, B.S., Parker, S., Yellowitz, M., & Kaplan-Sanoff, M. (1998). Child development and pediatrics for the 21st century: The Healthy Steps approach. Journal of Urban Health: Bulletin of the New York Academy of Medicine, 75(4), 704–723. McLoyd, V.C. (1990). The impact of economic hardship on black families and children: Psychological distress, parenting, and socio-emotional development. Child Development, 61(2), 311–346. Minkovitz, C., Strobino, D., Hughart, N., Scharfstein, D., & Guyer, D. (2001). Early effects of the Healthy Steps for Young Children program. Archives of Pediatrics and Adolescent Medicine, 155(4), 470–479. Moore, J.F., & Jansa, N. (1987). A survey of policies and practices in support of breastfeeding mothers in the workplace. Birth, 14(4), 191–195. Moore, K. (1996). Comparing the Commonwealth Survey of Parents with Young Children to other national data. Child Trends. Morelli, G.A., Rogoff, B., Oppenheim, D., & Goldsmith, D. (1992). Cultural variations in infant’s sleeping arrangements: Questions of interdependence. Developmental Psychology, 28, 604–613. Musick, J., & Stott, F. (2000). Paraprofessionals revisited and reconsidered. In J.P. Shonkoff & S.J. Meisels (eds.), Handbook of Early Childhood Intervention. Cambridge: Cambridge University Press. National Center for Children in Poverty. (1996). One in Four. America’s Youngest Poor. Abridged Version. New York: Columbia University School of Public Health. National Commission on Children. (1991). Final Report. Beyond Rhetoric: A New American Agenda for Children and Families. Washington, DC: National Commission on Children. Newacheck, P.W., McManus, M., Fox, H.B., Hung, Y.Y., & Halfon, N. (2000). Access to health care for children with special health care needs. Pediatrics, 105(4, part 1), 760–766. Olds, D.L., Eckenrode, J., Henderson, C.R., Kitzman, H., Powers, J., Cole, R., Sidora, K., Morris, P., Pettitt, L.M., & Luckey, D. (1997). Long-term effects of home visitation on maternal life course, child abuse and neglect, and children’s arrests: fifteen year follow-up of a randomized trial. Journal of the American Medical Association, 278(8), 637–43. Orr, S.T., & Miller, C.A. (1995). Maternal depressive symptoms and the risk of poor pregnancy outcome. Review of the literature and preliminary findings. Epidemiology Review, 17(1), 165–171.

412

Halfon, McLearn

Osofsky, J.D., & Thompson, D. (2000). Adaptive and maladaptive parenting: Perspectives on risk and protective factors. In J.P. Shonkoff & S.J. Meisels (eds.), Handbook of Early Childhood Intervention. Cambridge: Cambridge University Press. Parke, R.D. (1995). Fathers and families. In M.H. Bornstein (Ed.), Handbook of Parenting. Vol. 3. Status and Social Conditions of Parenting. Mahwah, New Jersey: Lawrence Erlbaum. Phillips, M., Crouse, J., & Ralph, J. (1998). Does the black-white test score gap widen after children enter school? In C. Jencks & M. Phillips (eds.), The Black-White Test Score Gap. Washington, DC: Brookings Institution Press. Radke-Yarrow, M., Cummings, E.M., Kuczynski, L., & Chapman, M. (1985). Patterns of attachment in two and three year olds in normal families and families with parental depression. Child Development, 56, 884–893. Sameroff, A.J. (1987). The social context of development. In N. Eisenberg (ed.), Contemporary Topics in Developmental Psychology. New York: Wiley. Schor, E.L. (1995). The influence of families on child health. Family behaviors and child outcomes. Pediatric Clinics of North America, 42(1), 89–102. Shonkoff, J.P., & Meisels, S.J. (2000). Handbook of Early Childhood Intervention. Cambridge: Cambridge University Press. Shonkoff, J.P., & Philips, D. (eds.). (2000). From Neurons to Neighborhoods: The Science of Early Childhood Development. Washington, DC: National Academy Press. Smith, J. (1999). Healthy bodies and thick wallets. The dual relation between health and economic status. Journal of Economic Perspectives, 13(2), 145–166. Suomi, S.J. (1999). Developmental trajectories, early experiences, and community consequences. Lessons from studies with rhesus monkeys. In Developmental Health and the Wealth of Nations. New York: Guilford Press; 185–200 Sutherland, C., McCroskey, J., & Halfon, N. (2001). The Challenges of Measuring the Impact of Pioposition 10. In N. Halfon, E. Shulman, Building Community Systems for young children, UCLA Center for Healthier, Families and Communities. Sylva, K. (1997). Critical periods in childhood learning. British Medical Bulletin, 53(1), 185–197. Texas Department of Health. (1999). Vital Statistics Records. http://www.tdh. state.tx.us/bvs/stats97/ANNR PDF/97t03.PDF. Accessed 9/25/00 Wadsworth, M. (1999). Early life. In M. Marmot & R.G. Wilkinson (eds.), Social Determinants of Health. New York: Oxford University Press. Whitehurst, G.J., & Lonigan, C.J. (1998). Child development and emergent literacy. Child Development, 69(3), 848–872. Wilkinson, R.G. (1992). Income distribution and life expectancy. British Medical Journal, 304(6820), 165–8. Wilkinson, R.G. (1996). Unhealthy Societies: The Afflictions of Inequality. New York: Routledge. Zeanah, C. (1993). Handbook of Infant Mental Health. New York: Guilford Press. Zuravin, S.J. (1989). Severity of maternal depression and three types of mother-tochild aggression. American Journal of Orthopsychiatry, 59, 377–389.

Index

Abuse. See Child abuse; Physical abuse; Sexual abuse Access, to health care. See Health care, access to Addiction, drug, 358 Age, child in disciplining, 152, 156 in parent–child activities, 137 Age, mother, 57, 58, 60, 71 in breastfeeding, 190, 191, 193, 194, 195, 197, 198 in prenatal care use, 275–76, 278–79 Age, parent in anticipatory guidance, 328 on child IQ and learning, 138 in depression, 218, 219, 230 in parent–child activities, 138 in spanking, 157, 160, 161, 162, 163 Aggravation, parental. See also Frustration, parental depression and, 242, 243, 244, 246 in spanking, 160–64 Aggression, spanking and, 150 Alcohol use, parental, 24 Alcoholics, 358 children of, 265 American Academy of Pediatrics on anticipatory guidance, 322 Breastfeeding Work Group of, 199 on health care access, 316 on physical punishment, 150, 151

Anticipatory guidance. See Guidance, anticipatory Arguments, parent-child, in spanking, 154 Attachment behavior, parental depression on, 221 Authoritarian parenting style, 149 Authoritative parenting style, 148 Baby Friendly hospitals, 203 Balanced Budget Act of 1997, 314–15 Baumrind, Diana, 148–49 Behavior, parent. See specific types, e.g., Parent–child activities Beyond Rhetoric, 5, 377 Biobehavioral organization, parental depression on, 222 Bioecological systems perspective, 23, 378–79, 397–98 Birth order, in parental depression, 229, 230, 231 Birthing centers, on breastfeeding practice, 186–88, 202–4 Birthweight, low ethnicity/race on, 280 income on, 280–81 prenatal care in, 270, 280–81 Brain, of young child, 5 Brain development better understanding of, 349–52 breast milk in, 180 cortisol in, 350–51 413

414 Braveman, P., 272 Brazelton Neonatal Behavioral Assessment Scale (BNBAS), 353–54 Breastfeeding, 32, 178–210, 386 childcare on, 186 on cognitive development, 180–81 community services on, 203–4 continuation of, 194–97 cultural differences in, 184–85, 190 data on, 183 education on, childbirth, 191, 192, 193, 194, 195, 199–201 education on, maternal, 193, 194, 195, 196, 197 educational level on, maternal, 190, 191–92, 197, 198 ethnicity/race on, 183–84, 189–91, 193, 194, 195, 196, 197, 198 financial savings of, 182 governmental assistance programs on, 182, 190, 192, 193, 194, 204–5 on health, child, 181–82 on health, maternal, 182 health care system challenges and, 186–88 home visits on, 191, 204 hospitals on, 186–88, 202–4 income on, 190, 192, 193, 197, 198 initiation of, 192–94 marital status on, 190, 191, 193, 194, 195, 197, 198 maternal age on, 190, 191, 193, 194, 195, 197, 198 Medicaid services on, 200–201 mother’s age on, 190, 191, 193, 194, 195 oxytocin in, 180 parent–infant connection from, 179–80 policy implications for, 201, 203, 206, 208 prenatal care on, 190, 192, 197–200 prenatal services providers on, 199–200 prevalence of, 179, 189 public policy on, 205–6 rates of, surveillance of, 208

Index region on, 193, 194, 195, 197, 198, 203–4 return to work or school on, 206–8 study of, methods for, 188–89 as time resource, 32 unsuccessful, 272 U.S. Public Health Service goals for, 182–83 U.S. trends in, 182–85 vaginal vs. cesarean delivery on, 191, 192 welfare reform on, 186 in working mothers, 185–86 Breastfeeding Work Group, 199 Breastmilk on brain development, 180 nutritive value of, 181 Bright Futures, 315, 316, 322–23, 324 British 1946 National Birth Cohort Study, 2 Brooks-Gunn, J., 23, 154–55 Brown, G., 229 Bus, A.G., 11 California Proposition 10, 6, 379, 401 Canadian Center for Advanced Studies, 373 Caregiving responsibility, new parent, 84–85. See also Parenthood, new daily care in, 89–92 Carnegie Corporation Starting Points report, 5, 117, 320, 369–70, 401 Census Bureau Survey of Income and Program Participation (SIPP), 273 Cesarean delivery, 269–70 on breastfeeding, 187 postpartum stay length after, 281–82 Child abuse, parental. See also specific types, e.g., Physical abuse on spanking, 158, 162 Child and Adolescent Health Measurement Initiative (CAHMI), 317 Child care on breastfeeding, 186 nonparental, 5 Child care providers, shared reading with, 141

Index Child development disparities in, 2 ecological theories of, 23, 378–79, 397–98 environment in, 22–23 expanded services for, 355–57 family structure on, 3 identifying problems of, 356–57 information on, access to, 307, 309–10 long-term, family in, 395–96 parent health on, 352–53 as process, 22 promotion of, by pediatricians, 355–56 resources on, 23 social structures in, 3, 371 Child development models, in resource allocation, 22–24 Child development services, expanded, 355–57 Child health. See Health, child Child mental health care, lack of, 301, 303 Child rearing, parental sharing of, 368 Child rearing knowledge, from clinicians anticipatory (See Guidance, anticipatory) aspects discussed in, 331 factors in receipt of, 331–32 families most likely to benefit less likely to seek, 336 helpful steps already taken in, 338 lack of discussion on, 336 other sources of information on, 332–33 parents’ rating of quality of, 334 parents turning to other sources for, 337 parents valuation of, 336–37 parents’ willingness to pay extra for, 334–35 reasons parents not getting help in, 337 recommendations on, 338–40 Child rearing knowledge, parental acquisition of, 36–37 Child welfare movement, 321

415 Childbirth education. See Education, childbirth Childhood, changing views of, 3–4 Children on life at home, 34, 35 parental frustration with, 35–36, 106–8, 111 social policy since 1960 for, 374–80 Children and Families First Act, 6, 401 Children Now, 398 Children of the Great Depression, 3 Children’s Health Insurance Program, 314–15 ChildServ, 398 Clinton, President, 117 Cognitive development breastfeeding on, 180–81 with depressed parents, 220 in psychopathology, 3 Cognitive function, early neurobiology of, 5 Committee for Economic Development (CED), 380 Community-level policies, 400–404 Community services, on breastfeeding, 203–4 Community support, on parental depression, 228, 239–40 Confidence, parenting, 389. See also Self-confidence on parental depression, 228, 231, 233, 237, 240–41 Coping, 389 by new parents, 105, 111 Cortisol in brain development, 350–51 on learning, 351 Cuddling, parent–child, 119, 123, 128–29 on parental depression, 242–43 Culture. See also Ethnicity/race on breastfeeding, 184–85 on discipline, 148, 174 diversity of, 368–69 diversity of, in families, 368–69 Daily care, division of, by new parents, 89–92, 108–9

416 Daily routines, family, 119–20, 124–26, 129–30, 242–43, 385–86 depression and, 132, 133, 242–43, 245, 250, 252 Darwinism, neural, 350 Davies, M., 378 Day, R.D., 153–54 Delivery. See also Cesarean delivery; Vaginal delivery “drive-through,” 271 Delivery type on breastfeeding, 191, 192 on early postpartum discharge, 281–82 prenatal care on, 269–70, 279–80 Dental care, lack of, 303–4 Depression, childhood, 372 Depression, parental, 24, 217–55, 265, 358, 386, 388–89, 393–94 abuse of parent in, 230, 232, 235, 236, 238 age in, 218, 219, 230 aggravation in, 242, 243, 244, 246 analysis of, methods of, 225 on attachment behavior, 221 on behavioral organization, 222 birth order on, 229, 230, 231, 234–35 children as risk factor in, 251 on cognitive skills, child, 220 definition of, 224–25 education in, parenting, 253 educational level in, 218, 228, 229, 230, 232, 236 employment on, 232, 236 ethnicity/race in, 218, 228, 229, 232, 236 family profiles of, 226–27, 231–33, 236–37, 238 family support teams for, 254 with firstborn, 230, 231, 234–35 on frustration with parenting, 107, 111, 389 gender in, 218, 219, 225–26, 250 health on, 230, 232, 238–39 high demand-low control in, 251, 389 home visits and, 254 income in, 226, 228, 230, 232, 235, 236

Index irritable and punitive stance in, 249 marital status on, 228, 229, 230, 232, 236 medication on, 253 number of children on, 228, 229 parent–child interactions in, 131–32 parent–child interactions in: conflict-related, 241–50 parent–child interactions in: play, 242–43, 245, 250 parent–child interactions in: positive, 242–43, 248 parent–child interactions in: positive, low, 131–32, 246, 247, 249–50 parent–child interactions in: reading, 245, 250 parent–child interactions in: routines, 132, 133, 242–43, 245, 250, 252 on parenting, effective, 252 parenting confidence on, 228, 231, 233, 237, 240–41, 389 in parenting problems, 253–54 policy implications of, 250–55 prevalence of, 218–20 on psychopathology, child, 221, 250 on psychophysiology, child, 221–22, 250 research data on, generalizing, 251 as risk for child outcomes, 220–23, 251 screening for, 253 on social skills, child, 220–21 social support on, 228, 231, 233, 237, 239–40 sociodemographics of, 226–37 sociodemographics of: bivariate analysis, 226–29, 230–31 sociodemographics of: multivariate analysis, 229–35 on spanking, 157, 161, 163, 242, 245, 248, 252 stress in, parenting, 217 stressful prebirth events on, 228, 230, 232, 235, 236, 238, 248 survey questions on, 223–24 withdrawn stance in, 249

Index Developmental disabilities, 294 Developmental disparities, 373 Developmental gradients, 373, 385 Developmental psychology, of young children, 5 Developmental services, 316 Developmental trajectory, 371, 374, 395 Dike v. Orange County School Board, 205 Disabilities, developmental, 294 Discharge, early postpartum, 281–85, 287 demographic differences in, 281–83 parent concerns about, 284–85 parent information and infant health in, 283–84 Discipline, 383, 386–87. See also Punishment authoritarian style of, 149 authoritative style of, 148 Baumrind’s approach to, 148–49 of boys vs. girls, 154–55, 160 child age in, 152, 156 cluster analysis on, 165–70 cluster analysis on: high-interacters, 166, 168, 170 cluster analysis on: high negative discipline, 166, 169, 170 cluster analysis on: high overall discipline, 166, 169, 170 cluster analysis on: low-interacters, 165, 168, 170 context of, 164–70, 173 counseling on, 172–73 culture on, 148, 174 definition and scope of, 146–47 Disciplining Children in America survey on, 151–52 explaining in, 160, 242, 248 harsh words in, 153 hitting in, 152, 153, 155, 242, 246, 252 income on, 152, 154–55 National Family Violence Surveys on, 152–53 National Longitudinal Study of Youth on, 153

417 National Survey of Families and Households on, 153–54 vs. nurturing interactions, 164, 165 parents need for information on, 170–71, 173–74 permissive style of, 149 previous studies on, 151–55 vs. punishment, 147 by sexually abused parents, 152 shoving in, 152, 242 slapping in, 152, 155, 242, 246, 248 vs. spanking, 164, 165 spanking in, 147, 150, 152, 153–54, 155, 157–63, 172–73, 242, 245, 248, 252, 386–87 spanking in, aggression and, 150 Stern on, 148 survey questions on, 155 taking things away in, 160, 242, 248 time out in, 155, 171–72, 242, 248 in white vs. minority families, 153 yelling in, 152, 153, 155, 242, 245, 246, 248, 252 Disciplining Children in America, 151–52 Disparities, developmental, 373 Division of daily care, in new parenthood, 89–92, 108–9 of responsibilities, in new parenthood, 84–85, 108–9 Doctor–patient relationships. See Guidance, anticipatory Drug addiction, 358 Drug use, parental, 24 Dubay, L.C., 31 Dworkin, P.H., 356 Early and Periodic Screening, Diagnosis, and Treatment Program, 323 Early Childhood Public Engagement Campaign, 370 Early Head Start, 6, 370, 379, 400 Earned Income Tax Credit, 377 Ecological theories, of child development, 23, 378–79, 397–98

418 Economic model, of family decision making, 25–26 Economics. See also Income; Poverty of breastfeeding, 182 of early childhood intervention, 4, 373 of raising children, 3 of young families, 378 Education, childbirth, 394–95 on breastfeeding, 191, 192, 193, 194, 195, 199–201, 386 on parental play with child, 133, 136–37 on parental readiness, 52–53, 58, 59, 61, 62, 67–68, 76 on prenatal care, 268–69, 279 on shared book reading, 132–35 Education, hospital, on newborn care, 97–98, 100, 110, 272 Education, new parent, 85–87, 97–105, 110–11 by family, 97–100, 110 by health professionals, 99–100, 101, 110 by media, 100–101, 102, 103, 110 by parenting classes, 101–5, 110–11 Education, parenting, 383–84, 386, 394–95 on new parenthood, 100–101 on parental play with child, 133, 136–37 on parental readiness, 52–53, 58, 59, 61, 67–68, 76 pediatrician role in (See Guidance, anticipatory) on shared book reading, 132–35 Education for All Handicapped Act, 376–77 Educational level, maternal on breastfeeding, 190, 191–92, 197, 198 on depression, 218, 228, 229, 230, 232, 236 on prenatal care use, 267, 276, 278 Educational level, parental on anticipatory guidance, 328 on depression, 218, 228, 229, 230, 232, 236

Index on new parenthood, 111 on parent readiness, 58, 59, 63, 71, 75–76 on parent–child activities, 138 on prenatal care, 267 on resource allocation, 36 on spanking, 158, 160, 161, 162, 163 Elder, Glen, 3 Emotional development, in psychopathology, 3 Emotional health, parental, 35, 55 Emotional support, of parents, 54 Employment, 26–27, 30, 111–12, 381. See also Income on anticipatory guidance, 328 on depression, 232, 236 maternal, on breastfeeding, 206–8 on new parenthood, 90–91, 93–94, 96, 109, 110, 111–12 on parenting, 3 on spanking, 158 Employment Retirement Income Security Act (ERISA), 271 Environment, 371 in child development, 22–23 Ethnic diversity, 368–69 of families, 368–69 Ethnicity/race, 380–81 in anticipatory guidance, 329 in breastfeeding, 183–84, 189–91, 193, 194, 195, 196, 197, 198 in depression, parental, 218, 228, 229, 232, 236 in discipline, 148, 174 in early postpartum discharge, 281, 283 in health care access, 295 in parental readiness, 56, 58, 59, 63, 71, 75 in parent–child activities, 139–40 in pregnancy intention, 65–66 in prematurity and low birthweight, 280 in prenatal care, 267, 275–76, 278 in spanking, 158, 161 Explaining, 160, 242, 248 Exploration, child, intrusions on, 148 Eyeglasses, lack of, 301, 303

Index Families and Work Institute, 370 Family(ies), 367–407 changing nature of, 4, 383 changing views of, 3–4 child rearing in, changing context of, 367–74, 375 children’s long-term development in, 395–96 community-level policies on, 400–404 cultural and ethnic diversity of, 368–69 decision making in, economic model of, 25–26 demographics of, 367–68 health care and other social policies on, 397–400 health care system role in, 389–92 in new parent education, 97–100, 110 number of children in, 27–28 parent characteristics in, 380–84 parent child rearing in, 384–88 parent depression in, 388–89 parent mental health in, 358, 386, 388–89, 393–95 public health implications of, 392–93 social policy on, general, 404–7 social policy on, since 1960, 374–80 structure of, in parental readiness, 56, 58, 59–60, 62, 64, 65, 71 structure of, on child development, 3 Family and Medical Leave Act, 6 Family planning, 357 Family systems theory, 379 Family therapy theory, 379 Fathers in child rearing, 368 parenting role of, 84–85, 109–10, 382–83 Federal Even Start Family Literacy Program, 142 Federal legislation. See also specific legislation and programs recent, 6 Federal Women, Infants, and Children Supplemental Feeding Program (WIC), 182, 189, 204–5

419 Financial resources, 23, 27. See also Income; Poverty; Purchased resources family income in, 29–30 Firstborn, parents of, 57, 58, 64, 71 anticipatory guidance for, 328 caregiving sharing by, 89–90 parental depression in, 230, 231, 234–35 parental readiness in, 57, 58, 64, 71 Foundation for Accountability (FACCT), 317, 398–99 From Neurons to Neighborhoods, 5–6, 371–72 Frustration, parental, 163–64, 389 with child, 35–36 with child, in new parents, 106–8, 111 in spanking, 159, 160–62, 163–64 Gender, child on anticipatory guidance, 327 on spanking, 159 Gender, parent on anticipatory guidance, 328 in parental depression, 218, 219, 225–26, 250 on spanking, 157 Glascoe, F.P., 356 Goals 2000, 117 Government input, 37–39, 45 Governmental assistance programs. See also specific programs on breastfeeding, 182, 190, 192, 193, 194, 204–5 on parental depression, 230, 232, 236 Grandparent support, 39, 58, 59 benefits from, 55 Guidance, anticipatory, 320–40, 391 age-normed percentiles in, 325–30 alternative models of service delivery in, 339 analyses of, 330 aspects discussed in, 330 child welfare movement in, 321 evaluation of, expanding, 338–39 factors in receipt of, 330–32 families most likely to benefit less likely to seek, 336

420 Guidance, anticipatory (cont.) helpful steps already taken in, 338 lack of discussion with doctor on, 336 making it easier for physician provision of, 339 other clinician services offered and, 333 parental child-rearing concerns in, 322–24 parents’ rating of quality of, 334 parents’ willingness to pay extra for, 334–35 pediatrics in, rise of, 321–22 physician training in, improving, 339–40 reasons parents not getting help in, 337 recommendations on, 338–40 sources for, parents turning to other, 337 sources of information on, other, 332–33 study of, independent variables in, 330 study of, outcome variables in, 324–25 valuation of, by parents, 336–37 Guidelines for Health Supervision, 322 Halfon, N., 324 Handicapped Children’s Early Education Assistance Act, 376 Harris, T., 229 Harsh words, 153 Head Start, 6, 376, 377, 400 Early, 6, 370, 379, 400 Health, child breastfeeding on, 181–82 maternal health on, 182 parental health on, 5, 352–53 risks to, 3 Health, emotional, of parent, 35, 55. See also Depression Health, infant, early discharge on, 283–84

Index Health, maternal breastfeeding on, 182 on children’s health, 182 before conception, 357–58 promotion of, 360 Health, parent childhood health in, 5 on children’s health and development, 352–53 in depression, 230, 232, 238–39 Health care, access to, 293–318 delayed care on cost of, 304–6 early childhood development information and, 307, 309–10 ethnicity/race on, 295 health insurance on, 295–96, 314 health insurance on, breadth of covered services, 312 and health status, of U.S. children, 293–94 income on, 295 legislation on, new, 314–15 missed care on cost of, 301, 303–4 paying for, difficulty in, 312, 313 physician services use in, 306 policy implications of, 312–16 policy implications of, enhanced services, 315–17 policy implications of, traditional services, 313–15 population characteristics in, 298–99 quality in, 317, 391 receipt and usefulness of services in, 309, 311–12 statistics on, 294–95 study of, data and approach to, 296–98 study of, statistical analysis in, 298 study of, variable construction in, 297–98 usual source of care in, 299, 300, 301–2 usual source of care in, satisfaction with, 307, 308 Health care policies, 397–400 Health care system, 389–92 breastfeeding and, 186–88

Index Health insurance, 394 on anticipatory guidance, 328 children’s, 30–31, 33, 41, 44 on delayed and missed care, 304–6 on depression, parental, 230, 232, 236 government-provided, 38–39 on health care access, 295–96, 312, 314, 390–91 on home visitations, 283–84 on prenatal care, 267–68, 276–77, 278–79 Health outcomes, disparities in, 2 Health Plan Employer Data Information Set (HEDIS), 317 Health professionals. See also Guidance, anticipatory; Pediatric care on breastfeeding, 199–200 in new parent education, 99–100, 101, 110 in parent education on discipline, 171, 174 Health promotion strategies, 374, 375 Health status, 3 of U.S. children, 293–94 Healthy People 2000, 182 Healthy Steps, 360–62, 370, 397–98, 399 Hertzman, C., 373–74 Hitting, 152, 153, 155, 242 parental depression and, 242, 246, 252 Home environment, on literacy and school readiness skills, 140 Home Instruction Program for Preschool Youngsters (HIPPY), 142 Home production, 26, 31–32. See also Time resources in one- vs. two-parent families, 26 Home visits, 273 on breastfeeding, 191, 204 on depression, parental, 254 health insurance on, 283–84 income on, 283 prevalence of, 283–84 Hoobler, B.R., 320 Hospital(s) breastfeeding and, 186–88, 202–4

421 in new parent education, 97–98, 100, 110, 272 Hospital stay, postpartum, 270–73, 281–85, 287 Hugging parent–child, 119, 123, 128–29, 385 punishment and, 164, 165 Human capital resources, 23, 24, 36–37 Human development. See also Child development disparities in, 2 I Am Your Child Foundation, 370 Income, 29, 381. See also Employment on anticipatory guidance, 320 on breastfeeding, 190, 192, 193, 197, 198 on delayed and missed care, 304–6 on depression, parental, 226, 228, 230, 232, 235, 236 on discipline, 152, 154–55, 160 disparity in, 349 distribution of, 29–30 employment on, 30 on health care access, 295 on home visits, 283 on parental readiness, 56 on parent–child activities, 138–39 on postpartum discharge, early, 281, 283 on prematurity and low birthweight, 280–81 on prenatal care, 267–68, 274–75, 277, 278 on spanking, 157, 161, 162, 163 Individuals with Disabilities Education Act (IDEA) Part C, 6 Infant mortality, reductions in, 3 Institute of Medicine From Neurons to Neighborhoods report, 5–6, 371–72 Insulting, 153 Insurance. See Health insurance Investing in Our Children: Business and the Public Schools, 1, 380 Investments, per child, 21–22. See also Resource allocation

422 Jacobs, F., 378 Jaundice, neonatal, 272 Jobs. See Employment Kagan, Jerome, 147–48 Keating, D.P., 373–74 Kennedy, President, 376 Kenney, G.M., 31 Kin involvement. See also Grandparent support parenting skills and, 54–55 Klaus, M.H, 179 Kotelchuck, M., 267 Lactation management education, 199–200 Learning cortisol on, 351 early, 5 early, better understanding of, 349–52 Life course trajectories, interventions on family relationships in, 5 Literacy. See also Reading as continuous process, 351 skills in, home environment on, 140 Low birthweight. See Birthweight, low Lyons-Ruth, K., 253 Marital status on breastfeeding, 190, 191, 193, 194, 195, 197, 198 on depression, parental, 228, 229, 230, 232, 236 home production on, 26 on spanking, 158 on time spent with child, parent satisfaction with, 92–94 Maternal and Child Health Bureau, 315, 322–23 Medi-Cal Comprehensive Perinatal Services Program (CPSP), 200–201 Media, in new parent education, 100–101, 102, 103, 110 Medicaid, 399 Balanced Budget Act of 1997 and, 314–15

Index breastfeeding services in, 200–201 expansions of, 317, 377 for prenatal care, 267–68 well-child visits in, 323 Medications, lack of, 301, 303 Mental health, parental, 358, 386, 388–89, 393–95. See also specific disorders, e.g., Depression on parent–child activities, 131–32 Mental health care, child, lack of, 301, 303 Mental retardation, 376 Milk, human on brain development, 180 nutritive value of, 181 Miller, Representative George, 377 Minority groups. See Ethnicity/race Morbidities, new, 4 Mother, as primary caregiver, 90–91 Music listening to, punishment and, 164, 165 playing, with child, 32–33, 242–43 National Breastfeeding Policy Conference, 1998, 209 National Bureau of Economic Research, on early childhood intervention, 4 National Center for Children in Poverty, 370 National Commission on Children Beyond Rhetoric report, 5, 377 National Committee on Quality Assurance (NCQA), 317 National Committee to Prevent Child Abuse, 172 National Family Violence Surveys, 152–53 National Goals Panel, 117 National Health and Nutrition Examination Survey (NHANES), 273 National Health Interview Survey (NHIS), 273, 296–97. See also Health care, access to

Index National Longitudinal Survey of Youth (NLSY), 273 on discipline, 153 National Maternal and Infant Health Survey (NMIHS), 273 National Research Council, From Neurons to Neighborhoods report of, 5–6, 371–72 National Survey of Families and Households, 150, 153–54 National Survey of Family Growth (NSFG), 273 Neural Darwinism, 350 Neurobiology, of young children, 5 New Chance, 53 New models, of pediatric care. See Pediatric care, new models of New parenthood. See Parenthood, new Newborns’ and Mothers’ Health Protection Act, 271 North Carolina Smart Start Program, 6, 401 Obstetrical stays length of, 270–73 short, 281–85, 287. See also Discharge, early Omnibus Budget Reconciliation Act of 1981, 377 Oxytocin, 180 Parent(s) characteristics of, 380–84 child rearing activities of, 384–88 providing support to, 359 Parental education. See Education; Educational level, parent Parental employment. See Employment Parental health. See Health, parental Parental mental health, 358, 386, 388–89, 393–95. See also specific disorders, e.g., Depression on parent–child activities, 131–32 Parental readiness, 50–77 education in, parenting, 52–53, 58, 59, 61, 67–68, 76 ethnicity/race in, 56, 58, 59, 63, 71, 75

423 family structure in, 56, 58, 59–60, 62, 64, 65, 71 good parenting and, 72, 74 income in, 56 index of, 72, 73 kin involvement in, 54–55 mothers’ age and experience in, 57, 58, 60, 71 parents’ family experiences in, 57, 58, 60 in parents of firstborn, 57, 58, 64, 71 policy implications on, 74–77 pregnancy intention in, 51–52, 58, 59, 61, 65–67, 77 psychological resources in, 55–56, 58, 59, 70–72 risk factors for, 75–77 social support in, access to, 54–55, 58, 59, 62, 65, 68–69, 76 survey analytical techniques on, 60–61 survey indicators and variables on, 57–60 Parental resources fundamentals of, 25–28 human capital, 36–37 psychological, 23, 24, 34–36, 55–56, 58, 59, 70–72 purchased, 26, 28–31 time, 23–24, 26, 31–34 Parental warmth, 119 Parent–child activities, conflict-related, depression and, 241–42, 243–50 Parent–child activities, positive, 117–42, 383 child age in, 137 combinations of, 140 depression on, parental, 131–32, 241–43, 246, 247, 249–50 ethnicity/race on, 139–40 following daily routines, 119–20, 124–26, 129–30, 242–43, 245, 250, 252, 385–86 hugging and cuddling, 119, 123, 128–29, 242–43, 385 income on, 138–39 linguistic, 141

424 Parent–child activities, positive (cont.) by mothers vs. fathers, 138 parent age in, 138 parental education on, 138 playing, 32–33, 119, 122–23, 127–28, 137, 242–43, 385 reading, shared book, 118–19, 121–22, 126–27, 140–41, 384–85 Parenthood, new, 83–113, 115–16 advice from family on, 97–100 advice from health professionals on, 99–100, 101, 110 advice/information from media on, 100–101, 102, 103, 110 chi-square and ANOVA results on, 115–16 coping in, 105, 111 division of daily care in, 89–92, 108–9 division of responsibilities in, 84–85, 108–9 employment status on, 90–91, 93–94, 96, 109, 110, 111–12 frustration in, 106–8, 111 frustration in, depression and, 107, 111, 389 information, support, and advice on, sources of, 85–87, 110–11 information from hospital on, 97–98, 100, 110 parents’ perceptions of their abilities in, 105–8 parents’ satisfaction with time spent with child in, 92–96. See also Time spent with child perceptions of parenting in, 87 self-confidence in, 104–5 variables in, independent, 88 variables in, parenting, 87–88 Parenting. See also Families competency of, 34, 35 coping in, 389 employment in, 381 ethnicity/race on, 380–81 feelings and behavior in, 35 financial status on, 381 information on, 85–87, 97–103 instrumental conditions for, 381

Index on life at home, 34, 35 new, 87. See also Parenthood, new normal values in, 369 responsibilities for, division of, 84–85, 89–92, 108–9 risk factors for, 381–82 social expectations for, 369 social support networks for, 86, 111 style of, 24 Parenting classes. See Education, parenting Pediatric care, new models of, 347–62, 391–92 advocating for the basics in, 359–60 behavior problems in, identifying, 356–57 brain development in, and early learning understanding, 349–52 child development in, identifying problems of, 356–57 child development in, promotion of, 355–56 child development services in, expanded, 355–57 costs and benefits of, 361–62 expanded and enhanced care in, effectiveness of, 353–54 health in, parental, 352–53 Healthy Steps in, 360–62 income disparity in, 349 new team member in, 360–61 other services in, establishing links to, 358–59 social change in, pace of, 348–49 social disparity in, 349 support to parents in, providing, 359 two-generation approach to child health in, 357–58 Pediatric Pathways to Success, 360–62 Pediatrics rise of, 321–22 visits per year in, 347 Permissive parenting style, 149 Personal Responsibility and Work Act (PRWA), 371

Index Personal Responsibility and Work Opportunity Reconciliation Act of 1996, 186 Physical abuse, of parent on discipline, 152 on parental depression, 230, 232, 235, 236, 238 on spanking, 158, 161 Physical punishment, 149–51, 383. See also Discipline injury from, 172 Physician services. See also Education; Guidance, anticipatory; Pediatric care; Pediatrics use of, 306 Play, parent–child, 32–33, 119, 122–23, 127–28, 385 child age in, 137 depression and, 242–43, 245, 250 punishment and, 164, 165 Point-and-say routine, 119 Policy, 397 community, 400 health, 397 health care access, 312–16 health care access, enhanced services, 315–17 health care access, traditional services, 313–15 public. See also specific policies and legislation public, breastfeeding, 205–6 public, early childhood, 5–6 social, 404 social, family, 404–7 social, since 1960, 374–80 Poor. See also Income; Poverty working, 30 Postpartum hospital stay length of, 270–73, 281–85, 287 short, demographic differences in, 281–83 Poverty. See also Economics; Income child, 2, 38, 378 on discipline, 154–55 early childhood, 2 in negative outcome, 29

425 on parent–child activities, 138–39 Prebirth events, stressful. See also Child abuse; Physical abuse; Sexual abuse on parental depression, 228, 230, 232, 235, 236, 238, 248 Pregnancy intention, 51–52, 58, 59, 61, 65–67, 77 Pregnancy timing, on parental depression, 230 Prematurity ethnicity/race on, 280 income on, 280–81 prenatal care on, 270, 280–81 Prenatal care, 28–29, 263–65, 390 on breastfeeding, 190, 192, 197–200 childbirth education on, 268–69, 279 content of, 264 coordination and organization of services for, 268 delivery on, vaginal vs. cesarean, 269–70, 279–80 disparities in, 286 educational level on, maternal, 267, 276, 278 ethnicity/race on, 267, 275–76, 278 health insurance on, 267–68, 276–77, 278–79 hospital stay length and, postpartum, 270–73, 281–85, 287. See also Discharge, early income on, 267–68, 274–75, 277, 278 maternal age on, 275–76, 278–79 measuring, 266 policy implications for, 286–87 on prematurity and low birthweight, 270, 280–81 social factors in, 274–76 survey methods on, 273–74 use of, 274–79 use of, addressing barriers to, 267–68, 286 use of, multivariate analysis of, 277–79 use of, recent trends in, 266–67 women’s health and, 265–66

426 Prenatal service providers, on breastfeeding practice, 199–200 Prescription medications, lack of, 301, 303 Proposition 10, California, 6, 379, 401 Psychological resources, 23, 24, 34–36 in parental readiness, 55–56, 58, 59, 70–72 Psychopathology, child, parental depression on, 221–22 Psychosocial issues, pediatrician role in, 322–24. See also Guidance, anticipatory Public health, 392–93 Public policy. See also specific policies and legislation on breastfeeding, 205–6 in early childhood, 5–6 Public spending, 37–39 Punishment. See also Discipline; Physical abuse acceptable vs. unacceptable, 147 context of, 164–70 criteria for judging, 147–48 vs. discipline, 147 Kagan’s model of, 147–48 physical, 149–51, 155 Purchased resources, 26, 28–31. See also Financial resources key patterns of, 33 Race. See Ethnicity/race RAND on economics of early childhood intervention, 4 Investing in Our Children report of, 1, 380 Reach Out and Read (ROR) program, 142, 356, 385 Readiness parental. See Parental readiness school. See School readiness skills Readiness index, 72, 73. See also Parental readiness Reading, learning to, early failure in, 351–52

Index Reading, to child, 32–33, 385 child age in, 137 by child care provider, 141 depression and, parental, 245, 250 by parent, 118–19, 121–22, 126–27, 140–41 by pediatricians, 141–42 punishment and, 164, 165 Regalado, M., 324 Reiner, Rob, 370 Resilience, children’s, 40 Resource allocation, 21–46 child development models in, 22–24 financial, 23 government, 37–39, 45 human capital, 23, 24 number of children on, 27–28 other sources of support in, 39 parental: fundamentals, 25–28 parental: human capital, 36–37 parental: psychological, 34–36. See also Psychological resources parental: purchased, 26, 28–31, 34 parental: time, 23–24, 26, 27, 31–34, 45 psychological, 23, 24 risk factors in, 39–44 survey on, 24–25 Resources, on child development, 23. See also Education; Guidance, anticipatory; specific types, e.g., Parental resources Responsibilities, division of, in new parenthood, 84–85, 89–92, 108–9 Risk factors, 22 for parental readiness, 75–77 in parenting, 381–82 Risk factors, resource allocation, 39–44, 45 demographic, 41 health, 41, 44 identification of, 40–41 mental health, 41. See also Mental health multiple, 39–40, 42–44

Index Risk reduction strategies, 374, 375 Rockefeller, Senator John D., 377 Routines, family daily, 119–20, 124–26, 129–30, 385–86 depression and, 132, 133, 242–43, 245, 250, 252 Safe Start, 370 Sampson, Robert, 3 School readiness skills on breastfeeding, 206–7 home environment on, 140 strategies to improve, 401–3 Self, sense of, 148 disruptions to, 148 Self-confidence, 389 maternal, 55, 58, 62 of new parents, 87, 104–5 on parental depression, 228, 231, 233, 237, 240–41 Service organization, 404, 405. See also specific organizations Sexual abuse, of parent childhood, 57, 58, 60, 62, 63, 71 on depression, parental, 230, 232, 235, 236, 238 disciplining after, 152 on spanking, 158, 161 Shoving, 152 Singing, with child, 32–33, 385 Slapping, 152, 155 parental depression and, 242, 246, 248 Sleeping position, 97fn Smart Start Program, 6, 401 Smith, J.R., 154–55 Smoking, maternal, 265 Social capital, 39 Social disparity, on pediatric care, 349 Social outcomes, disparities in, 2 Social policies, 397–400, 404–7 since 1960, 374–80 Social skills, parental depression on, 220–21 Social structures, on child development, 3, 371

427 Social support, 54 access to, 54–55, 58, 59, 62, 65, 68–69, 76 for new parents, 86, 111 for new parents, advice in, 97–100, 110–11 on parental depression, 228, 231, 233, 237, 239–40 Spanking, 147, 150, 152, 153–54, 155, 157–63, 172–73, 386–87 aggression and, 150 depression and, 157, 161, 163, 242, 245, 248, 252 Specialty stores, early childhood, 369 Starting Points, 5, 117, 320, 369–70, 401 State Child Health Insurance Program (SCHIP), 399 Stern, D.N., 148 Straus, M.A., 153 Stress in parental depression, 217 prebirth. See Child abuse; Physical abuse; Sexual abuse prebirth, on parental depression, 228, 230, 232, 235, 236, 238, 248 Stressors, response to, 350–51 Success by Six, 6 Sudden infant death syndrome (SIDS), 97fn, 398 Support. See specific types, e.g., Social support Sure Start Program, 373 Survey of Income and Program Participation (SIPP), 273 Survey of Parents with Young Children, 26 Swearing, 153 Taking things away, 160, 242, 248 Teenage mothers, parental readiness in, 57, 58, 60, 71 Teenage Parent Demonstration, 53 Temporary Assistance for Needy Families, 26, 33, 112–13 Time out, 155, 171–72, 242, 248 Time resources, 23–24, 26, 27, 31–34, 45

Index

428 Time spent with child, parent satisfaction with, 92–96 child age on, 96 linear regression predicting, 94 in single-parent families, 94–96 in two-parent families, 92–94 Time spent with child, right amount of, 32, 33 Title XXI, Social Security Act, 314– 15 Toys, “brain stimulating,” 350 Trajectories developmental, 371, 374, 395 life course, interventions on family relationships in, 5 The Unfinished Agenda, 380 Uninsured. See Health insurance Urban Institute’s National Survey of America’s Families, 138–39 U.S. House of Representatives Select Committee on Children, Youth and Families, 377 U.S. UNICEF Baby Friendly Hospital Initiative (USBFHI), 202–3 Vaginal birth after cesarean, 270 Vaginal delivery, 269

postpartum stay length in, 281–82 van IJzendoorn, M.H., 11 War on Poverty, 376 Warmth, parental, 119 Welfare reform, 33–34, 112–13. See also specific legislation on breastfeeding, 186 Well-child care, 322–24. See also Guidance, anticipatory White House Conference on Early Childhood Development and Learning, 370–71 Why Child Care Matters, 380 Women, Infants, and Children Supplemental Feeding Program (WIC), 182, 189, 204–5, 376 Work. See Employment Work relationships, on parenting, 3 Working poor, 30 World Health Assembly Resolution 47.5, 205 Years of Promise, 117 Yelling, 152, 153, 155, 242, 246 parental depression and, 242, 245, 248, 252 Zero To Three, 370