The Cycle of Deviant Behavior
LONGITUDINAL RESEARCH IN THE SOCIAL AND BEHAVIORAL SCIENCES An Interdisciplinary Series Series Editors: Howard B. Kaplan, Texas A&M University, College Station, Texas Adele Eskeles Gottfried, California State University, Northridge, California Allen W. Gottfried, California State University, Fullerton, California THE ADDICTION-PRONE PERSONALITY Gordon E. Barnes, Robert P. Murray, David Patton, Peter M. Bentler, and Robert E. Anderson
THE CYCLE OF DEVIANT BEHAVIOR Investigating Intergenerational Parallelism Howard B. Kaplan and Glen C. Tolle Jr.
DRUG USE AND ETHNICITY IN EARLY ADOLESCENCE William A. Vega, Andres G. Gil, and Associates
DRUGS, CRIME, AND OTHER DEVIANT ADAPTATIONS Longitudinal Studies Edited by Howard B. Kaplan
FROM ADOLESCENCE TO ADULTHOOD IN THE VIETNAM ERA Timothy J. Owens
PREMARITAL PREDICTION OF MARITAL QUALITY OR BREAKUP Research, Theory, and Practice Thomas B. Holman and Associates
RESILIENCE AND DEVELOPMENT Positive Life Adaptations Edited by Meyer D. Glantz and Jeannette L. Johnson
IN SYNC WITH ADOLESCENCE The Role of Morningness-Eveningness in Development Anna-Karin Andershed
TAKING STOCK OF DELINQUENCY An Overview of Finds from Contemporary Longitudinal Studies Edited by Terence P. Thornberry and Marvin D. Krohn
TEMPERAMENT Infancy through Adolescence Diana Wright Guerin, Allen W. Gottfried, Pamella H. Oliver, and Craig W. Thomas
A continuation Order Plan is available for this series. A continuation order will bring delivery of each new volume immediately upon publication. Volumes are billed only upon actual shipment. For further information please contact the publisher.
Howard B. Kaplan Glen C. Tolle, Jr.
The Cycle of Deviant Behavior Investigating Intergenerational Parallelism
Howard B. Kaplan and Glen C. Tolle, Jr. Texas A & M University Department of Sociology College Station, TX 77843-4351 USA
[email protected] [email protected]
Library of Congress Control Number: 2006923502 ISBN-10: 0-387-32643-X ISBN-13: 987-0387-32643-6
e-ISBN 0-387-0-387-32644-8
Printed on acid-free paper. © 2006 Springer Science+Business Media, LLC All rights reserved. This wok may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. Printed in the United States of America. 9 8 7 6 5 4 3 2 1 springer.com
(SPI/SBA)
To the first generation, Diane the next generation, Samuel, Jenny, Rachel, Jeremy and the one after that, Calvin Austin Kaplan Howard B. Kaplan
To my wife, April, my family, and friends for their love, support, and patience Glen C. Tolle, Jr.
Preface
This volume is about understanding the relationship between deviance and selected correlates of deviance in one generation and deviance and its selected correlates in the next generation. By examining the significance of these constructs in the parental generation as part of the explanation for the same constructs in the child’s generation, we contribute to an understanding of the phenomena. This contribution, however, is quite limited in the sense that we are examining in essence bivariate relationships—the association between first-generation and second-generation phenomena— while ignoring all of the other influences on the second-generation phenomena that do not stem from or account for the intergenerational relationship. Nevertheless, the study of intergenerational parallelism of deviance and its correlates justifiably has excited attention and resulted in a voluminous literature greater than might have been expected for any particular bivariate relationships because of the mystique surrounding ideas—cycle of violence, reproduction of culture, to name but a few—that are evoked by consideration of the association between such phenomena in one generation and the same phenomena in a successive generation. Because of the implications of studies of intergenerational parallelism for understanding continuities in deviant behavior and its antecedents and, indeed, for understanding culture stability and change, we systematically describe and elaborate upon simple observations of cross-generation parallelisms in deviant behavior and its putative antecedents. The elaborations take the forms of (1) specifying the conditions under which intergenerational parallelism is increased (i.e., circumstances that express or facilitate commonalities of experience) or vii
viii
Preface
decreased (i.e., circumstances that express or increase dissimilar life experiences), (2) specifying variables that mediate the association between first- and second-generation deviance, and (3) identifying variables that exercise intragenerational influences in each generation and manifest intergenerational continuity across generations. These systematic elaboration strategies along with baseline descriptions of the association between deviance in one generation and deviance in the succeeding generation compose a logic of procedure (i.e., methodology) for investigating intergenerational parallelism in deviant behavior and related phenomena. It will be recognized that, at a higher level of abstraction, this methodology reflects the procedures to be followed in investigating any bivariate relationship, not only intergenerational parallelism, to wit—establishing a temporal relationship between two variables, specifying the conditions under which the relationship is strengthened or weakened, decomposing the temporal linear relationship by modeling mediating variables, and attenuating the temporal linear relationship by identifying common antecedents of the two variables. The logic of procedure outlined above is applied to the study of intergenerational parallelism of deviant behavior and its correlates using data from a multigeneration panel study. In Part I, the salient concepts, theoretical framework, and methodology are described. In chapter 1, we discuss the concepts of deviance and intergenerational parallelism at length and outline in broad strokes the theoretical framework that guides the elaboration strategy for studying intergenerational parallelism of deviance. In chapter 2, the nature of the multigeneration panel study is described. Analytical techniques, and any other specific methodological details that are applicable to all of the ensuing chapters reporting findings, are presented. Procedural details that are unique to only one or some of the chapters reporting findings will be presented in the chapters in question. In chapter 2, we also consider methodological limitations in the extant literature on intergenerational parallelism. In Part II, we consider intergenerational parallelism of deviant behavior as a conditional relationship. In chapter 3, the literature on intergenerational parallelism of deviant behavior and selected correlates of deviant behavior is described and discussed. We present findings on intergenerational parallelism for our major indexes and put these in the context of the general literature on intergenerational continuity. In chapter 4, we note that the literature reports highly variable degrees of intergenerational parallelism of deviance and suggests that the strength of the association
Preface
ix
between first- and second-generation deviance is conditional on a number of circumstances. In this chapter, we report theoretically informed analyses that specify conditions said to impede or facilitate intergenerational parallelism. The findings are discussed in the context of the guiding theoretical framework and the literature bearing on the moderators of the relationship between first- and second-generation deviance. In Part III, analyses are presented that statistically decompose the linear associations observed between first- and second-generation deviance. In chapter 5, analyses are presented and discussed in the context of relevant literatures in which the observed association between first- and second-generation deviance (or salient correlates of deviance) is, in part, accounted for in terms of intervening processes. In these analyses, firstgeneration deviance (or its correlates) is interpreted as having a causal impact on some intervening variable that, in turn, is interpreted as having a causal influence on second-generation deviance (or its correlates). In chapter 6, analyses are reported in which observed relationships between first- and second-generation deviance (or its salient correlates) are explained partly in terms of intergenerational continuities between firstand second-generation variables that have a causal impact on deviance (or its salient correlates) in their respective generations. The continuities might be causal in nature such that a cause of deviance in the first generation has consequences that increase the likelihood of a comparable cause of deviance in the second generation. Alternatively, sociocultural or physical context associated with the genesis of first-generation deviance might continue to be shared by both generations. In Part IV, the results of the previous chapters are summarized and their implications for future research are considered. Chapter 7 reviews the results with regard to their contributions to the understanding of deviant behaviors, in particular, and more generally to the dynamics underlying intergenerational parallelism, evaluates the procedures followed in investigation of intergenerational parallelism as a worthy elaboration strategy for the study of any bivariate relationship, and outlines a research agenda for the future that promises to increase our understanding of the meaning of earlier-generation deviance for the occurrence of deviance in successive generations.
College Station, Texas
Howard B. Kaplan Glen C. Tolle, Jr.
Acknowledgments
This work was supported in part by research grants R01 DA02497 and R01 DA010016 and by a Research Scientist Award (K05 DA00136) to the first-named author, all from the National Institute on Drug Abuse. We are pleased to recognize the unwavering dedication of Holly Groves, “Sam” McLean, and Joseph Hall to the technical production of the manuscript. Where we recognize the origins of our ideas we acknowledge these sources by appropriate textual citations. However, many of our ideas, particularly as these are synthesized in our comprehensive theory of deviant behavior, are the end products of lifetimes of scholarly activity. The precise sources or originality of these ideas can no longer be determined. Thus, often we must leave it to others to make judgments about the historical roots, originality, or independent creation of the theoretical statements in this volume.
xi
Contents
Preface Acknowledgments
Part I
vii xi
Theoretical and Methodological Issues in the Study of Intergenerational Parallelism of Deviance 1 1. Toward an Understanding of Intergenerational Parallelism of Deviance 2. Method
Part II The Conditional Nature of Intergenerational Parallelism 3. Observing Intergenerational Parallelism in Deviance 4. Moderators of Intergenerational Parallelism
Part III Decomposing Intergenerational Parallelism 5. Intervening Processes 6. Intergenerational Continuity of Intragenerational Causes of Deviant Behavior and Its Correlates
Part IV Retrospect and Prospect 7. Summary and Conclusions References Index
5 41
57 59 73
125 127 157
193 195 217 235
xiii
Part I THEORETICAL AND METHODOLOGICAL ISSUES IN THE STUDY OF INTERGENERATIONAL PARALLELISM OF DEVIANCE
Part I introduces the intersection between two salient interests of social scientists—the etiology of deviance and intergenerational replication of psychosocial phenomena—by describing the conceptual/theoretical framework and methodology employed in this investigation of intergenerational parallelism of deviance. Chapter 1 deals with the conceptual issues and theoretical framework that guide the empirical investigation, the results of which are reported in Parts II–III. Deviant behavior is defined in terms of failure to conform to the prescriptions and proscriptions that compose specified socionormative systems that are taken to be the reference standards for evaluation. Intergenerational parallelism refers to the de facto correlation between an attribute or behavior pattern by a member of one generation and the same attribute or behavior pattern of their children assessed at the same developmental stage in each generation. The term intergenerational parallelism is used because, unlike other terms such as intergenerational influence, it does not connote assumptions about the nature of the explanation for the intergenerational correlation. Influence implies a causal connection such that the attribute in one generation has consequences that, in turn, cause the presence of the attribute in the children when the intergenerational correlation in fact might 1
2
Part I
be accounted for in terms of contextual continuities that have like outcomes in each generation. The integrative theory that guides the analyses to be reported in subsequent chapters is outlined. The theory in effect brings together a number of extant frameworks for the study of deviance. The presentation makes explicit relationships between these specific theoretical approaches and the integrative theory, and the relevance of these theories (and, by implication, the integrative theory) for explaining intergenerational parallelism in deviance. The guiding theoretical framework variously identifies processes that moderate the existence or strength of the intergenerational parallelism in deviance, variables that mediate the influence of first-generation deviance on second-generation deviance and intergenerational continuities that have like influences on deviance in each generation. The identification of these phenomena in the aggregate serves to explain a particular bivariate relationship of interest: the association between deviance in one generation and deviance in the next generation. At the same time, the procedure of systematically estimating theoretically informed moderators, mediators, and causally significant intergenerational continuities serves as a template for addressing explanations of any bivariate relationship. In chapter 2, the limitations of the literature on intergenerational parallelism, in general, and on intergenerational parallelism of deviance, in particular, are considered. A multigenerational panel study is described as an antidote to certain of the more salient limitations, namely the frequent failure to use representatives of the multiple generations as separate data sources (rather than having representatives of one generation provide data for both generations) and the failure to have the data reported by the earlier and later generation subjects at comparable stages in the life course (rather than, for example, having the first generation provide self-reports as parents and the second generation provide self-reports as children). The dataset to be used in estimating the theoretically informed models describing or explaining intergenerational parallelisms in deviance is described as consisting of self-reports of patterns of deviance and their correlates provided by a sample of eighth graders and by their children at roughly the same developmental stage. The variables variously reflect (1) patterns of deviant behavior (or their correlates), (2) putative moderators of the association between deviant patterns in the two generations, (3) variables that mediate the observed intergenerational parallelism in deviant behavior, and/or (4) putative common intergenerationally continuous antecedents of
Theoretical and Methodological Issues
3
deviance in each generation. The models will be estimated employing structural equation modeling methodologies focusing on three latent constructs reflecting deviant behavior, disposition to deviance, and negative self-feelings, respectively.
1 Toward an Understanding of Intergenerational Parallelism of Deviance
In broad terms, the purpose of this study is to contribute to an understanding of why people engage in deviant behavior. More particularly, the objective is to describe and explain the part played by deviance in an earlier generation in accounting for deviant behavior in a later generation. What is the relationship between deviance in an earlier generation and deviance in a later generation? Where a significant association is observed between deviance in one generation and deviance in the next generation, how is this relationship to be explained? Are there specifiable conditions that facilitate or impede the existence of such a significant correlation or that increase (decrease) the strength of the association between first- and second-generation deviance? Does first-generation deviance have identifiable consequences that, in turn, increase (decrease) the probability that second-generation deviance will occur? Is the observed correlation between deviance in successive generations accounted for, in part, by intergenerational continuity of identical intragenerational antecedents of deviance? This project should not be taken to be more or less than it is. In its simplest sense, the aim of the project is to account for deviant behavior in a population of second-generation youths during late childhood/early adolescence. Further, the focus is on the import of one explanatory factor, namely similar behavior by the parents of that child at approximately the same developmental stage. The objective of the study, then, is to examine 5
6
Chapter 1
the nature of the relationship between deviant behavior by a population of subjects and similar behavior by the children of those subjects assessed at similar points in the life course and to elaborate this intergenerational parallelism by examining (1) circumstances that moderate the strength of the relationship, (2) variables that mediate any observed intergenerational continuity in deviant behavior, and (3) the intergenerational parallelism of common intragenerational predictors of youthful deviance. One the one hand, the investigation is narrow in that in effect, the focus is on only one explanatory variable (first-generation youthful deviance), but, on the other hand, the nature of this relationship is elaborated greatly by specifying moderating, intervening, and common antecedent variables. This exercise, in addition to informing us about the dynamics of intergenerational continuity in deviant behavior, provides a template for fully investigating the extended meaning of any particular explanatory variable in its relationship to any particular dependent variable. In the present chapter we discuss the concepts of deviance and intergenerational parallelism, outline the theoretical frameworks that guide the empirical analyses, and provide the elaboration strategy that serves as an organizing framework for conduct of the analyses and presentation of the findings.
Defining Deviance and Intergenerational Parallelism The terms deviance and intergenerational parallelism are rife with conceptual ambiguities that should be resolved before embarking on reports of empirical findings.
The Nature of Deviance Although the behavior patterns that are the foci of the several analyses in this volume are widely recognized in the more inclusive society as examples of deviant behavior, the concept of deviance has more general applicability. The concept might refer to failure to conform to expectations in the context of a wide variety of interpersonal systems, including friendship groups, marital dyads, and work groups, as well as the general community. Indeed, the concept might refer to behaviors or attributes that conform to the expectations of one group but violate the expectations of
Toward an Understanding of Intergenerational Parallelism
7
another group from whose perspective the judgment of deviance is made. Deviance might refer to physical traits, social identities, experiences, behaviors, or a variety of other phenomena that describe a person. Deviance, then, refers to behaviors or attributes manifested by specified kinds of people in specified circumstances that are judged to violate the normative expectations of a specified group. “Shared normative expectations” refers to group evaluations regarding the appropriateness or inappropriateness of certain attributes or behaviors when manifested by certain kinds of people in certain situations. The indication that certain patterns of behavior in certain contexts (i.e., manifested by certain kinds of people in certain situations) are socially defined as deviant is the administration of negative sanctions. Members of a group that are said to share a normative system impose these negative sanctions, which are responses that, according to the perceptions of the group, are intended to serve as punishment for the (absence of the) attributes or behaviors in question. The consistent application of relatively severe sanctions in response to particular kinds of behavior serves as an indication that those behaviors are deviant according to the normative system that serves as a reference point. If the sanctions are applied only to certain kinds of people who perform the behaviors in question, then the implication is that the behaviors when performed by other kinds of people are not defined as deviant. If certain behaviors evoke negative sanctions regardless of the person’s social characteristics and other situational contexts with perhaps very rare exceptions, then the implication is that the behavioral (prescription) proscription is generally applicable except in extenuating circumstances. The ranking in the hierarchy of normative expectations of the evaluative standard that is violated by the behavior is reflected in the severity of the sanctions. Behaviors or attributes are deviant not only because they evoke negative sanctions but also because they would evoke negative sanctions if representatives of the socionormative system that defines the attributes or behaviors as deviant became aware of them. A group that shares a normative system might evaluate the behaviors or attributes of individuals who do not belong to the group and might apply negative sanctions for behaviors or attributes that are judged to deviate from the normative expectations believed to be incumbent on even nongroup members. Depending on the group’s access to sanctions that are meaningful to the nongroup members, the application of negative sanctions might have a great adverse impact on the outcomes of nongroup members.
8
Chapter 1
In some cases, individuals whose behaviors or attributes appear to deviate from the normative expectations of a group to which they do not belong are not judged to be deviant. This is because the individuals (perhaps because of their perceived inferiority) are not expected to be capable of conforming to the normative expectations. These “barbarians” or “subhumans” are judged to be deviant by virtue of not belonging to the group that evaluates them, but they are not otherwise punished for failing to conform to the specific normative expectations that define the shared normative system. At worst, the failure to conform to the normative expectations is taken to be a (further) indication of their primary deviation (i.e., not being part of the group that shares the normative system). In any case, the valuation of behaviors or attributes as deviant presumes that those making the judgment have taken into account the applicability of the normative expectations to the person and, more particularly, to the circumstances in which the person finds himself or herself. It is not required that the deviant actor identify himself or herself as a group member for the group to evaluate the actor’s attributes or behaviors. The judgment that certain behaviors or attributes deviate from the normative expectations might be made even if the deviant manifestations are beyond the individual’s control. Every normative system offers examples of evaluative standards that stigmatize individuals for manifesting undesirable attributes or behaviors or for failing to manifest desirable attributes and behaviors that are beyond their control. The de facto deviation from the expectations of specified normative systems, thus, might be motivated or unmotivated. Motivated deviance derives from either of two sets of circumstances. In the first set, the person is a member of a group that defines the attributes or behaviors in question as deviant. Because of his experiences in the group, the person loses motivation to conform to the normative expectations of the group and adopts deviant attributes or behaviors that consciously or unconsciously are intended to serve self-enhancing functions. In the second set of circumstances, the person is a member of a group in which the attributes or behaviors under consideration are normatively prescribed. The person is motivated to conform to the normative prescriptions as one who has been socialized in the group. The person is either unaware or considers it to be irrelevant that another group judges the attributes or behaviors to be deviant. The behavior is motivated, but the fact that the behavior or attributes are deviant does not contribute to the motivation. Rather, the person’s motivation (when the performance of behavior is problematic
Toward an Understanding of Intergenerational Parallelism
9
rather than habitual) stems from the need to conform to internalized group standards and to evoke approving responses from group members who share these standards for conforming to the group’s normative expectations. Unmotivated deviance refers to instances of failure to conform to the normative expectations of the person’s membership or reference groups where the failure to conform is contrary to the person’s volition. The person would conform if he or she were able to conform. The circumstances that contribute to unmotivated deviance are discussed in some detail below (see pp. 20–22). The patterns of deviance that are the subject of the analyses of this volume generally fall into the category of motivated deviance. Unmotivated deviance is relevant as an explanatory factor, rather than as a dependent variable. The involuntary possession of traits, and the involuntary performance of behaviors, that are defined as deviant influences judgment of deviance, the administration of sanctions, and correlates of these phenomena that, in turn, influence the onset of other deviant acts or the continuity of the deviant behaviors at the voluntary level. The models describing and explaining intergenerational parallelism in deviant behavior, the estimate of which will be reported in later chapters, focus on a latent construct that reflects behavior patterns widely viewed as deviant in our society. The latent variable, deviant behavior, is reflected in indicators of youthful theft, intergenerational violence, and substance use. Two other constructs will be modeled in explanatory frameworks as well. One of these reflects disposition to deviance, which indicates attitudes rather than specific behaviors, and demonstrates a tendency to disrespect others, to respond to provocations with antisocial responses, and to express disrespect for conventional patterns. The other construct reflects negative self-feelings as is reflected in self-derogating attitudes, anxiety, and depressive affect. These constructs are important explanatory constructs in accounting for deviant behavior. However, within a broad definition of deviance, they might be considered examples of deviance. As we have defined the term, deviant behavior encompasses these constructs as well as other explanatory variables such as school failure, abusive parenting, and stigmatized personal or social characteristics. Any or all of these might be implicated in the explanation of intergenerational parallelism in deviance, whether as moderating or mediating variables or as antecedents of deviance in each generation that are intergenerationally continuous. For example, intergenerational parallelism in deviant
10
Chapter 1
behavior might be stronger for first-generation stigmatized groups that feel alienated from and do not respond to conventional social controls. Thus, their behavior might be more available for modeling by their children. As an example of mediating influence, first-generation deviant behavior might lead to deviant parenting that, in turn, motivates the second-generation youths to engage in resentful antisocial responses. Intergenerational continuity in deviant dispositions or negative self-feelings that within each generation influences the onset of deviant behavior might account for observations of intergenerational parallelism in deviant behavior. Deviance could be expressed in relatively broad or specific categories of concrete behavioral manifestations. Thus, intergenerational parallelism in deviant behavior could be expressed in terms of broad or specific categories. In the latter case, for example, children of parents with schizophrenia might be significantly more likely to manifest schizophrenia than children whose parents did not have schizophrenia or children of parents who were diagnosed as alcoholics might be significantly more likely to develop problems with alcohol use than individuals whose parents were not alcoholics. In the former case, it might be observed that individuals whose parents had some form of mental illness were more likely to develop some form of mental illness, but not necessarily the same form as that of their parents, compared to individuals whose parents did not have some form of emotional disorder. As Avison (1999) observed, reviewing the literature on intergenerational transmission of several forms of emotional disorder: In summary, the studies reviewed examining the functioning of the offspring of parents with schizophrenia, depression, or alcoholism suggest that these children are at risk for developing a variety of emotional problems. Clearly, there is not a one-to-one correspondence between parental diagnosis and child functioning. Thus, for example, children of depressed parents not only exhibit higher rates of depression, but also higher levels of conduct disorder, global psychiatric symptoms, and multiple psychiatric diagnoses. Similarly, children of alcoholic parents are not only at elevated risk for alcohol and drug use, but also for diagnoses of major depression and anxiety. Finally, it is important to note that there are common problems exhibited by children in all three groups of parental disorders. These problems include difficulties at school, temper tantrums, headaches, problematic social functioning, and emotional disorder. (p. 500)
For present purposes, a more general construct of deviant behavior is adopted—a latent construct reflecting any of a diversity of acts encompassing theft, violence, and various patterns of substance use.
Toward an Understanding of Intergenerational Parallelism
11
Intergenerational Parallelism The degree to which one generation reproduces or is reproduced by another has long excited the interest of observers of the human condition. As Oliver (1993, p. 1316) noted: “It is hard to think of many topics so central to society in general and psychiatry in particular as the intergenerational transmission, by one means or another, of human attributes.” Generally, the discussion focuses on the opposition of forces that lead to psychological and social stability across generations, on the one hand, and influences that bring about intergenerational individual and social change, on the other hand. Thus, as Atkinson and Dodder (1990) observed: The longstanding concern over generational conflict and continuity may be because it is through generational transmission that culture is continued and the world of today becomes the world of tomorrow.... Generational transmission acts in two major ways to pass on culture. First, it serves as the mechanism through which values, attitudes, and ways of life are kept alive. Secondly, it serves as the major mechanism for social change. Each generation determines what aspects of human culture will be retained from the previous generation and what aspects of culture will be discarded or changed, determining what will be kept as tradition and what will be discarded and forgotten. (pp. 193–194)
Others have observed as well that the generations are continually subject to a tension between a pressure to replicate the earlier generation and circumstances that facilitate generational change. Bengtson (1987), commenting on parenting, grandparenting, and intergenerational continuity observed: Parents invest prodigious amounts of time, energy, and material resources in the uncertain hope of producing offspring who will be happy, healthy, and wise—and who will, hopefully, validate at least some of their parents’ principles. But their offspring are continually changing, as is the social environment within which they are growing up. One of the things parents strive for is continuity: indicators in the behavior of their children that they have achieved transmission of what is best or better than in their own lives. One of the things children strive for is distinctiveness: a better way of life, a more successful social order. (p. 435)
Thus, we observe a good deal of intergenerational continuity manifest in genetic and social parallelisms between parents and children, facilitated by the force of tradition. At the same time, however, according to Bengtson (1987),
12
Chapter 1 ...there is the process of generational turnover and replacement. This implies change—both biological and social. Children are not mere replicas of their parents; they represent subtly new genetic combinations in interaction with a unique developmental environment (Rossi, 1980).... The product of unique sociohistoric influences, children are the carriers of new perspectives and commitments that represent the potential for change in the existing social order (Elder, 1978). (p. 435)
The significance of events at any time and in any place is measured by the nature and moment of the adaptive responses by individuals or collectivities to these events. The extent to which adaptations endure is a function of the transmission of the responses over time from one generation to another and/or of the stability of the circumstances that evoke such responses. To understand the continuity of patterns of human social behavior over time is to understand the mechanisms through which adaptations are transmitted and the forces that maintain the circumstances that evoke similar response patterns over time. Conversely, to understand individual or social change is to understand the mechanisms that impede intergenerational transmission and/or that modify the conditions that demand personal or collective adaptations. Unfortunately, the investigation of these mechanisms is made more difficult by identifying the phenomenon under investigation with putative mechanisms that account for the phenomenon. Terms such as intergenerational transmission and intergenerational continuity are frequently used to signify both the phenomenon to be explained and the explanation of the phenomenon. Normally, intergenerational studies focus on individual- or social-level patterns that are observed to be similar in successive generations. The intergenerationally comparable patterns are explained in terms of the direct or indirect causal influences of the first-generation pattern on the second-generation pattern, intergenerational continuities on circumstances that have comparable causal effects on the patterns of interest within each generation, or factors that moderate the intergenerational influences or continuities. In order to avoid confusion, it is proposed to signify the phenomena to be explained (intergenerationally comparable patterns) as intergenerational parallelisms. Terms such as intergenerational continuities will be reserved for the discussion of the mechanisms that are intended to explain intergenerational parallelisms. Intergenerational parallelism refers to comparable cognitions, feelings, or behaviors across generations. The term intergenerational parallelism
Toward an Understanding of Intergenerational Parallelism
13
is employed in order to avoid prejudging the nature of the mechanisms that result in intergenerational parallelism. To employ a term such as intergenerational transmission is to imply a “process through which purposively or unintendedly an earlier generation psychologically influences parenting, attitudes and behavior of the next generation” (Van Ijzendoorn, 1992, pp. 76–77). Parallelism might be accomplished without such direct or indirect causal influence of the first generation on the second generation. As Van Ijzendoorn (1992) observed: If grandparents and parents have been rearing their children in about the same physical and social circumstances, their childrearing behavior and attitudes may be more alike, but the earlier generation may not have exerted any direct, psychological influence on the next generation’s parenting (Quinton & Rutter, 1984). (p. 77)
This latter phenomenon reflects intergenerational continuities. For present purposes, then, we will use intergenerational parallelism as a description of comparable phenomena observed in successive generations. This term will in itself have no implications for the mechanisms through which the intergenerational isomorphism occurs. We will use the term intergenerational transmission to refer to causal mechanisms (whether genetic or social-psychological) through which phenomena observed in the first generation influence comparable phenomena in the second generation. Intergenerational transmission might also refer to the influences of phenomena observed in one generation on noncomparable phenomena in the second generation, although these processes will be subordinate to our predominant interests in comparable intergenerational phenomena. The term intergenerational continuity will be reserved for instances in which intergenerational parallelisms are accounted for in part by intergenerationally stable or transmitted phenomena that have like causal outcomes on the phenomenon of interest in each generation and so, in part, account for intergenerational parallelism. In focusing on intergenerational parallelism, by no means is it suggested that the only, or even the major, cause of a phenomenon in the second generation is the same phenomenon observed in the first generation. For example, in one study only 24% of subjects who ran away earlier in life had parents who had also run away earlier in life (Plass & Hotaling, 1995). If the purpose were to explain the second-generation behavior or attributes, numerous variables other than the first-generation behavior or attribute would be incorporated as explanatory variables. Nevertheless,
14
Chapter 1
our purpose here is to examine the conditions under which intergenerational parallelism is or is not present and the circumstances under which the degree of intergenerational parallelism increases or decreases, as well as the explanation of the continuity in terms of mediating and intergenerationally (dis)continuous contextual variables. In examining the intergenerational parallelism phenomenon, the focus is, in effect, only on one predictor of a second-generation outcome. There are many other predictors of the second-generation phenomenon. However, for the moment, the spotlight is on one particular antecedent in an attempt to understand how that first-generation phenomenon operates to influence the second generation. As Egeland (1988) observed with regard to the intergenerational transmission of child abuse: The data presented here indicate that the parents’ histories of having been abused as children should be included as a major risk factor for maltreatment of their children. It must be stressed, however, that there are many additional risk factors, including the families’ life circumstances and stressors and a range of parental characteristics and caretaking skills. For purposes of early identification and prediction, a combination of risk factors must be considered. (p. 96)
Nevertheless, our purpose here is to examine the conditions under which intergenerational parallelism is or is not present and the circumstances under which the degree of intergenerational parallelism increases or decreases, as well as the explanation of the parallelism in terms of mediating and intergenerationally (dis)continuous contextual variables. Although deviant behavior in the earlier generation is only one predictor of deviant behavior in the succeeding generation, it still represents an important part of the explanation of deviant behavior in the next generation. For example, regarding the intergenerational transmission of abusive behavior, Pears and Capaldi (2001) reported: The 23% rate of intergenerational transmission found in this study is comparable to Kaufman and Zigler’s (1987) best estimate for the rate of transmission of 30%. This once again illustrates that having a history of abuse is not a guarantee that one will become abusive. However, those parents who children reported being abused were twice as likely to have been abused themselves than to have had no such history, confirming that having been abused is indeed a risk factor for transmitting abuse (Egeland, 1993). (p. 1454)
The remainder of this volume is devoted to describing, elaborating, and explaining patterns of intergenerational parallelism in deviance and two selected theoretically informed salient antecedents of deviance.
Toward an Understanding of Intergenerational Parallelism
15
Guiding Theoretical Framework Several different theoretical frameworks have been employed to interpret instances of intergenerational parallelism in deviance. All of these approaches are said to be incorporated within an integrative theory of deviant behavior that informs the analyses to be reported in Parts II and III of this volume. Here, in turn, several theoretical interpretations of observed instances of intergenerational parallelism are presented, the guiding integrative theory of deviant behavior is summarized, and the ways in which other theoretical approaches have been incorporated into the guiding explanatory framework are made explicit. Diversity of Explanatory Frameworks Explanations of intergenerational parallelism of deviance or its correlates have implicated various permutations and combinations of theories, sometimes focusing on only one or another framework. The following are presented only to illustrate the range of theoretical frameworks that are offered. Some approaches focus, in part, on the alienating effects of environmental stress and on social learning theories whereby abusive behavior (for example) is imitated, observed, or reinforced (Kaufman & Zigler, 1993; Muller, Hunter, & Stollak, 1995). Mechanisms relating to social learning, for example, have been considered with regard to the intergenerational transmission of early childbearing (Barber, 2001a): Socialization is one important mechanism that may produce an association between mothers’ and their children’s childbearing behavior. Existing research has suggested that children born to young mothers may hold attitudes, values, or preferences that are more favorable toward early childbearing, and this is why they reproduce their mothers’ behavior (Anderton et al., 1987; Kahn & Anderson, 1992; Thornton, 1991; Thornton & Camburn, 1987). (p. 221)
Social control theories have also been implicated in explanations of intergenerational transmission of early childbearing. Barber (2001a), for example, stated: Social control is another important mechanism that may explain the intergenerational transmission of first birth timing. This perspective stresses that the extent to which mothers are able to control their children’s dating,
16
Chapter 1 sex, and related behaviors influences how quickly they become parents. The hypothesis is that young mothers, particularly if they are unmarried, are less able to control their adolescents’ behavior, and thus the adolescents are more likely to engage in risky sexual behavior and experience of premarital pregnancy. For instance, Hogan and Kitagawa (1985) illustrated a strong effect of parental supervision on young Black women’s risk of a premarital birth. (p. 223)
The relevance of social stress-related variables are frequently noted in conjunction with (perhaps as an antecedent of) processes that appear to implicate social control. For example, variables that have been suggested as mediating intergenerational transmission of early age of childbearing— financial and marital instability—are interpretable as suggesting that these variables attenuate the child’s emotional dependence on the family and, therefore, the family’s ability to exercise control over the child’s behavior. Barber (2001a) further stated: Two consequences of early childbearing produce undesirable family circumstances that are particularly likely to propel children into early parenthood: financial instability and marital instability. The hypothesis is that family disruptions may weaken the child’s relationship with his or her parent or encourage the child to seek emotional support and intimacy outside of the family of origin (Wu & Martinson, 1993). Characteristics of the family of origin that make staying at home less attractive, such as low or unstable income, are similarly likely to propel young adults out of the family and into families of their own (Michael & Tuma, 1985; Wu, 1996). (p. 224)
While attenuating social controls, a disposition to spend more time away from the family might also increase opportunities to engage in risky sexual behavior that results in early childbearing. The relevance of labeling for understanding the social reproduction of a criminal class is noted by Hagan and Palloni (1990), who observed that: ...modern formulations of labeling theory are oddly silent on...the occurrence of labeling across generations. This silence is surprising because George Herbert Mead (1918) anticipated a focus on not only intergenerational labeling but also the notion of class reproduction in his early article ‘The Psychology of Punitive Justice’. Mead called attention to the inconsistency of what we call labeling and deterrence theories and to the potential consequences of the indiscriminate use of legal sanctions. In doing so, the father of symbolic interactionism encouraged a structural perspective on the intergenerational effects of legal sanctioning, writing that “a system of punishments assessed with reference to their deterrent powers not only works very inadequately in repressing crime but also preserves a criminal class” (p. 583).
Toward an Understanding of Intergenerational Parallelism
17
Mead was concerned that our system of legal sanctions provides few if any mechanisms for the revocation of the stigma involved in criminal proceedings. The implication of Mead’s analysis is that crime is concentrated across a generation in a permanent class of criminals. Mead’s concern about a permanent criminal class focuses attention on processes of intergenerational labeling and suggests the possibility of a further contingency in labeling theory, namely, that deviant behaviors that meet labeling responses early in a behavioral career may be most likely to lead to subsequent deviant behavior when the early actor is the child of parents who are also labeled deviant. Said more simply, labels may be most likely to affect the behavior of adolescents where they are imposed in the context of a family that has previously been labeled deviant. (pp. 267–268)
These statements illustrate both the range of theoretical frameworks and their interdependence as explanations of instances of intergenerational parallelism of deviance and its correlates. The need to simultaneously consider several different theoretical perspectives is nicely illustrated by a case study of intergenerational transmission of deviant patterns by Dunlap and her associates (2002): This paper examines some of the social mechanisms for intergenerational transmission of subcultural norms drawing heavily on social learning theory, strain theory, and theories regarding identity formation. We contend that girls growing up within severely distressed inner-city households are continually confronted in many ways: physically, sexually, and visually by the use of drugs within their household. Within their environment, a child’s sense of safety, sense of self, sense of worth, and sense of any alternative future within conventional society is continually assaulted. They become inured to what comes next—a process of desensitization that Goffman (1961) described as mortification of the self. Once prepared in this manner, girls learn the prevailing conduct norms in their households regarding drugs, sexual exploitation and interpersonal violence by modeling what others did to them. As adults, they effectively replicate these behavioral conduct norms, re-initiating the mortification process upon the next generation. (p. 2)
Both the range of theories that have been used to interpret instances of intergenerational parallelism and the empirical interdependence of variables that reflect theoretical constructs in several of these theories stimulate the use of an integrative theory of deviant behaviors that, in effect, incorporates virtually all extant theoretical approaches to the study of deviance and has proven useful in informing inclusive multivariate models of the antecedents and consequences of deviant behavior (Kaplan, 1972, 1975, 1980, 1982, 1983, 1984, 1986, 1995; Kaplan & Johnson, 2001).
18
Chapter 1
An Integrative Theory of Deviant Behavior Intergenerational parallelism is studied in the context of a general theory of deviant behavior that addressed how people become motivated to engage in deviant behavior, the factors that influence acting out deviant dispositions, and the factors that influence intragenerational (dis)continuities or (de)escalation of deviance. As the theory is applied to an understanding of intergenerational parallelism in deviance, firstgeneration variables are modeled as (in)directly mediating or moderating the influence of first-generation variance on second-generation deviance or as common intragenerational influence on deviance in each generation that reflect or have consequences for parallelism across generations. Consideration is given in turn to processes that reflect or influence: motivation (or disposition) to engage in behaviors defined as deviant; acting out deviant dispositions; and (dis)continuity of deviant behavior. Motivation to Engage in Acts Defined as Deviant
The most immediate cause of deviant behavior is the motivation or disposition to engage in such acts. Motivation to behave in ways that are defined as deviant arise in either of two broad sets of circumstances. In the first case, a person is socialized to internalize standards for selfevaluation. Circumstances result in the person experiencing failure and rejection. These experiences result in distressful negative self-feelings that cause (1) alienation from (loss of motivation to conform to) the conventional normative system and (2) motivation to adopt alternative deviant patterns that promise to forestall experiences of failure and rejection or to assuage the negative self-feelings that result from such experiences. In the second case, the standards are internalized prescribed behaviors that are judged to be deviant by groups other than those in which the person holds membership, but that are evaluated as appropriate according to the standards of the person’s membership group(s). The person is motivated to engage in the behaviors (that are defined as deviant by other groups) because they conform to the expectations of other group members on whom the person depends for satisfaction of his/her quotidian instrumental and socioemotional needs and because they conform to self-imposed expectations as well. Failure to conform to these expectations would evoke rejecting attitudes from salient others
Toward an Understanding of Intergenerational Parallelism
19
and self-devaluing attitudes, either of which would result in distressful negative self-feelings. Each of these circumstances is considered in turn. Motivation to engage in acts that are deviant by the standards of one’s membership group(s) is the result of negative self-feelings that are the consequences of chronic failure to evoke positive attitudes toward oneself from significant others and to approximate salient self-evaluative standards. In the course of the normal socialization process, universally and characteristically the youth (based on childhood dependence on others for satisfaction of basic biological needs) develops, first, a need for the approval of significant others and, later, a need for self-approval. The child learns early on that the approval of others (and, ultimately, self-approval) is contingent on meeting (self-) evaluative standards regarding the possession of highly valued characteristics and the performance of role-appropriate behaviors. Indeed, one of those standards is the expectation of eliciting approving responses from significant others. The chronic failure to elicit approving responses from self and others, resulting from the failure to approximate the now internalized selfevaluative standards, results in the experience of stable negative self-feelings. Because the person’s experiences in conventional membership group(s) have culminated in self-perceptions of failure and social rejection and consequent negative self-feelings, the person loses motivation to conform to the normative standards of the conventional group(s) that have failed to provide opportunities to succeed and to evoke positive attitudes from others: self-approval and correlated positive self-feelings. Nor has conventional society provided the mechanisms that might have permitted the person to assuage the negative self-feelings that are the concomitants of failure and rejection. The inability to satisfy the need for self-approval (stimulated by the self-derogation associated with chronic failure and rejection in the course of conventional group experiences) motivates the person to seek and adopt alternatives to the conventional patterns that were ineffectual in satisfying the need to avoid or assuage negative self-feelings and to experience positive self-feelings. These alternative deviant patterns would serve these needs if the deviant patterns (1) facilitated achievement of conventional goals through illicit means, as when theft permits the procurement of articles that signify material success—a standard for positive self-evaluation, (2) permitted perceptual distortions that allowed the rejected and failed person to think well
20
Chapter 1
of himself/herself, as when drug misuse precludes self-awareness of failure and rejection or allows misperceptions of personal potency or grandiose achievements, (3) permited interpersonal avoidance of conventional others so as to forestall experiences of rejection and failure, as when the person seeks personal isolation or withdraws into his/her own personal world, and (4) resulted in the invalidation of the conventional standards according to which the person was judged to be a failure and rejected by self and others or that led to the adoption of an alternative deviant value system that offered more easily attainable standards as a route to self-approval, as when vandalism or interpersonal violence serve to symbolically or in fact attack representations or representatives of the conventional value system or when attraction to a deviant subculture provides a potentially more achievable set of standards (at the same time expressing contempt for conventional values) and precludes interaction with conventional others who have judged the person and found him/her wanting. The chronic experience of failure and rejection and of consequent distressful self-feelings, are a function of numerous circumstances, virtually all of which implicate the characteristics, experiences, and behaviors that compose the parental histories whether as genetically transmitted, constitutionally given, socially ascribed or transmitted, ecologically contextualized, or socially evaluated. The youth’s experiences of rejection and failure in large measure reflect, or are occasioned by, what has been labeled earlier as unmotivated deviance. That is, the youth possesses characteristics, performs behaviors, or suffers experiences that are contrary to his/her will but nonetheless, as a result, is forced to be the object of negative social sanctions that indicate to him/her outcomes worthy of judgments of failure and social rejection. The self-perception of these outcomes evokes negative self-feelings. Some of these unwelcome and sanctionable attributes are genetically transmitted/constitutionally given such as physical limitations, a less than amiable temperament, or psychological deficits related to cognitive or executive functions. Other intergenerational outcomes relate to a paucity of resources, including those related to level of wealth, education, reputation, and group membership. The absence of resources on the part of the parental generation increases the likelihood of failure and social rejection, the adoption of deviant adaptations, and consequent social stigmatization. The experience of stigmatization is suffered through association by the child as well.
Toward an Understanding of Intergenerational Parallelism
21
Still other parental experiences relate to the occurrence of stressful life circumstances and the lack of effective conventional coping, adaptive, or defense mechanisms to forestall or assuage the distressful consequences of the stressful life experiences. The inadequate coping patterns are transmitted to the child, who, in turn, deals with life experiences through the adoption of maladaptive and unconventional response patterns. At the same time, the experience of stressful life events by the youth is increased by the concomitants of the deviant lifestyle of the parents, which occasions family conflict, parental abuse and neglect, and stigmatization. Even under circumstances where the child accepts the family climate as normative, the youth is vulnerable to the negative social sanctions that function to deprive him/her of resources (individual and interpersonal, instrumental and socioemotional) that are prerequisite to the resolution of life crises. The net effect of such intergenerational influences is the genesis of selfrejecting feelings in the youth. Particularly where the experience of self-rejection is associated with conventional membership group experiences, the youth will attenuate any feelings of attraction to the conventional group(s) that he/she had and will become disposed to seek deviant alternatives as responses to distressful self-feelings. The second scenario under which youths become motivated to engage in behavior that is defined as deviance specifies the existence of shared normative expectations that endorse the behavior as conventional within the group although the behavior is defined as deviant by groups other than those in which the youths hold membership; perhaps more politically influential groups. The group that endorses the “deviant” patterns preexists the youths who come to share the expectations relating to the response pattern. The youth might be born into the group and be socialized to accept these patterns as normative by intergenerational parental influences, socialization into intergenerational continuous peer groups, or observation of the prevalence of the pattern throughout the population that is said to share the “deviant” pattern across generations. Alternatively, the youth might opt to adopt the deviant patterns later in life, having been attracted to the group as a more salubrious alternative to more conventional groups from which the youth was alienated because of the groups’ influence on the genesis of chronic experiences of rejection and failure and concomitant distressful self-feelings. The “deviant” group might offer the promise of a more easily achievable set of standards and
22
Chapter 1
consequent acceptance by others, escape from a more draconian socionormative system that offers the likelihood of continuing experiences of failure, and a shared antipathy to, and denigration of, the standards according to which the youth was judged to have failed and so suffered social rejection and distressful self-feelings. Acting Out Deviant Dispositions
Although youths might be motivated or disposed to engage in deviant activities, whether they in fact act out those motives is contingent on a number of circumstances. As noted earlier, in the normal course of socialization, the person develops the needs to elicit positive attitudes from significant others and to possess those characteristics and perform those behaviors that elicit those attitudes from significant others. Whatever the source of the deviant motivations, whether it stems from the experiences of failure and rejection in the groups that define the acts as deviant or from the perception that the acts are endorsed by one’s membership groups, the performance of the motivated acts depends on the need of the person to conform to the group’s expectations. If the person needs to evoke positive responses from the group members who view the act as objectionable, the youth will be inhibited from acting out his/her deviant disposition. However, if the person has attenuated emotional ties with group members, then their perceived disapproval of the deviant acts will not have an inhibiting effect on acting out the deviant disposition. Similarly, if the individual in the course of the developmental process internalized needs to approximate certain self-values such as being law abiding, then the perception of a projected act as contravening that standard would inhibit the acting out of the behavior in which the person is otherwise motivated to engage. However, if the need to be law abiding had not been internalized, then the recognition that the act was illicit would not inhibit the acting out of the deviant disposition; or, if the person needed to feel in control of his/her emotions, the recognition that the use of psychoactive substances would reduce self-control would threaten that self-value and inhibit the behavior in which the youth was otherwise motivated to engage, In the absence of that need, however, the anticipated loss of self-control would not in itself be inhibiting. Although some self-values such as those reviewed above are intrinsically valued, others are regarded as instrumental to the achievement of
Toward an Understanding of Intergenerational Parallelism
23
other ends. If the person anticipated that a deviant act that he/she was disposed to engage in would have consequences that deprived the person of the resources that were essential to the achievement of higher order values, the person would be inhibited from acting out the deviant motivation; but in the absence of recognition that receiving instrumental resources is contingent on not performing deviant acts, the person will not be inhibited from acting out the deviant disposition. The internalization of the needs to evoke positive attitudinal responses from conventional others (i.e., one’s membership group) and to approximate other self-evaluative standards depends on first-generation representatives effectively transmitting these needs to the youths of the next generation. Such effective transmission depends on these same representatives having, themselves, internalized these values, having the opportunity (including ongoing physical presence) to transmit the values, and being exposed to effective parenting techniques (including the ability to elicit positive affective identification from children). In the absence of those contingencies, the youth will not be socialized to need positive responses from others and to otherwise approximate salient self-attitudes; and, so, the youth will not be inhibited from acting out deviant dispositions, gained from whatever source. However, even in circumstances under which the youth was effectively socialized by conventional others, as had been suggested earlier, long-term experiences of failure and social rejection might attenuate emotional ties to conventional others and values and, thus, weaken the inhibitory influence on acting out deviant dispositions that might otherwise have been experienced. Even in the presence of inhibitory influences (i.e., where the youth is emotionally committed to the conventional individuals and standards), the youth might yet commit deviant acts if he/she is able to justify it in terms that are compatible with the broader social structure. Thus, interpersonal violence might be legitimated as a justifiable extension of prevalent patterns in social institutions such as those related to recreation (sports violence, television programming) or global political strategy (war); or, illicit activities by adolescents might be rationalized as youthful exuberance. The acting out of deviant dispositions, in addition to requiring the absence of inhibiting circumstances and the presence of rationalizations that are compatible with conventional socionormative systems, also requires opportunities to act out motivations to engage in deviant acts. The opportunity to engage in deviant acts requires resources, whether
24
Chapter 1
these take the form of a ready supply of illicit drugs, an unguarded house, an available victim of hostile urges, or other physical, personal, or interpersonal prerequisites. The opportunity to act out deviant dispositions also encompasses the situational context that provides the occasions or stimuli for the deviant acts, as when instruments of violence interact with predispositions to aggression to instigate interpersonal assault or when recreational settings stimulate drug use. The situational context and deviant motivation are mutually influential. Whereas the context facilitates activation of deviant motives, the disposition to deviance instigates the search for facilitating resources and occasions to commit the deviant acts. The opportunities to engage in deviant acts are a function of numerous factors, including the range of deviant acts that might reflect the disposition and the extent of the person’s involvement in conventional social roles. If the person is disposed to engage in any deviant act simply because of its function as a symbolic rejection of the conventional order from which the person feels alienated, then a wide range of opportunities would facilitate acting out this general predisposition. However, if drug dependency were the motivating force, only a narrower range of opportunities would suffice to permit acting out the predisposition, namely drug-related resources and supportive interpersonal situations. The greater involvement one has in playing conventional social roles, the less time a person who is otherwise motivated to engage in deviant behavior has to become aware of or participate in situations conducive to engaging in these activities. Indeed, immersion in normative roles might provide alternative conventional pathways to forestalling experiences of failure and social rejection or to the reduction of the distress that accompanies such experiences. (Dis)continuity of Deviant Behavior
Whether a person continues or discontinues, increases or decreases, performance of deviant acts following initiation of deviant behavior is influenced by the more or less direct consequences of the initial performance (e.g., self-enhancing feelings, personal injury, approval by deviant peers, stigmatization by conventional others) and by changes in the person’s circumstances (e.g., developmental processes across the life course) that influence the strength of motives that mitigate the disposition to deviate and the opportunities to act out those motives.
Toward an Understanding of Intergenerational Parallelism
25
Continuation or Escalation of Deviant Behavior
Continuation or escalation of deviant behavior is influenced by three sets of variables: those that influence (1) positive reinforcement of deviant acts, resulting from the satisfaction of salient needs that are traceable to the initial deviance, (2) the weakening of social constraints that might have forestalled continuity or escalation of initial deviance, and (3) the establishment or continuity of opportunities to engage in deviance. Virtually all of these implicate intergenerational processes. Regarding positive reinforcement, deviant behavior satisfies important needs of the individual, some of which precede and motivate the initial deviance, and another of which is a consequence of stigmatizing social responses to the initial deviance that motivate continued or enhanced deviance as a solution to the resultant need for self-justification. The experience of the satisfaction of needs that preceded and motivated the initial deviance increases the likelihood that the deviance will be repeated (or escalated) when the need recurs in the anticipation that satisfaction of the need will ensue once again. The several needs that might be said to motivate the deviant acts relate to the more general need for positive self-feeling. This need is occasioned by self-derogatory evaluations that stem from related experiences of social rejection and failure from significant others in the person’s life. The deviant acts might satisfy the person’s need for positive self-feelings in any of a number of ways, including (1) permitting the attainment of conventional self-evaluative standards through illicit means, (2) facilitating intrapsychic or interpersonal avoidance of circumstances that occasion experiences of rejection and failure, (3) attacking, symbolically or otherwise, the validity of the socionormative system that occasioned self-perceptions of failure and social rejection, and (4) substituting a more forgiving and achievable set of deviant evaluative standards for the conventional one(s) that occasioned self-derogatory attitudes in response to judgments of failure and concomitant rejection by self and others. The performance of deviant acts, in addition to satisfying preexisting needs, also creates need. By virtue of evoking stigmatizing responses to initial deviant acts as formal and informal sanctions, the deviant actors develops the need to assuage the distressful self-feelings associated with being an object of stigmatizing responses. The deviant actor accomplishes this by self-justification, or through redefinition of the deviant act in positive rather than negative terms, and by evidencing the “positive” value of
26
Chapter 1
the “deviant” acts through repetition or escalation of these acts. The deviant actor, in the former instance, uses conventional rationalizations of the behavior, and in the latter case, the actor rejects the validity of the “conventional” socionormative order in favor of one centered on deviant value, identities, and behaviors. As Coleman (1986, p. 225) suggested, the positive reevaluation of deviant identities and behaviors permits the deviant actor to “regain their identity through redefining normality and realizing that it is acceptable to be who they are.” Having come to positively value the deviant identity, the “deviant” actor is motivated to conform to the expectations that the “deviant” patterns are good by conforming to these expectations through the repetition or escalation of these acts. The evaluative redefinition of deviant behavior is facilitated by the weakening of emotional ties to the conventional society and entry into a network of deviant actors that provides social support for the adoption of deviant identities and the correlated performance of deviant acts. In addition to the positive reinforcement of deviant acts, a second circumstance that increases the likelihood of continued or escalation of deviant behavior is the weakening of social controls that previously worked to forestall the acting our of deviant impulses. Social controls operate through engendering personal needs to avoid adverse consequences of deviant acts (e.g., incarceration, loss of resources, rejection by significant others) or to attain valued consequences (e.g., respect from valued others, material rewards, a sense of satisfaction from having a law-abiding personal identity). These controls will be weakened as effective constraints against deviant behavior if engaging in deviant acts is not perceived as a threat to the satisfaction of these needs or if the deviant actor no longer needs these outcomes. In the former case, deviant behaviors will not be considered a threat if adverse consequences are not observable either because they do not in fact occur or because the deviant actors have been expelled from conventional interactions and thus are not longer vulnerable to the experiences of adverse sanctions from conventional others. The deviant actors, because of the social distance created by negative sanctions, are no longer able to be aware of any sanctions being administered. In the latter case, the need for conventional social approval and the trappings of conventional achievement are weakened by the same processes that endangered the initial disposition to engage in the deviant acts. The experiences of rejection and deprivation of resources that follow upon initial deviance, along with the past and ongoing experiences of
Toward an Understanding of Intergenerational Parallelism
27
failure and rejection in the conventional world, lead to defensive rejection of the persons and standards that are associated with the world perceived as responsible for the distressful self-feelings that accompany the adverse outcomes. In self-defense, the deviant persons no longer care about what conventional others think of them or about the standards that are used to judge the deviants (and that were once used by the deviants to judge themselves). No longer needing the good opinions and resources of the conventional world, the deviants are no longer restrained from continuing to engage in or escalating involvement in deviant acts by the fear that they will not achieve the approval of conventional others and success, as this is defined by conventional standards. Along with positive reinforcements that result from deviance and the attenuation of social controls that might have constrained the acting out of deviant impulses, increased opportunities to engage in deviant acts influence the continuity or escalation of initial deviance. A number of factors conspire to increase the interaction between those who engage in initial deviance and deviant group members. First, the nature of sanctions that are administered in response to initial deviance (incarceration with other deviants, expulsion from school, avoidance by conventional others) increases the opportunities for (or constrain) interaction with deviant peers while (by) decreasing the likelihood of interaction with conventional others. Second, the stigmatizing sanctions administered by conventional others decrease attraction to conventional groups and increase the attractiveness of deviant groups as potential sources of satisfaction of needs that have not been, nor are likely to be, satisfied in the future by conventional groups. Third, having been stigmatized by virtue of the initial deviance and as a result of being the object of punitive responses, the deviant person appears to deviant others as an appropriate recruit into their deviant networks. The increased interaction with deviant networks increases the opportunities (as well as social support) for deviant acts. The youth becomes socialized through the process of observation and positive reinforcement. The “recruit” learns new patterns of deviance and is provided with a setting that occasions such actions. Just as positive reinforcement of deviant behaviors, the attenuation of social controls, and increased opportunities influence the continuation or escalation of deviant behavior, so it is to be expected that the lack of positive reinforcement of deviant behavior, the continuity of conventional social controls, and increased opportunities to satisfy needs
28
Chapter 1
through conventional patterns would decrease the opportunity to pursue deviant mechanisms for need satisfaction. Thus, whether the deviant individuals were born into and socialized within deviant networks or were recruited into such networks as deviant adaptations to experiences of failure and rejection in conventional groups, the failure of the deviant networks and behaviors to enhance positive self-feelings or, indeed, the tendency of deviant groups to be sources of failure and rejection would decrease the disposition to continue to engage in (deviant) behaviors that failed to enhance self-feelings or were the source of negative self-feelings (Kaplan & Lin, 2000a, 2005). Thus, discontinuance of or decrease in deviant behaviors would ensue following recognition of the continuity or onset of negative self-feelings associated with the deviant adaptations. Given the expectations that conventional mechanisms were available that gave hope of self-enhancing outcomes, the deviant actors would be good candidates for eschewing deviant adaptations in favor of conventional ones. Discontinuity of deviant adaptations is likely to occur also under circumstances whereby the continuity of deviance begins to threaten new needs that theretofore were not relevant, as when becoming a parent leads to the commitment to fulfilling associated role expectations—a need that would be threatened by ongoing deviant acts. In any case, despite the alienation engendered by experiences of failure and rejection in conventional groups, the needs to conform engendered by early and intense socialization processes would not be completely obviated; thus, contemplation of continued deviance would generate, at the very least, ambivalence and, at most, antipathy. Thus, especially given the possibility of successfully pursuing conventional values, the deviant actor is vulnerable to forces contraindicating continuity or escalation of deviance. Finally, if the person’s needs, or the opportunities to meet those needs, change, the use of deviant mechanisms to meet one’s needs might be rendered obsolete. With maturation, the adoption of adult roles might provide opportunities for success (e.g., through occupational pursuits) that were theretofore denied the deviant actor; with the assumption of occupational, marital, and parental roles, the time and related opportunities to become aware of and adopt deviant alternatives would be diminished. This outline is provided only as a brief review of the complex integrative theory of deviant behavior that guides the analyses to be reported in chapters 3–6. Other sources provide more detailed coverage (Kaplan, 1972, 1975, 1980, 1986, 1995, 1996; Kaplan and Johnson 2001). This
Toward an Understanding of Intergenerational Parallelism
29
theory effectively integrates a large number of special theories that address particular aspects of explanations of deviant behavior. Our intention was not to integrate these theories. However, following the implications of our initial premises, it was inevitable that the general theory would encompass these other perspectives. By way of justifying the use of the integrated theory to inform the empirical analyses, the ways in which the integrative theory overlap with numerous extant more specific theories are highlighted, particularly regarding their applicability to the explanation of intergenerational parallelism. The Integrative Theory and Other Deviance Theories The guiding theoretical framework generally is recognized as encompassing a range of traditional theories of deviance. Thus, Akers (2000) stated: Howard B. Kaplan (1975) proposed a self-esteem/self-derogation theory of adolescent deviance that brings together deviant peer influences (social learning theory), family and school factors (control theory), dealing with failure to live up to conventional expectations (strain theory), and self-concept (symbolic interactionism and labeling theory). (p. 248)
Earlier, Wells (1978, p. 190) cited this general theory as one of two major exemplars of the rapprochement between two theoretical viewpoints— respectively named structural interactionism and socialization-control analyses—that in the 1960s were among the “dominant perspectives on the use of the self concept in the study of deviance.” Structural interactionism focused on deviance in subcultural terms as a collective response to social variables. For Cohen (1955), gang delinquency emerged as a collective response to status-frustration that resulted “from the intersection of social dysfunction and the fundamental motivation of people to enhance or validate their self-identities through social interaction” (Wells, 1978, p. 190). This work was an attempt, then, to articulate motivational, interpersonal, and situational considerations, on the one hand, with social structural theories (notably anomie theory). Further attempts to link social structural conditions (differentially distributed resources, experiences, and values) with the interpersonal events by which they are produced took the form of self-role theory. Within this perspective, a self-identity is said to arise within ongoing social interactions that are, in turn, influenced by social structures of available or
30
Chapter 1
appropriate identities. The socially organized sense of self guides the construction of new actions and is influenced by others’ responses to past actions. Wells cites, as examples of this approach in the study of deviance, motivational models of deviance adoption (Cohen, 1965, 1966; Cohen & Short, 1966) and differential identification theory emphasizing the selective action of significant others or valued reference groups on reflected appraisal. From the second perspective, socialization-control analyses—and particularly in the early and later development of containment theory (Reckless, 1967; Schwartz & Tangri, 1965)—self-concept was treated as a developing personality variable rather than as an interactional process intervening between individual and group-level events. A favorable selfconcept, as a mechanism of inner containment and as an insulator against deviance, was introduced to explain deviance along with the previously propounded social control factors (i.e., socially institutionalized structures of inducements and constraints). In the course of the socialization process, the development of a favorable self-concept functions to inhibit dispositions toward deviance. The synthesis between structural interactionists and socializationcontrol analysis would appear to be the outgrowth of the recognition that social structural differences are produced through the distribution of socialization experiences and role-learning opportunities, as well as through the distribution of resources. “It reemphasizes that socialization is a social process consisting essentially of interpersonal associations and patterns which connect individuals interactively to larger social contexts” (Wells, 1978, p. 194). Both Kaplan (1975) and Hewitt (1970), the other exemplar, are said to relate self-concept and deviance to the social structure by the following proposals: 1. Commitment to the normative order is influenced by adequacy of level of self-esteem. 2. Level of self-esteem is a cumulative product of socialization experiences variously distributed across different social sectors or interpersonal associations.
When the situational structure is unable to maintain an acceptable level of self-esteem, the person will be disposed to seek individual or collective deviant alternatives that might provide more positive experiences. The actual adoption of the alternative activities will also be a function of the external structure of rewards, opportunities, and resources.
Toward an Understanding of Intergenerational Parallelism
31
The comprehensive nature of the general theory under consideration is apparent when attention is directed toward any of several traditional perspectives on deviance, including strain, subculture/differential association, control, and labeling theories/perspectives. Several of these are considered, in turn, with particular attention paid to their relevance, and therefore the relevance of the integrative theory, for explaining instances of intergenerational parallelism of deviance. Strain-Related Theories
From the structured strain perspective (Cloward & Ohlin, 1960; Merton, 1938), deviant responses are viewed as outcomes of the disjunction between culturally prescribed (and personally internalized) goals and institutionalized means for achieving these goals. Compatible with this view, in the general theory such disjunctions reflect both the failure to achieve culturally valued goals and the absence of instrumental resources. These, in turn, increase the likelihood of pervasive self-rejection and the consequent adoption of deviant responses that are consciously or unconsciously intended to enhance self-attitudes. Among the available mechanisms for accomplishing this is the rejection of culturally prescribed values. Thus, the theory is particular congruent with Cohen’s (1955) view of delinquent gangs as collective solutions to the frustration of being unable to attain conventional goals by conventional means. The solution, of course, is to reject the legitimate structure. More recently, Agnew (1992) provided a more inclusive version of strain theory that went beyond equating strain with the disjunction between goals and means. Agnew conceived of three categories of strains, which included inability to achieve positively valued goals, loss of positively valued relationships and other stimuli, and the experience of being the object of negative actions by others. This version also is compatible with the general theory insofar as all of these “strains” implicate threats to self-acceptance and increase the probability of self-derogation. With regard to intergenerational parallelism, social stress theories, and more particularly, self-derogation theory, also operate via intervening mechanisms, intergenerational parallelism of intragenerational causes of deviant behavior, and representation of moderating influences. Thus, the consequences of deviant behavior, including evoking distressful social sanctions, might mediate the intergenerational parallelism of deviant behavior. Alienation resulting from negative social sanctions leads to
32
Chapter 1
performance of the deviant behavior by adult first-generation subjects and thus provides the occasion for modeling such behavior by the secondgeneration children. With regard to intergenerational parallelism of intragenerational antecedents of deviance, self-derogation in each generation and related psychological distress motivate deviant adaptations. The psychological distress is intergenerationally continuous insofar as selfderogation forestalls the learning of effective conventional coping mechanisms that otherwise would be transmitted to the next generation. Self-derogation also serves as a moderating variable insofar as the association of self-derogation with deviant behavior in the first generation might weaken the intragenerational parallelism of deviant behavior because it has failed to assuage self-derogation and, in fact, might contribute to it. The attenuation of intragenerational deviance, in turn, diminishes the occasions for modeling of deviant behavior by the next generation. Subcultural/Social Learning Theories
Once Cohen’s (1955) collective solutions to shared frustrations are viewed as a stable response system, this perspective might be classified as either a subculture or a differential association theory (Akers, 1973; Miller, 1958; Sutherland & Cressey, 1974), in that “deviant” responses reflect de facto conformity to shared normative expectations in particular membership groups. In the context of the general theory of deviant behavior under consideration, such responses are viewed similarly as the outcome of normal socialization processes whereby the individual conforms to expectations in order to earn self-acceptance and correlated benign outcomes, such as acceptance by other group members. The failure to conform to these group expectations is deviant from the group perspective and initiates the chain of events leading to motivated deviant responses (i.e., responses that purposely contravene the group’s expectations). Motivated deviance would not ordinarily result from conforming to the group’s expectations (which happen to be judged as deviant from the perspective of other groups), except under conditions where the other groups, as a result of their judgment that the behavior in question was deviant, had the power to engineer outcomes for the subject that were sufficiently self-devaluing in their implications to more than counterbalance the self-enhancing effects of the subject’s commitment to the normative structure. According to the stress paradigm, intergenerational parallelism in deviance is mediated by the induction of stress by the parents’ deviance.
Toward an Understanding of Intergenerational Parallelism
33
The substance-use-induced stress in the children evokes the previously learned stress-reduction response. In this instance, the stress paradigm presumes a social learning model as well. In the social learning model, the transmission of the deviance pattern is direct. Either deviance by the first generation is modeled by the second generation or deviance as a stressreducing mechanism is learned directly by the second generation. In short, the situational context of deviance is learned. Social learning theory and subcultural theory in the context of the integrative theory provide for the purposeful teaching of deviant patterns to children, the modeling of observed deviant patterns, and related processes reflecting “cultural or characterological processes by which parents, through child-raising conditions and practices reproduce in their children the characteristics that lead to crime” (Hagan & Palloni, 1990, p. 266). Subcultural or social learning theories might also suggest that circumstances that facilitate the teaching and modeling of deviant behavior within each generation are themselves intergenerationally continuous. Thus, a criminogenic environment might facilitate the genesis of deviant dispositions in one generation and be continuous across generations so that the same or a similar environment influences deviant dispositions in the second generation as well. Social learning theory and subcultural theory also suggest moderating influences for intergenerational parallelism. Thus, in social environments where particular deviant patterns are epidemic, deviant patterns performed by representatives of one generation are more likely to be modeled by the second generation. Social Control/Bonding Theory
In the context of control theories (Briar & Piliavin, 1965; Hirschi, 1969; Polk & Halferty, 1966) and containment perspectives (Reckless, 1967; Reckless, Dinitz, & Murray, 1956; Schwartz & Tangri, 1965; Voss, 1969), impulses toward deviant responses are checked by a variety of socially induced factors. Control theories propose that such factors are related to a person’s emotional commitment to the normative order based on earlier gratifications achieved in the course of the normative socialization process. Containment theory introduces a favorable self-concept as an insulator against deviance, in addition to the earlier noted social control factors that inhibit the expression of deviant dispositions. These theories, in effect, are incorporated under the general theory in a number of ways. To the extent that a subject’s gratifications (particularly
34
Chapter 1
noteworthy among which are positive self-feelings) are associated with the normative structure, the subject will develop positive affective ties to the normative structure, in addition to the self-concept of one who appropriately seeks valued goals through legitimate means. Conversely, to the extent that normative relations are associated with self-rejecting feelings and correlated adverse outcomes, the person’s emotional ties with the normative structure will be attenuated, as will the need to think of himself or herself as a person who conforms to normative expectations. Given the attenuation of these controls, the awareness of alternative deviant routes to self-enhancement and associated ends is likely to eventuate in the onset of deviant responses. The adoption of deviant responses further attenuates one’s ties to the social order insofar as the deviant actor becomes the object of negative social sanctions, thus motivating rejection of the normative system in order to justify the earlier deviant responses. Social control theories are also implicated in explanations of intergenerational parallelism of deviant behavior. As an intervening variable, deviant behavior might be thought of as evoking negative social sanctions (i.e., social control responses). These responses, contingent on a number of moderating conditions, either cause desistance or escalation of deviant behavior. Escalation of deviant behavior would occur when the social control mechanisms alienate the individual from the society by foreclosing opportunities to rejoin the normative structure in any rewarding fashion. The alienation from conventional society would attenuate the efficacy of social controls on the first-generation subject. The ensuing performance of deviant acts would then be modeled by the second-generation subject (contingent on the degree of identification of the second-generation subject with the first-generation subject). Under conditions where the individual identified strongly with the normative structure, being the object of negative social sanctions might decrease the subsequent performance of deviant acts and so decrease the modeling of the deviant behavior by the child. Regarding the intergenerational parallelism of intragenerational antecedents of deviant behavior, being the object of negative social sanctions has consequences within each generation for the escalation of deviant behavior (under specified conditions). Being the object of negative social sanctions in response to deviant behavior in one generation has consequences for the parallelism of being the object of negative social sanctions insofar as the labeling effect of negative social sanctions alienates the subject from conventional society and so attenuates the efficacy
Toward an Understanding of Intergenerational Parallelism
35
of the social controls. The increased deviant behavior would be modeled by the second-generation child and would, in turn, increase the probability of being the object of negative social sanctions in the second generation. As a moderating variable, being the object of negative social sanctions (i.e., being amenable to social control) would modulate the degree of intergenerational transmission of deviant behavior. Under conditions where social control was strong, whether formal or informal in nature, intergenerational parallelism in deviant behavior would be attenuated. Labeling Theory
Aspects of a fourth theoretical perspective are also implicit in the general theory. The labeling perspective (Becker, 1963; Kitsuse, 1962; Lemert, 1951; Scheff, 1966) focuses on responses to initial deviance by agents of social control who, in effect, define the social identity and, thereby, the self-identity of the actor as deviant. A deviant self-identity, in turn, influences the stabilization of deviant careers. These aspects of the labeling perspectives that are implicit in the general theory include propositions regarding the influence of labeling on the onset of deviant responses, the self-enhancing/self-devaluing consequences of labeling, and the influence of labeling on the stabilization of deviant response patterns. Regarding the first aspect, the subject’s anticipation of the reaction of others (including labeling) to proposed deviant adaptations in the service of the self-esteem motive will influence his or her expectations regarding the net self-enhancing/self-devaluing consequences and thereby his or her adoption of the deviant response. Regarding the second implication, the labeling phenomenon has self-enhancing effects to the extent that it facilitates (1) avoidance of self-devaluing membership group experiences by attenuating the relationship between the subject and the membership groups in which self-rejecting attitudes develop, (2) the attack of the basis of the subject’s self-rejection since acceptance of the label symbolizes the subject’s opposition to the very normative expectations that were the basis of his failure, and, (3) substitution of self-enhancing opportunities for selfdevaluing experiences by permitting the deviant identity to attract positive attitudinal responses from others similarly labeled. Insofar as selfenhancing effects are experienced, the subject will gain a positive emotional investment in the deviant identity.
36
Chapter 1
Finally, a third implication of the labeling phenomenon relates to the parallelism of deviant careers. Quite apart from any influence on the parallelism of a deviant pattern exercised by the resultant positive emotional investment in the deviant identity, labeling might be hypothesized to have at least two other influences on parallelism. First, a deviant label raises the psychological costs of attempted reentry into the normative system where, were it not for these labeling-induced barriers, the subject might have adopted newly available, normative, self-enhancing or self-protective mechanisms. Second, societal labeling increases the subject’s felt need to justify his or her initial deviant responses in order to assuage the selfrejecting feelings deriving from the negative labeling experience. Insofar as this justification is successful, repetition of the deviant act is facilitated. Labeling theory also has implications for explaining intergenerational parallelism in deviant behavior. As an intervening variable, being the object of negative social sanctions labels an individual as deviant and so alienates the person from the social order as a consequence of being so labeled. The labeling, in turn, leads to escalation of the first-generation subject’s deviant behaviors and so provides more circumstances for modeling the deviant behavior by the second-generation child. Within each generation, the labeling process influences deviant outcomes (depending on the person’s expectations of regaining rewarding entry into the conventional world). The labeling experience itself is intergenerationally continuous by virtue of the increased first-generation deviance and the modeling of that deviance by the second-generation subject. The intergenerational labeling experience is also mediated by the sins of the father being visited on the son—that is, by virtue of the secondgeneration subject being stigmatized because of the stigma associated with the first-generation parent. Intergenerational (dis)continuity in deviant behavior is moderated by experiencing labeling by virtue of the deviant behavior. Where labeling occurs and the individual has no expectation of being accepted back into conventional society, intergenerational continuity of deviant behavior is likely to be greater. Where labeling does not occur and social support for the deviant behavior is present in the individual’s interpersonal world, continuity of deviant behavior is likely to be increased. For the individual who experiences labeling and anticipates being accepted by the conventional world, the likelihood of discontinuity of deviant behavior across generations is likely to increase by virtue of the fewer occasions available for the second generation to model the behavior.
Toward an Understanding of Intergenerational Parallelism
37
The labeling of first-generation behavior stigmatizes the secondgeneration child. Where deviant behavior by the second-generation child might be punished, if the deviant behavior is known to have been committed by an individual who grows up in the family context of a stigmatized parent, then the child is more likely to be stigmatized for the deviant behavior, to become alienated from the conventional world, and to adopt deviant behavior as a career contingency as a way of adjusting to a stigmatized identity. The comprehensive nature of the guiding general theory might have been attested to by a consideration of a number of other special theories or by examination of any of the wide variety of integrative theories that purport to bring together several of these special theories (Akers, 2000). However, these special and integrative theories are also regarded as compatible with or capable of being subsumed under the present general theory (Kaplan, 2003).
Applying an Integrative Theory Each of the major theoretical orientations addressing the etiology of individual deviance also might be implicated in the intergenerational continuity of deviance. The involvement of these theoretical perspectives variously asserts intervening variables, intergenerational continuity of intragenerational antecedents, and moderating influences. Virtually every construct employed in the intergenerational literature relates on a face valid basis to traditional theories of deviant behavior. For example, intergenerational modeling appears to relate to social learning/subcultural theories of deviance transmission. Experiences of subjective distress or antecedents of subjective distress such as marital conflict or abusive parenting practices reflect stress theories or, more particularly, self theories. Parental influences whether as contextual or mediating variables are interpretable in terms of degree of social control that is exercised over children who are otherwise predisposed to engage in deviant behavior or in terms of the distress caused the second-generation subject as a result of being abused, ignored, or otherwise derogated by the parent in the course of the socialization experiences. The mediating, moderating, and common antecedent variables that are observed to play a role in intergenerational parallelism are interpretable at one level of abstraction in terms of a relatively small number
38
Chapter 1
of theoretical variables that together compose the above-summarized integrative theory of deviance. As the integrative theory is applied to the empirical investigation of intergenerational parallelism of deviance, theoretical constructs are operationalized and are specified as being related to each other either as reflections of, or contributions to, the explanation of intergenerational parallelism. Intergenerational parallelism of deviance is a special case of bivariate relationships in which a pattern of deviance observed for firstgeneration subjects is associated with a like deviant pattern observed for second-generation subjects. The explanation of the bivariate relationship, informed by the integrative theory, takes the form of specifying (1) direct or indirect causal effects of first-generation (G1) deviance on secondgeneration (G2) deviance, (2) moderators of the linear relationships, and (3) intergenerational continuity of intragenerational causes of deviance. The theoretically informed models of intergenerational parallelism of deviance are illustrated in Figure 1.1. The general theory specifies theoretical conditions (y) under which the empirical association between G1 and G2 deviance (a) is strengthened or weakened. For example, under
Figure 1.1. Explaining Intergenerational Parallelism (or its absence) in Deviant Behavior
Toward an Understanding of Intergenerational Parallelism
39
conditions where the amount of interaction or emotional ties between the two generations are lessened, the degree of intergenerational parallelism will be weakened, as it would under conditions of rapid social change disrupting commonality of attitudes and values. Conversely, where circumstances facilitate social learning and where positive affective ties are maintained across generations, G1–G2 parallelism in deviance would be increased. Where the empirical association between G1 and G2 deviance is hypothesized to be causal in nature, the effect of G1 deviance on G2 deviance is decomposed in terms of theoretically informed intervening processes. For example, G1 deviant behavior (through specified intervening processes) leads to G1 adult deviance (e) that is imitated by G2 adolescents (f), or G1 deviant behavior influences G1 adult deviance (e). which influences stigmatization of the G2 adolescent (g) that, in turn, leads to G2 deviant adaptations to the concomitant distressful selffeelings (c). Each of these causal linkages is contingent on specified circumstances (moderators). For example, the linkage between G1 adolescent deviance and G1 adult deviance (e) is contingent on the effects of the G1 adolescent deviance on the G1 adolescent’s self-feelings (s); the stigmatizing effect of G1 adult deviance on the G2 adolescent (g) is contingent, in part, on the G1 adult coming to the attention of the authorities (r); and the direct effect of G1 adult deviance on G2 adolescent deviance (f) depends, in part, on the visibility of the G1 adult deviance (t). Instances of intergenerational parallelism of deviance might be explained on theoretical grounds by processes other than, and in addition to, the more or less direct effects of G1 deviance on G2 deviance. These processes involve the parallel operation of intragenerational causal processes on deviance (b, c) and causal relations between the parallel intragenerational causes (d). For example, in each generation, the experience of self-devaluing feelings might influence the adoption of a deviant pattern as an attempt to enhance self-feelings (b, c). The self-devaluing feelings of G1 subjects might have an intergenerational causal impact on G2 self-devaluing experiences, whether more or less directly (d). Less directly, for example, G1 self-devaluing feelings as adolescents might lead to G1 adult self-rejection (h) that, in turn, is modeled by the G2 adolescent (g). Thus, whether G1 deviance has a causal effect on G2 deviance, instances of observed associations between G1 and G2 deviance, in part, might be accounted for by the intergenerational causal
40
Chapter 1
relationship between variables that have parallel intrageneration effects on deviance. As in the case of causal relationships between G1 and G2, all of these linear relationships have their own moderators. For example, the intragenerational effects of self-rejection on deviance (b, c) are moderated by the presence of social supports or self-enhancing alternatives that might blunt the disposition to engage in deviance or of social control mechanisms that might mitigate the acting out of the self-rejection-engendered deviant impulses (z, x); also, the influence of G1 adult self-rejection on G2 self-rejection (g) is contingent on such factors as the physical presence and positive affective bonds between the parent and child (r). The intragenerational causes of deviance have their own causes (i, j). For example, in each generation, abusive parenting might cause selfrejection (i, j) that, in turn, causes deviant behavior adaptations (b, c). As in the case of the primary intragenerational causes, the indirect causes might have an intergenerational (more or less direct) causal relationship, and all of the linear effects have their own moderators (all of which are represented in Figure 1.1 by unlettered arrows). Following a presentation of method in chapter 2, the results of the estimation of representative theoretically informed models are reported in Parts II and III. In Part II, models are estimated that focus on moderators of intergenerational parallelism of deviance and selected antecedents of deviance. In chapter 3, models are estimated that reflect instances of intergenerational parallelism of deviance and its correlates. In chapter 4, the estimated models reflect theoretical conditions that facilitate or impede intergenerational parallelism of deviance and theoretically salient antecedents of deviant behavior. In Part III, hypothetical causal relationships are decomposed in terms of theoretical mediating variables (chapter 5) or intragenerational causes of deviance and its correlates that are continuous across generations (chapter 6). Together, the results of the analyses as these are summarized in Part IV (chapter 7) offer (1) a methodological template, a logic of procedure for studying the bivariate relationship that is intergenerational parallelism of deviance in particular (and, indeed, for studying and elaborating any bivariate relationship), and (2) theoretically informed substantive contributions to the understanding of intergenerational parallelism of deviance.
2 Method
The research on the intergenerational parallelism of deviance to be reported in Parts II and III should be evaluated against a consideration of the adequacy of the procedures that have been followed in extant intergenerational research. Presented, in turn, is a critical review of the methodological characteristics of others’ research on intergenerational parallelism and a description of the research procedures employed in the present study to address what are perceived as methodological limitations in current research.
Methodological Limitations of Current Research The research to be reported in subsequent chapters tries to assess and explain the degree to which parent’s deviant behavior in late childhood or early adolescence predicts similar behavior in their own children at comparable ages. A large literature exists that attempts to accomplish these same objectives. A number of observers have presented what they see as desirable characteristics of such research, implying that a good deal of research on (what is here called) intergenerational parallelism fails to approximate these ideal conditions. Thus, It has been argued that the design of intergenerational research should properly meet three criteria (Cairns, Cairns, Xie, Leung & Hearne, 1998): First, such investigations presuppose the study of people who are observed at approximately the same age (or developmental stage) in two or more successive generations. The information from each generation should be
41
42
Chapter 2 longitudinal to use time and sequence as primary variables and to clarify the potential mediational processes. Second, the longitudinal information should be prospective rather than retrospective. Given the limitations of retrospective reports—forgetting, distorting, and interpreting the past in light of the present—there are few alternatives to the prospective longitudinal study. Third, the information should be multilevel and not restricted to a single measurement source or domain. As Radke-Yarrow, Campbell, and Burton (1968) emphasized, independence of assessment provides a buffer against confounding in interpretation, whether the analyses are contemporary, retrospective, or prospective. (p. 1163)
Similarly, Van Ijzendoorn (1992) suggested with regard to studies of the intergenerational transmission of parenting that ...longitudinal studies should be carried out measuring parenting with comparable instruments at comparable times across the lifespan. Furthermore, contextual factors should be taken into account because the transmission may be stronger or weaker depending upon the influence of these contextual accounts on two or three generations. (p. 97)
More recently, Smith and Farrington (2004) concurred that the ...true nature of intergenerational continuities in behavior can be established more clearly using independent information about youth behavior in two successive generations assessed prospectively at roughly the same ages... (p. 231)
Taken together, these descriptions of desirable methodologies include the following: observation of the different generations at comparable developmental stages; use of prospective longitudinal designs; specification of mediating causal mechanisms, intergenerationally stable causes of the phenomenon of interest in each generation and (by implication) moderators of intergenerational parallelism; and independent sources of data for the separate generations as the more salient features of ideal research designs for the investigation of intergenerational parallelism. Comparable Developmental Stages Several researchers have noted the virtues of controlling on developmental stage, and the biases that might be introduced by failing to do so, in studies of intergenerational parallelism. For example, Cairns and associates (1998) observed:
Method
43
Early parenting studies have added significantly to the understanding of socialization processes and the various ways that parents and young children interact (Parke & Buriel, 1997). However, there are difficulties inherent in invoking parent–child comparisons to explain continuity across generations. Large differences exist between adults and children in the organization, form, and functions of their social and cognitive patterns. Ignoring these disparities may invite superficial evaluations of parent–child similarities and their functions. More important, because of the dynamic and ephemeral nature of interactional adaptations, it cannot be assumed that similarities observed in early transactions will be maintained across successive periods of developmental reorganization. Given these concerns, the rigorous assessment of intergenerational continuity seems to require the study of child–child comparisons across generations (see de Beer, 1958; Cairns, Elder, & Costello, 1996; Cairns, Gariepy, & Hood, 1990; Elder, Caspi, & Downey, 1986; Hood & Cairns, 1988; Magnusson & Cairns, 1996). The term intergenerational development refers to the relations between the ontogeny of behavior and cognition in children and the ontogeny of the same characteristics in their parents when they, in their own time, were children. To determine whether mother–child similarities persist over time, parent-to-child comparisons can be embedded in a long-term child-to-child comparisons across generations. When integrated over time, such linked longitudinal studies may provide a picture of the relations between the childhood of the parents and the childhood of the children and what the mediators of these effects are. Such studies also seem critical in the identification of possible mechanisms of cross-generational continuity and change (Burgess & Youngblade, 1988; Caspi & Elder, 1988; Furstenberg, Brooks-Gunn, & Morgan, 1987; Serbin, 1996; Werner & Smith, 1982). (p.1162)
The failure to control on stage in the life course across generations would be to ignore that individuals might have different life-courserelated needs that lead to biased reporting. Thus, according to generation stake theory (Acock & Bengtson 1980, p. 512): “Each generation has an investment in the generational bond. But, for youth, the ‘stake’ is more toward maximizing a sense of separate identity; for parents, the investment pays off in maximizing continuity.” The young adult children would be motivated to perceive relationships with parents as more conflictual than they actually are, whereas the parents might be more motivated to view the relationships in more positive terms that belie the realistic picture of the relationship. In support of this perspective, Aquilino (1999) cited studies of parents and children at different stages of the life cycle, including pairs of parents and adolescents (Noller, Seth-Smith, Bouma, & Schweitzer, 1992), parents and college students (Thompson, Clark, & Gunn, 1985), and
44
Chapter 2
middle-aged children and their parents (Bond & Harvey, 1991). As another example, Van Ijzendoorn (1992), citing a study (Hanson & Mullis, 1986) that evaluates parenting and child-rearing attitudes of female college students and their parents, observed that by comparing attitudes of two generations at different phases of the lifespan, it is not evident that the birth of a baby would or would not change the students’ attitudes to children and to child-rearing in one or another direction, which could have been ascertained had the female college students been tested at the same developmental stage as their parents. As these illustrations suggest, studies of intergenerational parallelism vary according to whether the developmental stages of the representatives of the separate generations are comparable. Data might be provided by or about representatives of the two generations at different or comparable developmental stages. In the former case, for example, adults might provide information about themselves as adults and about their children when they were preadolescent, adolescent subjects might provide information about themselves during adolescence and about their parents’ attitudes and behaviors, or adult subjects might provide information about their own attitudes and behaviors and their adolescent children might provide data about their attitudes and behaviors. In the latter case, where the data referred to comparable developmental stages of the representatives of the successive generations, subjects might report about the attitudes and behaviors of their young adult children and about their own attitudes and behaviors when they were young adults, or self-reports about adolescent subjects in one generation might be compared with the self-reports of adolescents in the next generation (i.e., their children) at the same developmental stage. For the above-noted reasons, controlling on developmental stage is the much preferred design feature in studies of intergenerational parallelism.
Prospective Longitudinal Studies An important limitation of studies of intergenerational parallelism in deviant behavior is the correlational nature of these investigations— that is, the failure to establish temporal priority among parental deviance, putative intervening circumstances, and deviance of the children (Downey & Coyne, 1990). Cairns and his associates (1998, p. 1162) noted that “only a handful of empirical researchers have assessed,
Method
45
prospectively, the behavioral and cognitive development of children relative to the development of their parents when they, in their own time, were children (but see Eron & Huesmann, 1987; Furstenberg et al., 1987; Serbin et al., 1991).” When studying intergenerational continuities, many research designs do not permit disentangling the influence of parents on children from the influence of children on parents. This is less of an issue in circumstances such as the present study in which measures on the parental generation were taken well before parenthood was reached. Thus, if causal processes are implicated at all, temporal priority of the baseline measures of the first-generation subjects have been clearly been established to be temporally prior to second-generation outcomes. This is not to say that more proximal processes might not involve reciprocal influences. As others have noted, particularly during late adolescence and early adulthood, relationships tend to be increasingly egalitarian (Vollebergh, Iedema, & Raaijmakers, 2001). In any case, particularly where the specification of intervening processes is an issue, prospective longitudinal designs facilitate less equivocal interpretations of observed instances of intergenerational parallelism.
Mediators, Common Antecedents, Moderators Research protocols that aim at a more complete understanding of instances of intergenerational parallelism must permit the specification and measurement of hypothetical mediating, moderating, and common antecedent variables. Any complete understanding of intergenerational parallelism must focus on intervening mechanisms of intergenerational transmission. As Van Ijzendoorn (1992) observed in connection with a review of studies of intergenerational transmission of parenting: Remarkably little is known about the mechanism of intergenerational transmission of parenting. Learning to be a parent and to acquire a certain parenting style may be the outcome of modeling, coaching, or other cognitive processes, and we are not able to derive from the studies reviewed here which (combination of) learning process(es) is most supported by the empirical evidence. Most studies are restricted to just showing that a relation between infant and adult characteristics exists, and do not give insight into the causal mechanism. (p. 95)
46
Chapter 2
Specification of intervening mechanisms provides insight into some of the circumstances that moderate instances of interpretational parallelism. Any variable that is observed to mediate the intergenerational transmission could be assumed to be a moderator of that relationship. Thus, if the intergenerational continuity of poverty is mediated by the influence of poverty on the educational level of the next generation, then the degree of continuity depends on the level of education of the second-generation subject. Under conditions whereby the second-generation subject gains a higher level of education, intergenerational continuity of poverty is less likely to be observed. Where poverty in one generation decreases a parent’s commitment to educational level in the second generation, first-generation parental commitment to the educational aspirations of the second generation will modulate the intergenerational transmission such that under conditions of high educational aspirations by the first-generation parent, intergenerational continuity of poverty is less likely to be observed. Although many instances of intergenerational parallelism might be explained by consequences of first-generation behavior or characteristics (and, by extension, moderators of the parallelism), these same or other instances might be explained, in part, by intergenerational continuities in factors that within each generation are significant causes of the phenomenon of interest (i.e., the construct that manifests intergenerational parallelism). Intergenerational parallelisms might be explained at many different levels. Thus, intergenerational continuities might result from continuities in social context or by intergenerational communication. As Vollebergh et al. (2001) observed with regard to intergenerational congruity in cultural orientation: Comparable orientations in both parents and their children may result from sharing the same environment or sharing comparable social status. Explanations for the intergenerational transmission of attitudes are therefore on at least two levels of analysis: on the level of the intergenerational transmission of social status and social positions from parents to their children (Glass, Bengtson, & Dunham, 1986) and on the level of direct transmission of cultural orientations of parents to those of their children through communication within the family (Acock & Bengtson, 1980; Beck & Jennings, 1975; Moen, Erickson, & Dempster-McClain, 1997; Petit, Clawson, Dodge, & Bates, 1996). Past empirical studies have suggested that both explanations are valid and complementary (Glass et al., 1986; Moen et al., 1997; Vollebergh, Iedema, & Raaijmakers, 1999). (pp. 1185–1186)
Clearly, then, the research design that attempts to account for, rather than merely describe the existence of, intergenerational parallelism requires
Method
47
the specification of intervening processes, variables that moderate the presence and magnitude of intergenerational parallelism, and intergenerational continuities that have causal impact on the outcome of interest in each generation.
Independent Sources of Data Studies of intergenerational parallelism vary according to whether a representative of one generation provides data about both generations or information on each generation is provided by a representative of that generation. In the former instance, for example, youth, adolescent, or adult subjects might provide data about themselves and about their parents or grandparents or adult subjects might provide information about themselves and about their children. Responses by the person regarding comparable attitudes or behaviors of the two generational representatives are taken to be an indication of intergenerational continuity or parallelism. In the latter instance, data are provided by a representative of each of the generations. Thus, parents might provide responses about their own behavior and their children might provide data about their own behavior. Where the separate responses indicate comparable behaviors or attitudes, intergenerational continuity or parallelism is said to have occurred. Studies of intergenerational transmission too frequently use reports by one or the other generation of the phenomenon that is examined across generations. Thus, for example, findings indicating that children of depressed parents demonstrate poorer psychological functioning than children of parents who are not depressed are often based on parents’ reports of the children’s functioning rather than on the children’s self-reports or independent observations of the children. Thus, it is possible that the effect of parental depression on children’s psychological functioning might be the spurious result of the tendency of the depressed parents to bias their perceptions in a negative direction (Avison, 1999; Rickard, Forehand, Wells, Griest, & McMahon, 1981). As another example, Velleman and Orford (1990) reported data for themselves as adults and for their parents with drinking problems. The importance of having independent reporters for each generation is apparent when contrasting methods in which a single reporter was used with those in which in separate representatives of the multiple generations reported. Thus, Chassin and her associates (1998), examining the
48
Chapter 2
continuity of parenting across generations, observed such continuity when the mothers offered data on their own parenting behavior assessed in adulthood. However, intergenerational continuity was not observed when the children’s perceptions of maternal support were used. Studies in which the same subject provides data about both generations, whether the data are provided about the different generations at different developmental stages, are vulnerable to criticism on numerous methodological grounds. Validity might be compromised by problems associated with method variance whereby associations between measures of the two generations are artificially inflated due to, for example, common underlying personality traits. As Simons, Whitbeck, Conger, and Wu (1991) observed: In the issue of intergenerational transmission of harsh parenting, it may be that aggressive individuals tend to see others, including their parents, as displaying high rates of aggressive behavior (Straus et al., 1980) thereby producing an artifactual relationship between descriptions of their own behavior and that of their parents. (p. 160)
In support of this assertion, although studies based on adolescents’ perceptions of congruence reveal intergenerational continuity regarding values dealing with educational goals, career, and major life concerns, the relatively few studies dealing with congruence based on reports from parents and their adolescent children observe relatively little congruence (Gecas & Seff, 1990). Another problem with collecting data on the two generations from a common source is that the relationship might be artifactually reduced if we are dealing with socially undesirable data. The cross-generational data are further threatened when a representative of only one of the generations provides data about both generations due to retrospective distortion. As Oliver (1993) observed with regard to the intergenerational transmission of child abuse: Omission, confusion, irrationality, distortion, and most bizarre of all, idealization in accounts of cruel, rejecting, or neglectful grandparents (G1) by parents (G2) with rearing difficulties and/or ill-treated children (G3) are universal findings. (p. 1320)
The work of Aquilino (1999), alluded to earlier in connection with the need to control on stage in the life course, is relevant also in regard to requiring that each representative of successive generations provide data on
Method
49
their own situation rather than having one or the other representative provide data for both. It apparently does matter who provides self-report data on such matters as relationship quality contact and interaction patterns. Systematic differences exist in the perspectives of parents and their adult children regarding the nature and quality of their relationships. In general, the parents are more likely to provide a happier view of the intergenerational relationships than a sample of adult children. This might be due, in part, to the two generations having different stakes in the relationship—that is, different psychological needs of the two generations at different points in the life cycle (Acock & Bengtson 1980). Thus, young adult children might exaggerate conflictual relations with parents in order to facilitate their need for separation from the family of procreation. The older parents, on the other hand, might experience a need for continuity of preexisting relationships and, therefore, might need to perceive a more positive relationship with the adult children. As Aquilino (1999) concluded: ...theoretical and empirical scholarship on intergenerational relations would benefit from more attention to the issue of divergent perspectives in families. I believe there is a strong case for making the collection of equivalent data from both generations a standard practice in designing research on parents and adult children. (p. 869)
A number of studies can be mentioned that conform to some of the features of an ideal study design but not to the others. Thus, although very few studies exist that permit assessments of antisocial behavior across “two generations of children at comparable ages, employing independent reporters and comparable measurements” (Smith & Farrington, 2004, p. 231), Smith and Farrington (2004) cite a number of studies that have these desirable characteristics: The Concordia Longitudinal Risk Project followed female subjects from childhood over a span of 20 years, and assessed a subsample of 89 Canadian females and their children at ages 5 to 13 (Serbin et al., 1998) Here a weak correlation between aggression in successive generations was found, although this was largely mediated by the mother’s educational attainment. Farrington (1993) utilized Cambridge Study data to assess bullying and antisocial behavior in a sample of 411 London males from the age of 8, and in their resident children aged 3–15 when the men were ages 32. This study found continuity in bullying, with 30% of men who had been identified as bullies, at age 14 reporting that their children were bullies, compared with 17% of men who had not been bullies. Huesmann and his colleagues (1984b) assessed aggression in over 600 subjects followed from age 8 to age 30, and
50
Chapter 2 found that in the subjects at age 8 predicted aggression in the subject’s children 22 years later when they were aged 8. (p. 231)
Other studies of intergenerational parallelism (whether or not focused on antisocial behavior) might be cited as well, in which data are provided by these representatives of the several generations at comparable developmental stages (Blee & Tickamyer, 1986; Harburg, Gleiberman, DiFranceisco, Schork, & Weissfeld, 1990; Lefkowitz, Huesmann, & Eron, 1978). Although these studies display a number of desirable design features, they are limited in other ways. For example, Smith and Farrington’s (2004) study of continuities in antisocial behavior and parenting across three generations uses data on only 411 inner-city (London) males. Data from the earlier point in time come from first-generation mothers, whereas data at the later point in time come predominately from second-generation fathers; that is, data on the index child’s (second generation) conduct problems were provided by the child’s mothers, whereas data on the children of those subjects (grown up) were provided by the fathers. In Capaldi and Clark’s (1998) investigation, antisocial behavior was measured by different indicators for the parents as opposed to their 10-year-old boys. Other studies failed to address the issue of explaining observed instances of intergenerational parallelism in terms of intervening processes, moderating variables, and intergenerationally stable effects that have causal implications for the variable of interest in each generation. The present research design meets these criteria. The subjects in each generation are observed at approximately the same developmental stage (early adolescence) in two successive generations. The information for the first generation is longitudinal in nature, the subjects having been assessed several times between early adolescence and the fourth decade of life. The second-generation subjects have been or will be interviewed at different points in the life course. The longitudinal information is prospective rather than retrospective. Information on the respective generations is provided by the separate representatives of the generations rather than being provided for both generations by the representative of one of the generations. The sample size is adequate to estimate reasonably complex models. Analyses take into account moderators, intervening processes, and intergenerationally continuous factors that have common causal effects within each generation on the object of the investigation of intergenerational parallelism. We now turn to a description of the multigeneration prospective longitudinal study that provides the data for the investigation of intergenerational parallelism of deviance.
Method
51
A Multigeneration Prospective Study Data Collection The data used in this study were collected from a first-generation (G1) panel starting in 1971 (G1T1) from a target population made up of the seventh-graders (N = 9,056) enrolled in a random half of all the junior high schools (N = 18) in the Houston Independent School District. These adolescents were surveyed again in 1972 (G1T2) and in 1973 (G1T3). The survey instrument consisted of 209 items and was identical in all three waves with the exception that waves two and three provide (for the most part) self-reported deviant behaviors for the previous year, rather than the previous month, as in the first wave. Since then, the subjects have been surveyed through personal interviews (including the T1–T3 variables) using greatly expanded instruments in the 1980s (G1T4) and again in latter part of the 1990s (G1T5), which was the final G1 data wave with a total of 5,467 subjects. At G1T5, the adult respondents were asked about the number, ages, sex, and addresses of their biological, step, adopted, and foster children. The parent’s permission was then sought to interview their children (G2), who would be contacted to participate in a second-generation study (G2T1). If the child subject was younger than the age that their parent was when he or she was first interviewed, then that child was not interviewed until they were at least 12 years old. At the end of the G2T1 data collection wave, a total of 7,519 second-generation subjects had been interviewed. It was from these parents and their children that the sample for this study was drawn. The selection criteria needed for the G1 subjects to be included in the sample were that they provided data in both the Time 1 (T1) and Time 2 (T2) data collection waves and had a least one child interviewed at G2T1. The selection criteria for the G2 subjects were that they provided data at Time 1, they were biologically related to the parent subject, and they were one to 2 years younger or older in age than the parent subject at the time he or she was interviewed. These criteria produced an initial sample size of N = 2,721. The sample was then aggregated, so that G1 subjects who had multiple children in the sample had their children’s responses averaged within their respective families. This means that the variable values were first added together for all children within any given family and then were
52
Chapter 2
divided by that same number of children in each family independently. This produced a one-to-one correlation of one G1 subject to an averaged G2 subject for each intergenerational relationship, for a total sample size of N = 1,658.
Missing Data Ignorable missing data are usually a product of two types of mechanism: missing completely at random (MCAR) and missing at random (MAR). Data are MCAR when a subject’s nonresponse to a question is not dependent on any other measured or observed variable related to the subject, study, or the question itself. If a subject’s nonresponse to a given question is contingent on subject characteristics or a previous response, but not dependent on the question itself, then the data are considered MAR (Enders & Bandalos, 2001; Rubin, 1976). It should be evident that MCAR is the stronger assumption, because data that are MCAR are also MAR. Missing data in the variables reported here are assumed to be the less restrictive MAR type. However, given the nature of the variables, it is possible that the subjects’ responses might not even meet MAR assumptions. There are several methods for addressing missing data. Such methods include theory-based direct maximum likelihood (ML) or full information maximum likelihood (FIML), listwise and pairwise deletion, and different forms of multiple imputation. In general, the majority of recent research into the efficiency of missing-data methods has shown that direct ML techniques outperform all other methods (Enders, 2001; Little & Rubin, 1987). One drawback of the direct ML method is that it assumes multivariate normality similar to all ML estimation methods. However, little is known about how these methods work in the presence of nonnormal data and/or clustered data such as used for the study. If it follows other ML estimation techniques, then parameters will be increasingly biased as the degree of nonnormality and clustering increase. An alternative form of imputation, and the one used for this study, called the similar response pattern method has been implemented in PRELIS 2, which is a preprocessor for the LISREL program (Joreskog & Sorbom 1994a). The method attempts to impute real values from another case with similar observed values by using a minimization routine based on a set of matching variables. If the routine cannot find a case with
Method
53
complete data using the matching variables, then the missing value for that variable is not imputed into the case and remains missing. A study by Brown (1994) found that compared to listwise and pairwise deletion, mean imputation, and hot-deck imputation, similar response pattern imputation produced the least bias overall with regard to structural and measurement model parameters. However, he did find that there was some positive bias in the error estimates, indicating that Type 1 error rates would be larger than normal. Although there is no statistical theory that would support this method over direct missing-data methods, the fact that it imputes values from similar cases is attractive because of the clustered nature of the second-generation data. If it is plausible that children from the same family would have more similar responses to each other than to children from other families, then possibly imputing a value from a respondent’s sibling does have some validity. As suggested in the PRELIS manual, a large number of matching variables were used, including subject identification numbers, that were not otherwise used to select the G1 or G2 subjects or used in any of the model estimations as moderators, indicators, or other variables.
Variable Construction All variables, except for the binary moderators, are cumulative indexes made up of a number of binary items. The items were coded such that as the value of the variable increased the negative effect of the variable increased. Specifically, the first-generation deviant behavior variables were taken from a combination of the Time 1 and Time 2 surveys. As mentioned earlier, the items that were used to construct these indicators were based on two different time references. Time 1 questions asked whether each subject preformed the deviant behavior within the last month, except for the alcohol use question, which asked if it had happened within the last 2 weeks. The same questions were asked at Time 2, but the time reference was within the last year, except for the alcohol use question, which again asked if it had happened within the last 2 weeks. The Time 2 data were the primary responses used for G1 variable construction. If G1 subjects had a missing value at Time 2, then the Time 1 response was substituted in its place. If, for any given item, the value was missing at both Time 2 and Time 1, then that response was coded as a
54
Chapter 2
missing value. The effect of this scheme minimized missing data at Time 2 and best replicated deviant behavior ever engaged in by the G1 subject despite the fact that the questions only covered behavior within the last 2 years of the subject’s life. Given that the G1 variables were indeed simple additions of similar items with less than seven categories, all of them were treated as true ordinal-scale variables The survey questions for the G2 subjects specifically asked whether they had ever performed the deviant behavior in their lives up to the point of the survey. Thus, only the G2T1 survey responses were used to construct their indicator variables. Because in many instances there was more than one G2 subject from a family in the sample, after the variable indicators were constructed they were then added together with their siblings and then divided by the total number of their siblings within each family. This averaging produced more values for each G2 variable than the SEM software would classify as ordinal, so the G2 indicators were treated as continuous such that when used with the ordinal indicators of the G1 subjects, polyserial correlations were adopted in model estimation. For both G1 and G2, the items used to construct the disposition to deviance and negative self-feeling variables were not time referenced. In general, missing values for these items were very low, so only Time 2 responses were used to construct the G1 variables. As with the deviant behavior items, only the G2T1 survey responses were used to construct the G2 variables.
Variable Nonnormality As previously mentioned, the scale of the G1 variables were treated as ordinal and the G2 variables were treated as continuous. There is some ambiguity and debate about how to classify variables measured on an ordinal scale when there are only three to five categories, as is the case with many of the G1 variables. Given that these variables were cumulative indexes composed of face valid, similar items, with relatively high reliability coefficients, it is plausible that the values between categories were equidistant. It is also plausible that the relationship between the categorical measured variable and the underlying theoretical variable that it was supposed to measure had a linear relationship. However, given the nature of deviant behavior in general, both the G1 and G2 variables measuring the behavior were, in a univariate sense, both heavily skewed and kurtotic.
Method
55
Another way to investigate the degree of nonnormality is to estimate the bivariate polyserial correlations between all of the indicators in any model to be estimated. PRELIS 2 has a unique routine that allows the evaluation of the degree of nonnormaily of any given polyserial correlation by a root mean square error of approximation (RMSEA) score (Joreskog & Sorbom, 1994a). All of the variables used in the models and their corresponding polyserial correlations with the other variables in any specific model fell within the acceptable range.
Statistical Methods and Latent Variable Model Estimation SPSS 11.1 and LISREL 8.54 (Joreskog & Sorbom, 1994b) were used for variable recoding, calculation of univariate statistics, data imputation, bivariate normality evaluation, and any needed secondary analyses. The software used for the estimation of the latent variable models was Mplus 3.12 (Muthen & Muthen, 2004). The output from Mplus includes univariate and multivariate statistics, model fit, and residual information that was used to assess the hypotheses and theories that motivated this study. Structural equation modeling specifies hypothesized relationships between observed variables and their latent constructs and the structural relationships among the latent constructs in the model. The measurement model describes the hypothesized relationship between the observed variables and the unobserved constructs that are presumed to underlie the indicators. The relationship between indicator and construct is expressed in terms of factor loadings and error terms for both variable and construct. The structural model reflects the hypothesized causal relationships between latent constructs as regression coefficients. For ordinal models, the interpretation of the estimates/coefficients for paths to a categorical outcome in Mplus, such as paths from predictors to an observed categorical dependent variable, are probit regression coefficients. The test statistic for the indicators is the estimated parameter divided by its standard error. The asymptotic distribution of Est./S.E (Estimate divided by Standard Error) can be treated as standard normal with a z-distribution (Muthén, & Satorra, 1995). The models computed with Mplus used the WLSMV estimator, which produces weighted least square parameter estimates utilizing a diagonal weight matrix with robust standard errors and a mean and
56
Chapter 2
variance modified chi-square test statistic. The models were evaluated in terms of the strength of the indicator regressions, magnitude of the structural parameter(s), and the goodness of fit. The three fit statistics used were the chi-square value, Steiger’s RMSEA as described by Steiger and Lind (1980), and the Comparative Fit Index (CFI) as described by Bentler (1990). The chi-square value was the primary evaluator, but given its susceptibility to bias, if it was found to be too high given sufficient sample size and simple model structure, then the other two fit statistics were used.
Multigroup Analysis Binary moderators were used to divide the sample into two subgroups. These subgroups were then compared for invariance across the models of interest. In general, two types of comparison can be employed to test for invariance, model form, and parameter congruence (Bollen, 1989). Given the relative simplicity of the models, it was assumed that the form of the models would be identical. Thus, only the model parameters were tested for invariance. The strategy of the multigroup analysis was to first estimate the baseline models for each subgroup separately to see if the separate models fit the data. This is an important first step, because if the models do not fit the data for one of the subgroups, then it would be unlikely that the two subgroups would be equivalent on a significant number of the model parameters, which would negate the need for further tests of invariance. If it was found that the models fit both of the subgroups via inspection of their respective fit indexes, then the measurement parameters were tested for invariance by placing across group equality constraints on the factor loadings for each group and estimated as a stacked model. The p-value of the chi-square statistic was inspected and if it was found to be significant, then the invariance of the measurement model was supported. When the measurement model was found to be invariant, the third step of constraining the structural parameter to equivalence across the subgroups was performed. The difference in chi-square statstic between the constrained measurement parameters and the fully constrained model and its respective p-value was calculated to test whether the structural parameter was invariant across the two subgroups.
Part II THE CONDITIONAL NATURE OF INTERGENERATIONAL PARALLELISM
Part II considers both the observed nature and magnitude of intergenerational parallelisms and the conditions under which the intergenerational continuities will be observed in greater or lesser degree. In chapter 3, the literature on intergenerational parallelisms is reviewed with regard to putative correlates of deviant behaviors and patterns of deviant behavior. Patterns of deviant behavior encompass psychiatric disorders of mood and cognitions, socially disvalued behaviors, and deviant role performance in the context of traditional social institutions. The literatures reflect a plethora of instances of intergenerational parallelism, with some exceptions, although the magnitudes tend to be quite modest. Consistent with the literature, intergenerational parallelisms in deviant behavior, disposition to deviance, and negative self-feelings were observed at statistically significant but less than appreciable levels. Variability in the significance and magnitudes of intergenerational correlations suggest that the relationships are conditional ones. Chapter 4 considers the literature on moderators of interpersonal parallelism and reports the estimation of models that specify the conditions under which intergenerational continuities will be observed to be stronger or weaker. The focus is on moderating circumstances that facilitate or impede intergenerational parallelism in deviance and in two important constructs known to be implicated in the onset and continuity of deviant behavior, namely disposition to deviance and negative self-feelings.
57
3 Observing Intergenerational Parallelism in Deviance
The behavioral and social science literatures are rife with instances of observed intergenerational parallelism. In this chapter an overview of the literatures is offered as a context for considering observations of intergenerational parallelism of deviance observed in our own multigeneration study.
Literature on Intergenerational Parallelism Ample evidence exists that at least modest degrees of intergenerational parallelism might be discerned for traits and behavioral dispositions regardless of their dysfunctional or nondysfunctional nature (Belsky & Penksy, 1988). The latter category can be thought of, in a very general way, as correlates of deviant patterns rather than as reflections of deviant patterns per se. Correlates of Deviant Behavior Transgenerational continuities have been observed with regard to a wide variety of phenomena that are interpretable as intragenerational antecedents/causes of deviant behavior, including similarity in occupation (Biblarz, Raftery, & Bucer, 1997), parental bonding style (Miller, Kramer, 59
60
Chapter 3
Warner, Wickramaratne, & Weissman. 1997), attitudes toward consumption of specific foods (Stafleu et al., 1995), parenting of siblings (Kramer & Baron, 1995), and disciplinary practices (Fry, 1993; Simons et al., 1991). Arguably, the most frequently reported parallelisms relate to parenting patterns. A number of studies reflect continuity across generations of parenting patterns. That is, individuals who as children experience certain kinds of parenting responses are likely to replicate those responses when they themselves become parents (Chen & Kaplan, 2001; Simons, Beaman, Conger, & Chao, 1993; Simons et al., 1991). Evidence of intergenerational transmission of parenting patterns is summarized by Van Ijzendoorn (1992) and Putallaz and associates (1998), also cited by Smith and Farrington (2004). These reviews conclude “that higher-order parenting dimensions such as supportive/attached and punitive styles of parenting show the most intergenerational stability and there are a few longitudinal studies that illustrate this” (Smith & Farrington 2004, p. 232). Numerous studies have reported on the intergenerational transmission of parenting practices, whether of the abusive (Egeland, Jacobvitz, & Papatola 1987; Putallaz, Constanzo, Grimes, & Sherman 1998) or constructive (Chen & Kaplan 2001) varieties. Several studies have reported intergenerational parallelism in positive parenting. In general, parents who recalled their own parents as being warm, more accepting, sensitive, supportive, less intrusive, and nonconflictful tended to be more responsive and to adapt better to their own infants (Heinicke, Diskin, Ramsey-Klee, & Given, 1983; Main, Kaplan, & Cassidy 1985; Ricks 1985). Whereas the preceding studies focused on the second-generation parents’ responses to their infants, similar results were found for adolescent children. Simons and his associates (1993) also reported intergenerational continuity of supportive parents. The respondents’ recalled their own early upbringing, whereas their own parenting behavior was reported by their adolescent children as well as being manifest through household observations. Evidence also exists for the intergenerational transmission of specific parenting attitudes. Thus, Chassin and her associates (1998, p. 1200) reported: “Mothers who, in adolescence, perceived their parents as opposing their smoking later grew up to provide more smoking-specific discussion to their own children.” In addition to continuities in parenting patterns, intergenerational parallelisms have been reported for a wide variety of other dimensions of
Observing Intergenerational Parallelism in Deviance
61
human response. Thus, parallelism in sex role attitudes between mother and daughter has been observed (Smith & Self, 1980) and intergenerational transmission of sexual behavior has been noted (Newcomer & Udry, 1984), as has intergenerational parallelism in interpersonal competence (Filsinger & Lamke, 1983). Vollebergh et al. (2001) reported parallelisms between parental worldviews and the worldviews of the adult children of those parents, as others have observed (Acock & Bengtson, 1980; Dalhouse & Frideres, 1996; Jennings & Niemi, 1981; Miller & Glass, 1989; Raajimakers, 1999), and Bengston (1975) reported intergenerational continuity in value orientations. Instances of intergenerational parallelism have been reported for virtually every social institution: Harvey, Curry, and Bray (1991) reported intergenerational transmission of relational patterns of individuation and intimacy and Amato and Booth (1997) reported evidence for intergenerational transmission of marital quality; Wickerama and associated (1999) reported evidence of intergenerational (parent to adolescent child) transmission of health-risk behavior; Bao and associates (1999) provided evidence of intergenerational transmission of religious behavior and attitudes; and Cairns and his associates (1998, p. 1162) noted that “the academic competence of mothers when they were children was significantly linked to the academic competence of their children at school age.” Finally, intergenerational parallelism in income has been reported frequently (Israel & Seeborg, 1998): Economists have generally found a relatively small but significant correlation between parents’ income and their children’s earnings (Behrman & Taubman, 1985; Corcoran, et al. 1990, 1992; Krein & Beller 1988; Peters, 1992; Solon, 1992; Solon, Corcoran, Gordon, & Laren, 1991). For example, using a sample of parent/child pairs from the National Longitudinal Surveys, Peters (1992) estimated that parents’ log income explains between 9% and 11% of the variation in children’s log incomes, and Solon (1992) found substantial father/son correlations in hourly wages and family income. (p. 755)
Deviant Patterns More germane to the present study are observations of intergenerational parallelism of patterns of deviance. These include psychiatric disorders reflecting aberrant patterns of mood or thinking, a variety of socially
62
Chapter 3
disvalued behaviors, and deviant responses in institutionalized social roles. Each category is considered in turn. Psychiatric Disorders
Intergenerational parallelism of psychiatric disorder is well established (Velleman, 1992). Children of psychiatric patients are significantly more likely to manifest persisting emotional behavioral difficulties relative to children whose parents were not psychiatric patients (school classroom controls; Rutter, Quinton, & Yule, 1976). Individuals with schizophrenic parents are significantly more likely to develop schizophrenic symptoms than children of nonschizophrenic parents (Cohler, Gallant, & Grunebaum, 1977; Emery, Weintraub, & Neale, 1982; Fisher & Jones, 1980; Sameroff, Seifer, & Zax, 1982; Willerman & Cohen, 1990). Kaplan and Liu (1999) observed a substantial intergenerational continuity in antisocial character between mother-daughter dyads tested at the same developmental stage (early adolescence). These investigations vary with regard to whether the observations of the multiple generations are made at comparable developmental stages. In some instances, the observations of psychiatric disorders are made at different stages of the life course for the two generations. Thus, studies of the intergeneration transmission of depression might observe the effect of adult parental depression on the infant’s or child’s psychological difficulties (Field, Healy, Goldstein, & Guthertz, 1990; Lee & Gotlib, 1991). In other studies, stage in the life course is held constant. Thus, in the case of intergenerational transmission of alcoholism, it is frequently observed that children of alcoholic parents as adults are at increased risk for alcoholism (Sher, 1991). Socially Devalued Behaviors
Intergenerational parallelism in deviant behavior has been observed for a wide variety of deviant patterns, including substance abuse in general (Sheridan, 1995), child maltreatment (Zuravin, McMillen, DePanfilis, & Risley-Curtiss, 1996), eating disorders (Steiger, Stotland, Trottier, & Ghadirian, 1996), running away (Plass & Hotaling 1995), alcoholism (Johnson & Bennett 1995), aggression (Doumas, Margolin, & John 1994), antisocial behavior (Tapscott, Frick, Wooton, & Kruh, 1996), and harsh parenting techniques (Simons et al. 1991). Regarding other substance
Observing Intergenerational Parallelism in Deviance
63
abuse, intergenerational continuity in such behavior has been observed between parents and adolescent children (Fawzy, Coombs, & Gerber, 1983). Intergenerational parallelism has been noted, for example, between the drinking practices of first-generation adults and second-generation adolescents (Peterson, Hawkins, Abbott, & Catalano, 1994; Velleman, 1992). Cross-generational parallelism in smoking behavior also has been noted (Chassin et al., 1998). Numerous studies have reported cross-generational parallelism in indices of antisocial behavior (Cairns et al., 1998; Serbin et al., 1998). For example, an intergenerational association between paternal antisocial behavior and children’s conduct disorder has been reported (Tapscott et al., 1996). Evidence is available also for the intergenerational transmission of running away (Plass & Hotaling, 1995). The children of parents who ran away from home are more likely to run away themselves than are children whose parents did not run away. Regarding deviant patterns in the school environment, Kaplan, Kaplan, and Liu (2000) reported a significant intergenerational effect of parents’ junior high school school-related deviant behaviors and their children’s school-related deviant behaviors observed at a comparable developmental stage. School-related deviance was modeled as a latent construct reflected in measures of school failure, being the object of formal sanctions (suspension/expulsion, taken to the office for punishment), rejection by teachers, truancy, and disobedience to teachers. Intergenerational continuity in school failure has been noted by others as well (Cairns et al., 1998). In addition to instances of intergenerational isomorphism in deviance where the index of deviance was similar or identical for both first- and second-generation subjects, forms of deviant behavior manifested in one generation are observed to be associated with other forms of deviance in the next generation. Thus, substance abuse in one generation is linked to other forms of antisocial behavior in the children of the substance abusers (Hayes & Emshoff, 1993). Deviance in Social Role Performance
Many instances of intergenerational parallelism in deviance refer to failure to conform to role expectations, particularly in the context of the family. Four patterns of intergenerational parallelism in familial deviance are particularly noteworthy in the literature. These relate, respectively, to marital conflict, divorce, early pregnancy/childbearing, and abusive parenting.
64
Chapter 3
Intergenerational transmission of marital conflict has been noted frequently (Cowan, Cowan, Schulz, & Heming, 1993). Growing up in families characterized by marital conflict increases the likelihood of marital conflict in the next generation (Tallman, Gray, Kullberg, & Henderson, 1999). In examining the intergenerational transmission of marital aggression, Kalmuss (1984) distinguished between two types of modeling: Generalized modeling occurs when childhood family aggression communicates the acceptability of aggression between family members and thus increases the likelihood of any form of family aggression in the next generation. Generalized modeling does not necessarily involve a direct relationship between the type of aggression in first- and second-generation families. Specific modeling occurs when individuals reproduce the particular type of family aggression to which they were exposed. Intergenerational modeling of marital aggression appears to involve specific more than generalized modeling. Severe marital aggression was more strongly related to observing parental hitting than it was to being hit by one’s parents. (p. 15)
That is, intergenerational aggression appears to be role-specific. Exposure to aggression between specific family members teaches children the appropriateness of such behavior between the inhabitants of those family roles. Thus, parental hitting teaches the acceptability of marital more than of parent-child aggression. Similarly, parent-child hitting should more effectively communicate the acceptability of parent-child than of marital aggression. (Kalmuss, 1984, p. 17)
Numerous studies have reported intergenerational parallelism in the likelihood of being divorced (Amato, 1996; Bumpass, Martin, & Sweet, 1991; Greenstein, 1995; McLeod, 1991; Wolfinger, 1999). Adult children of divorced parents were significantly more likely to have their own marriages end in divorce than the children of parents who did not divorce (Amato 1996; Bumpass et al., 1991; Glenn & Kramer, 1987; Keith & Finlay, 1988; McLanahan & Bumpass, 1988). The risk of intergenerational continuity in divorce is even greater if both of the first-generation spouses have divorced (Amato 1996). Amato and DeBoer (2001) reported that the likelihood of offspring divorcing is approximately doubled if parents had divorced. This relationship could not be accounted for by the distress that precipitated the parental divorce because children who had parents whose marriages were distressed but who remained married did not have an increased likelihood of divorce. The observation that divorce is more likely to be transmitted
Observing Intergenerational Parallelism in Deviance
65
intergenerationally if the parents reported a low level of discord (Amato & DeBoer, 2001) is compatible with reasoning that ...if parents exhibit relatively little overt discord prior to divorce, then children may conclude that the marital promise can be broken even if the marriage is not seriously troubled. Divorces that occur among low-discord are more likely to weaken offspring’s commitment to lifelong marriage than divorces that occur among high-discord couples. (p. 1041)
Transgenerational parallelism in early childbearing has been noted in several studies (Capaldi & Clark, 1998; Hardy, Astone, Brooks-Gunn, Shapiro, & Miller, 1998; Serbin et al., 1998), thus supporting the findings of earlier investigators (Card, 1981; Furstenberg, Levine, & Brooks-Gunn, 1990; Haveman, Wolfe, & Peterson, 1997; Kiernan & Diamond, 1983). Manlove (1997) reported intergenerational parallelism in early motherhood independent of the influence of several family, school, and individual variables. Thus, daughters of teen mothers have a significantly higher likelihood of teen childbearing themselves compared to daughters of older mothers (Kahn & Anderson, 1992). More recently, Barber (2001b) reported on intergenerational transmission of age at first birth among sons as well as among daughters and particularly when the intergenerational transmission relates to premaritally conceived first births. Finally, abusive parenting is frequently the subject of investigations of intergenerational parallelism: Pears and Capaldi (2001) reported data congruent with the existence of a direct effect for intergenerational transmission of abusive behavior. Parents who had a history of being abused tended to be abusive toward their own children; Simons and his associates (1995) reported a correlation between parental reports of having been the object of harsh discipline during childhood and reports of their own children being the object of corporal punishment; Rand (1992) reported on the intergenerational transmission of rejecting parenting independent of the intergenerational transmission of depression; Doumas et al. (1994) reported that exposure to aggression anticipates aggressive behavior across three generations; and McCloskey and Bailey (2000, p. 1032) observed the following: “The presence of a maternal sexual abuse history placed girls at 3.6 times the risk for sexual abuse as other girls in the sample.” Apparently, then, there is no shortage of studies of intergenerational parallelism of deviant patterns. Thus, there would be no need to offer
66
Chapter 3
further analyses that provide such instances were it not for the following circumstances. First, many of the studies that comprise the literature on intergenerational parallelism of deviance are characterized by severe methodological limitations. In the great majority of studies of intergenerational continuities, some methodologically related distortions might have been introduced because the data were collected from only the parent or the child, or from the parents and children at different developmental stages in the life course. Second, these studies do not provide a systematic logic of procedure for explaining instances of intergenerational parallelism. In order to address these concerns, three instances of intergenerational parallelism are reported in which the data are provided by the parent and child, respectively, at comparable developmental stages. These relationships are used, then, as a basis for illustrating a systematic methodology by which the instances of intergenerational parallelism are explained by specification of moderating, mediating, and intergenerationally stable intragenerational causal variables.
Estimates of Intergenerational Parallelism As a baseline for gauging the influence of moderators, mediators, and common antecedents of intergenerational parallelism in deviant behavior and its correlates, the extent of intergenerational continuity in deviant behavior and two theoretically informed antecedents of deviant behavior was estimated. The two antecedents of deviant behavior were disposition to deviance and negative self-feelings. The three constructs were selected to reflect a basic theoretically informed model that was previously tested on the first-generation sample in a multigenerational study (Kaplan & Johnson, 2001). On theoretical grounds, it was specified that negative self-feelings associated with experiences of failure and rejection in the conventional world disposed the individual to engage in deviant behavior. The disposition to deviant behavior reflects both an alienation from the conventional world that was associated with the genesis of negative self-feelings and an attraction to deviant patterns that offered alternative responses that would assuage the distressful self-feelings and would provide experiences of success and acceptance that would be associated with more positive self-feelings. The disposition to engage in deviant behavior, in turn, would increase the likelihood of engaging in
Observing Intergenerational Parallelism in Deviance
67
such behavior through any of a number of routes, including the association with deviant peers. Deviant behavior, disposition to deviance, and negative self-feelings are measured as latent constructs, each of which is reflected in three measured variables. Deviant behavior is reflected in three multi-item indexes indicating engagement with theft, violence, and substance use, respectively, for both the first (G1) and second (G2) generations. For the G1 youths, the responses were reported for the previous year, whereas for the G2 youths, the responses reflected having “ever” engaged in the responses. Because the G2 youths were, on average, younger when responding to the baseline questionnaire and the G2 youths tended to report lower frequencies of deviant acts, the use of “ever” for the G2 youths increased the likelihood of tapping into deviant behavior for the second-generation subjects. The five-item theft index reflected taking things of different value (worth less than $2, between $2 and $50, more than $50), stealing from someone’s desk or locker, and joyriding (taking a car for a ride without the owner’s knowledge). The six-item interpersonal violence index affirms instances of beating up “someone who had not done anything to you,” using “force to get money or valuables from another person,” taking “part in a gang fight,” starting a fistfight, carrying “a razor, switchblade or gun as a weapon,” and “getting angry and breaking things.” Substance use was reflected in using “wine, beer, or liquor more than two times in the last week” and smoking marijuana. Disposition to deviance is reflected in three multi-item measures indicating, respectively, antisocial coping responses to stress (“If someone insulted me I would probably hit him,…take it out on someone else, …think about ways I could get even,…insult him back”), disrespect of others (self-reports of not being kind to others, not being a fairly honest person, telling lies, not having good manners), and rejection of conventional norms (affirmation of “I have a better chance of doing well if I cut corners than if I play it straight,” “The kids who mess up with the law seem to be better off than those who play it straight,” “If you stick to law and order you will never fix what is wrong with the country,” “The law is always against the ordinary guy”). Negative self-feelings is conceptualized as the confluence of negative self-evaluations and distressful self-feelings. Negative self-evaluations is measured by a seven-item index (self-derogation) indicating a need to have more respect for oneself, being dissatisfied with oneself, feeling that one does not have much to be proud of, feeling like a failure, not having
68
Chapter 3
a positive attitude toward oneself, thinking that the person is no good at all, and feeling useless at times. Distressful self-feelings is measured by a nine-item index of symptoms of anxiety (e.g., hands sweating, avoidance of competition in things that one is not good at, pressure or pain in the head, nervousness, trouble concentrating) and a six-item index of depressive affect (wishing to be as happy as others seem to be, feeling downcast and dejected, not getting a lot of fun out of life, most of the time not feeling in good spirits, on the whole not being a fairly happy person, not seeing the bright side of things). The structural equation models were estimated specifying the intergenerational parallelism of the three constructs. Each is considered in turn.
Deviant Behavior The estimation of the structural equation models specifying intergenerational parallelism in deviant behavior is summarized in Figure 3.1. The figure indicates the loading (unstandardized and completely standardized) of the measurement variables on the latent constructs as well as the structural effect indicating the degree of intergenerational parallelism for the 1,658 intergenerational units (the G1 youth’s deviant behaviors and the aggregated deviant behaviors of their children, the G2 youths, reported
Figure 3.1. Intergenerational Parallelism of Deviant Behavior
Observing Intergenerational Parallelism in Deviance
69
at approximately the same developmental stage by the representatives of the respective generations). The same data are reported for disposition to deviance (Figure 3.2) and negative self-feelings (Figure 3.3). As hypothesized, and consistent with the literature on intergenerational parallelism, a statistically significant but modest effect of G1 deviant behavior on G2 deviant behavior was observed (b = 0.240, B = 0.239). The measurement variables all had appreciable and statistically significant loadings on the respective G1 and G2 latent constructs. By the criteria of the Comparative Fit Index (CFI) index (0.996) and the the root mean square error of approximation (RMSEA) (0.026), a very good fit between the hypothesized and observed models existed. The fit was
Figure 3.2. Intergenerational Parallelism of Disposition to Deviance
Figure 3.3. Intergenerational Parallelism of Negative Self-Feelings
70
Chapter 3
improved by specifying an association between the theft and violence measures for the G1 deviance latent construct (b = 0.237, B = 0.237).
Disposition to Deviance As in the case of deviant behavior, a structural equation model was estimated that specified an intergenerational effect between G1 disposition to deviance and G2 disposition to deviance. The results are summarized in Figure 3.2. Again, the measurement variables manifested appreciable and statistically significant loadings on G1 and G2 latent constructs. The structural effect was statistically significant but modest in magnitude (b = 0.311, B = 0.216). Again, by the criteria of the CFI (0.990) and RMSEA (0.025), a quite good fit was observed between the data and the hypothesized model.
Negative Self-Feelings The estimation of the structural equation model specifying intergenerational parallelism for negative self-feelings is summarized in Figure 3.3. The measurement variables manifested strong and statistically significant loadings on the latent constructs for G1 and G2 negative self-feelings. As in the cases of disposition to deviance, a statistically significant but modest intergenerational effect was observed between G1 and G2 negative selffeelings (b = 0.385, B = 0.228). Again, by the criteria of CFI and RMSEA values, the observed model was a good fit to the hypothetical model. Discussion Although intergenerational transmission of deviant behaviors has been noted frequently, the association between outcomes in the earlier and later generations is less than perfect (Serbin et al., 1998): The intergenerational cycle of risk for poverty, crime, psychological distress, and illness is a well-established and stubborn social phenomenon, leading sociologists and journalists to refer to the existence of a ‘permanent underclass’ in North America (Chase-Lansdale & Brooks-Gunn, 1995; Furstenberg, Levine, & Brooks-Gunn, 1990; Wilson, 1987). Children growing up in disadvantaged conditions are likely to become the
Observing Intergenerational Parallelism in Deviance
71
parents of another disadvantaged generation, who, like their parents, are born with a high risk of serious psychosocial and health problems. However, longitudinal studies have revealed that psychosocial risk is, as the term denotes, probabilistic. Many children from ‘high risk’ backgrounds grow up to have reasonably prosperous and productive lives, despite their poor prospects at birth (Elder & Caspi, 1988; Furstenberg, Brooks-Gunn, & Morgan, 1987; Hardy et al., 1997; Rutter, 1987; Werner & Smith, 1992). (p. 1246)
Indeed, the incompleteness of cultural transmission might in many ways be a desirable state of affairs. Phalet & Schönpflug (2000) stated: Although cultural persistence is essentially a question of transmission (Cavalli-Sforza and Feldman, 1981), cultural transmission is not and cannot be complete. Rather, culture is being shaped and reshaped by the ongoing interactions between persons and groups and their environment. The process of cultural transmission does not lead to the replication of culture in successive generations; it falls somewhere in between an exact transmission (with hardly any difference between parents and offspring) and a complete failure of transmission (with hardly any similarity between the generations). Functionally either extreme would be problematic for a society. Complete transmission would not allow for novelty and change, and hence the ability to respond to new situations, whereas failure of transmission would not permit coordinative action between generations (Boyd and Richerson, 1985). (p. 489)
Not only are all instances of intergenerational parallelism far from perfect, in some cases no significant intergenerational parallelism is observed at all. For example, Smith and Farrington (2004) reported that second-generation child conduct problems did not predict thirdgeneration child conduct problems in a study of continuities in antisocial behavior and parenting across three generations. However, in that study, antisocial parents did predict conduct problems in second- and third-generation children. Generally, however, intergenerational overlap is observed to be statistically significant but modest in magnitude (Van Ijzendoorn 1992): Our review shows that the effect sizes indicating the amount of intergenerational transmission differ strongly between different research programs. The traditional research program using rather large samples and quite global questionnaire measures does not yield much evidence for a relation between parenting across generations. Effect sizes are somewhere between 3 and 15% of explained variance, and even these figures may well be inflated because of lack of control for contextual continuity. (p. 93)
72
Chapter 3
The above-reported results for the present study are not strikingly different from the range of effects reported in the literature; that is, we observed statistically significant but modest degrees of intergenerational parallelism in deviant behavior. However, that variation in the fact and degree of intergenerational parallelism has been observed and the modest magnitudes of parallelism that have been reported stimulates the question: Under what conditions is the magnitude of intergenerational continuity strengthened or weakened? The literature and analyses directed toward answering this question are considered in the following chapters.
4 Moderators of Intergenerational Parallelism
As noted previously, and as will be apparent throughout the volume, any given phenomenon in the second generation is not perfectly predicted by the same phenomenon in the first generation. For example, although there is an increased likelihood that children who run away had parents who ran away, only 24% of runaways in one report had parents who had also been runaways (Plass & Hotaling, 1995). Thus, the question is raised as to the conditions under which a phenomenon in the second generation is predicted by the same phenomenon in the first generation. Under what conditions will the association across generations be high or low? That is, under what conditions does intergenerational (dis)continuity occur? The less-than-universal observation of intergenerational continuity is noted by others as well (Oliver, 1993): The crude rates of intergenerational transmission of child abuse according to the studies reviewed are as follows: one-third of child victims grow up to continue a pattern of seriously inept, neglectful, or abusive rearing as parents. One-third do not. The other one-third remain vulnerable to the effects of social stress on the likelihood of their becoming abusive parents. (p.1315)
Whether this is surprising depends on one’s expectations at the outset. On the one hand, one would expect strong intergenerational parallelism in behaviors and traits because of the influence of one generation on another, genetic transmission, and stable environmental factors across generations. On the other hand, one would expect low levels of intergenerational parallelism because of changes that occur in the capacity of the 73
74
Chapter 4
first-generation subject to influence the second generation or in disruption of cross-generational stability of influences on the behaviors or characteristics in question, whether due to historical changes or to idiosyncratic influences on the (in)stability of causal influences. In any case, as others have observed (Rutter, 1998), the only moderate degrees of continuity noted in the cross-generational studies suggests the need to model discontinuities as well as continuities. In attempting to explain the variable degrees of intergenerational parallelism that have been reported, numerous investigators have turned to specifying the conditions that moderate the strength of the intergenerational relationship. The literature on moderators of degree of intergenerational parallelism will be discussed first, as background for the presentation of the theoretically informed estimation of models from the present study that specify the conditions for greater or lesser degrees of intergenerational parallelism in deviance and selected correlates of deviance.
Research Literature on Moderators The research literature on moderators of intergenerational parallelism variously speaks to general moderators without regard to the nature of the parallelism or to moderators of specific instances of intergenerational parallelism. In the former case, the issues variously relate to methodologically relevant moderators or to substantively significant social-psychological and social structural contingencies. When considering specific patterns of intergenerational parallelism, some studies do not necessarily focus on instances of deviant behavior but are, nevertheless, reviewed because they might be implicated in explanations of deviance. Other studies, however, explicitly focus on moderators of intergenerational parallelism of patterns that are generally recognized as deviant.
General Issues Consideration of moderators of intergenerational parallelism have considered both the methodological and substantive implications of findings.
Moderators of Intergenerational Parallelism
75
Methodological Moderators
Methodological features of research designs influence the degree of intergenerational correspondence that will be observed. The greater the similarity in circumstances between the two generations, the greater is the likelihood that the indicator of intergenerational parallelism will be stronger. For example, the stability between generations in an “inhibited behavior pattern” was stronger when parents had been assessed at an age closer to that of their toddler children (Cairns, Cairns, Xie, Leung, & Hearne, 1998). The need to assess the phenomenon of interest across generations at the same developmental stage was noted by Patterson (1998): For example, simply correlating children’s trait scores with parents’ trait scores immediately introduces two possible confounds. A possibility is that a trait score measured in children at 5 years of age may mean something quite different when it is compared with the trait measured in the parents at 25 years of age. It may also be that the trait scores are measured in very different ways for parents and for children. (p. 1263)
Substantive Moderators
Among the factors that are thought to moderate intergenerational parallelism are those related to broad social changes, family environment (whether related to such social changes), and intrafamilial relations. Thus, under conditions of rapid social change, we should expect a decreased degree of intergenerational parallelism than during periods of stability. Broad social changes might be noted, for example, with regard to increases or decreases in secularism in particular societies (Wadsworth & Freeman, 1983). If the dimension of social change is a correlate of deviant behavior (or some other phenomenon of interest), then broad sociohistorical changes over the intergenerational period should attenuate the relationship between first- and second-generation values on the variable of interest. If the second generation grows up in a more secular context than the first generation and religiosity is inversely related to deviant behavior, then fewer constraints would be placed on a individual disposed to engage in deviant behavior and less correspondence would be observed between the first and second generations’ acting out of deviant dispositions. Numerous intrafamilial factors have been considered regarding moderation of intergenerational (dis)continuities. As Simons and his
76
Chapter 4
colleagues (1992) suggested with regard to intergenerational transmission of parenting beliefs: Future research needs to focus upon the possible sources of this discontinuity. It may be, for example, that children are only influenced by the parenting behaviors displayed by their mother or father when they have a close, involved relationship with the person. Or, degree of consonance between parents with regard to parenting practices may be important. Thus parents may have a strong impact on parenting beliefs of their children when they engage in similar parenting styles, but exert little influence when they demonstrate dramatically different approaches to the role of behavior. Future studies need to investigate the extent to which such factors account for the modest degree of intergenerational transmission of parenting beliefs. (p. 834)
With regard to parent-child relations, presumably the more a child identifies with the mother and feels positive affect for the mother, the more the child will emulate the attitudes and behaviors of the mother. Conversely, intergenerational modeling might be moderated by the degree of conflict that characterizes the families insofar as such modeling is less likely to occur in the presence of conflict (Hetherington & Frankie, 1965). When parents and children have more satisfying relations, intergenerational parallelism appears to be stronger. Thus, although the association between value endorsements of parents and children were generally quite strong, the correlations tended to be appreciably stronger for low right-wing authoritarian parents who (it is speculated) might have been more responsive to the needs of their children (Rohan & Zanna, 1996). Consistent with this, intergenerational transmission of values on sexual permissiveness was hypothesized and observed to be moderated by the quality of family interaction (Taris, Semin, & Bok, 1998). Intergenerational transmission of values was greater when family interaction was said to be characterized by mutual respect and understanding (but see Taris, 2000). Among the salient moderating variables that have been suggested in accounting for intergenerational continuities is stage in the life cycle. For example, intergenerational parallelism in personality patterns (particularly the inhibited behavior pattern) is stronger when the characteristics are assessed earlier in life. However, this might suggest the operation of genetic rather than, or in addition to, environmental influences. Some moderators may be more distal in their influence, while other variables may be more proximal in their effects on the degree of intergenerational continuity or on the effects of putative antecedents of deviant patterns. More distal moderators may have their own correlates that are
Moderators of Intergenerational Parallelism
77
more proximally influential on the nature of such relationships. A case in point is gender. The moderating influence of the gender of the first generation parent is apparent in a number of studies. For example, data suggest that the similarity between parent and adolescent child level of individuation may be stronger for mothers than for fathers (Bartle & Anderson, 1991). Wickrama and associates (1999) reported a moderating influence of thechild’s gender on intergenerational continuity of healthrisk behaviors: The findings of the separate models for boys and girls demonstrated that (1) father’s health-risk lifestyles affected only boys’ health-risk lifestyle, whereas (2) mother’s health-risk lifestyle affected only girls’ health-risk lifestyle. (p. 258)
Variables such as gender of the first- and second-generation subjects and of the congruence between them might be surrogates for more proximal moderating variables. Thus, daughters in early adolescence are at a cognitive developmental stage when they can perceive similarity to a person with whom they have an important relationship and to idealize and identify with such persons. Mothers might be such significant role models who influence their daughters’ self-esteem through their role relationship, their nearness, as well as their common gender and sex role. In addition, they are characterized by psychological closeness (Curtis, 1991). A young girl’s identification with the mother continues throughout life, whereas a young boy’s identification with the mother is broken and switched to the father (Chodorow, 1978). At the same time, mothers identify with their daughters because of shared gender, projection of feelings about self, and the disposition to act toward the daughter as part of herself (Eichenbaum & Orbach, 1983). Women experience less social pressure to differentiate from their mothers than men experience to differentiate from their fathers (Flax, 1981). Further, women’s identity is based on different values than are men’s identities. A woman’s identity is based on values of affiliation and mutual relationships, perhaps arising from the women’s identification with a caregiving mother (Gilligan, 1982; Miller, 1986; Noddings, 1984).
Specific Intergenerational Relationships The literature contains numerous exemplars of reports of (1) moderators of intergenerational parallelism of phenomena that are not ordinarily
78
Chapter 4
taken to be (but are open to interpretations as being) instances of deviant behavior, but are frequently reviewed as correlated with, or causally implicated in, the genesis of deviant behavior and (2) moderators of intergenerational parallelism of phenomena that are generally interpreted as instances of deviant behavior. Correlates of Deviant Behavior
The literature is replete with studies of moderators of intergenerational continuities of phenomena that, where they are not regarded as instances of deviance, are reviewed as implicated in the etiology of deviant patterns. Among intergenerational patterns that are more frequently attended to with regard to moderating influences are teenage fertility, divorce and related familial patterns, religious behavior, poverty, and parenting patterns. A variety of factors are said to moderate intergenerational (dis)continuity in teenage parenthood. Thus, Hardy and her associates (1998) reported regarding intergenerational patterns of age at first birth: Girls who avoided teenage parenthood when being reared by an early childbearer were more likely to have mothers who were married, were not poor, were not on welfare, and had graduated from high school. In addition, these girls had higher IQ scores. So, even though they initiated sexual activity prior to age 16 at the same rates, had similar grade repetition rates, and at similar ages at menarche, they did not become mothers. We suspect that their households were characterized by more parental supervision and monitoring, as well as possibly by greater emphasis placed on education (Brooks-Gunn & Chase-Lansdale, 1995). (p. 1230)
Race has also been observed to moderate intergenerational patterns of teenage fertility. Kahn and Anderson (1992) observed: For whites, we find that intergenerational patterns of premarital fertility are explained almost completely by measures of socioeconomic and family context. Intergenerational patterns of teen marriage and marital fertility persist, however, even after these factors are controlled. For blacks, we only find intergenerational patterns of premarital fertility, and they are explained only partially by the socioeconomic and family context in which teenagers grow up. (p. 55)
They further speculated that broad social changes might increase the likelihood of recapitulation of the earlier generation’s behavior by the later generation (Kahn & Anderson, 1992):
Moderators of Intergenerational Parallelism
79
In contrast, intergenerational patterns of premarital teen fertility are likely to reflect the weakening of norms against nonmarital childbearing. One need only look at the declining rates of legitimation and relinquishment to see that childbearing outside of marriage simply has become more acceptable in our society. Within families, a mother’s prior experience with premarital fertility may serve as a role model for her daughter. Although our results generally support such an intergenerational pattern of premarital childbearing, it appears the impact may be weakening among more recent cohorts. Yet, before we become too excited about this trend, we should emphasize that the intergenerational pattern is weakening only because of the increasing prevalence of premarital childbearing in all sectors of the population. Thus, unless teens truly believe that a premarital birth will severely limit their future life options, there is little reason to believe that the overall trends will change. (p. 55)
Intergenerational parallelism in divorce is also moderated by a number of variables. The relationship between parental divorce and divorce by the children is stronger where the marriages of the children were of short duration and where the parents’ divorce(s) occurred when the children were 12 years of age or younger (Amato, 1996). Gender of children moderates the observed transmission of divorce from one generation to another. Thus, evidence has been reported of intergenerational transmission of divorce from parents to daughters but not from parents to sons (Feng, Giarrusso, Bengtson, & Frye, 1999). Broad social changes represent an important class of moderators of the intergenerational transmission of divorce. Between 1973 and 1996 a substantial decline in intergenerational transmission of divorce has been observed (Wolfinger, 1999). The attenuation of intergenerational continuities in divorce are compatible with explanations based on broad social changes that have occurred over the same period. In general, Americans have become more accepting of divorce (Cherlin, 1992; Philips, 1991; Thornton & Camburn, 1987; Wolfinger, 1999). These changing attitudes have had outcomes that might account for the attenuation of intergenerational transmission of divorce (Wolfinger, 1999): First, as divorce has become more common and people have become more accepting of it, children in divorced families may suffer less stigmatization, either real or imagined. In the past, when divorce was less common (see, for example, Cherlin 1992), single mothers and their children were frequently harassed or ostracized (Philips 1991). Under these conditions, children may have been less likely to develop normal relationships with their peers, mothers, or grandparents. This experience may have made divorce more traumatic for children, thereby leading to higher rates of divorce transmission earlier in the study period.
80
Chapter 4 A second consequence of the increased acceptance of divorce may be the changing circumstances under which couples choose to end their marriages. In the absence of no-fault divorce laws, a couple desiring a divorce often needed to demonstrate the total and absolute deterioration of their relationship. Normative expectations persuaded quarreling couples to ‘stick it out’ under circumstances such as domestic violence that today would be readily recognized as reasonable grounds for a divorce. When couples finally ended their marriages, the situation may have deteriorated far more than is typical in divorces today, thereby bringing greater harm to children. (p. 415)
Intergenerational discontinuity in religious beliefs was moderated by family disengagement. The greater proportion of individuals reporting change in religious beliefs (including abandonment) was observed in the group who reported seeing their parents least often (Wadsworth & Freeman, 1983). Consistent with this, intergenerational transmission of religious beliefs and practices appears to be moderated by perceptions of parental acceptance by the adolescent children (Bao, et al., 1999): Mothers’ church attendance and religious belief affected all aspect of boys’ religiousness when they perceived high or moderate acceptance from their mothers. This indicated that, for boys, a supportive mother-son relationship was an important facilitator in their initiation and modeling of their mothers’ religiousness. However, fathers’ religious beliefs and practices had substantial impact on girls’ concept of God when girls perceived low levels of acceptance from their fathers. Studies of projections of images of God have indicated that for females the God-father relationship was crucial and that the father’s image was important to the girl’s concept of God . . . .our results suggest that low perceived acceptance from fathers may be compensated by high religiousness on the part of the father in predicting girls’ concept of God . . . It probably happened to girls because girls’ concept of God was more related to parents’ attributes and parenting styles than boys’ concept was (Dickei et al., 1997). (pp. 370–371)
Kirkpatrick and Shaver (1990) also reported a significant interaction effect in that parent-child attachment had relatively little effect on the child’s religiousness when the parents’ religiousness was high. Intergenerational change in religious beliefs was moderated by education levels such that more educated young adult children were more likely to have changed religious beliefs; also, intergenerational discontinuity in religious beliefs was associated with self-assessments of probability of changing one’s social class (Wadsworth & Freeman, 1983).
Moderators of Intergenerational Parallelism
81
Although some degree of intergenerational continuity in poverty has been observed, the relationship is far from perfect. Children who grow up in poor families might not become poor adults; and downward social mobility might occur as well. Presumably, as Shlonsky (1984) noted: ...some selection takes place in the process of the transmission of poverty from parents to their offspring. Either not all the offspring in some poor families inherit their parents’ social position or not all the poor parents transmit their poverty to their offspring. Or, perhaps, the intergenerational transmission does not take place evenly in all social spheres. For instance, while sons and daughters may continue in their parents’ low social position in some aspects, they might advance into higher positions in the hierarchy in other regards. Thus, a son who falls, like his father, into the lowest income bracket, may have surpassed his father in terms of the level of education he attained. (p. 457)
In short, the degree of correspondence between first- and second-generation poverty is a conditional relationship. Intergenerational continuities in poverty depend, in part, on social norms and practices (Harper, Marcus, & Moore, 2003): Public action to tackle poverty and interrupt poverty transfers clearly also reflects social norms. Where there is a strong sense of collective responsibility for social welfare, there may be stronger support for public safety nets, resource distribution or good quality education for all, than where the wellbeing of the next generation is viewed largely as a private familial matter (Esping-Andersen, 1990). (p. 541)
Degrees of intergenerational continuity are moderated by a number of other factors as well. Thus, several constructs might be suggested to account for the observation that with each successive generation, a smaller degree of intergenerational transmission of socioeconomic stratum was observed (Biblarz, Bengtson, & Bucur, 1996): Haut (1984, 1988) shows that education may condition the relationship between occupational origins and destinations—the association between occupational origins and destinations was not significant among respondents with college degrees in a recent period. The occupational destinations of G2s may have been less determined by their origins than those of the G1s because the G2s may have had greater access than the G1s to higher education. The G3s may have had even more access in higher education than the G2s, but they also may have been the first of the three generations examined here to experience their parents’ divorces during childhood. Like education, the experience of family disruption also reduces intergenerational inheritance (Biblarz and Raftery, 1993), in part through its effect on family resources
82
Chapter 4 available to children. At the same time, family size (both the size of the family that children grow up in and the number of children they have as adults) and parental childrearing values and practices covary with each generation and may affect occupational inheritance and social mobility.... An alternative possibility is that the mother’s occupation begins to supplant the father’s occupation as a primary determinant of their offspring’s occupation. Perhaps the family transmission of occupation has not so much weakened as shifted from one parent to the other. Women have joined the paid labor force in increasing numbers over recent decades. As more mothers are employed (and as fewer families have a father present), the mother’s occupation may become a better proxy than the father’s occupation for family resources that affect the intergenerational transmission process. (p. 197)
Frequently, moderating variables are antecedents of moderators that are more proximal conditions for intergenerational (dis)continuity. A case in point is the moderating influence of being black on intergenerational continuity of poverty (Israel & Seeborg, 1998): Cohen and Tyree (1986) find that being black is one of the strongest predictors of the income mobility of the poor. Both economists (e.g., Borjas 1992; 1995) and sociologists (e.g., Wilson, 1987) have explored mechanisms making upward mobility difficult for blacks, especially lack of employment opportunities, neighborhood effects, and effects of government programs. These analyses are most often couched in terms of intervening variables. For example, Wilson (1987) argues that being black increases the probability of exposure to adverse social and economic conditions (i.e. underclass environment) which in turn reduces the probability of intergenerational movements out of poverty. (p. 757)
That is, being black has consequences that, in turn, moderate the degree of intergenerational continuity of poverty. Perhaps the most frequently investigated intergenerational patterns, particularly with regard to moderators, relates to parenting—whether the focus is on intergenerational transmission of abusive or more benign responses. The experience of stressful life events also appears to moderate the intergenerational continuity of abusive parenting (Egeland, 1988): Mothers abused as children who are currently abusing their children had very high life stress scores and lived in chaotic and disruptive environments. Living in poverty and experiencing large amounts of life stress appear to increase the likelihood of continuity of abusive patterns across generations. (p. 94)
Consistent with this, the experience of anxiety and depression by mothers who had experienced abuse appeared to increase the likelihood of intergenerational continuity of abusive parenting (Egeland, 1988):
Moderators of Intergenerational Parallelism
83
The continuity group had extremely high scores suggesting that the functioning of mothers in this group may have been impaired because of high levels of anxiety. Effective caretaking of an infant requires, among other things, a sensitivity to the baby, responsiveness, and flexibility. High anxiety is likely to interfere with the mother’s ability to be flexible and to tolerate frustration which often accompanies the care of young children. For example, high anxiety sensitizes people to noise, so a crying baby would be more difficult to tolerate. The highly anxious mother also is less likely to learn new approaches to child care and instead to rely on approaches with which she is most familiar. (p. 95)
However, under certain conditions, dysphoric affect might decrease physical abuse. Parental depression and posttraumatic stress disorder have been observed to moderate the intergenerational continuity of experiences of abuse (Pears & Capaldi, 2001): There was a significant interaction between parent depression and PTSD and a history of abuse, such that parents who had experienced high levels of abuse but who also demonstrated high levels of depression and PTSD were less likely to be abusive than parents who had experienced high levels of abuse but had low levels of depression and PTSD. It may be that parents who are depressed and experiencing symptoms of PTSD tend to withdraw from interactions with their children, making it less likely that they will be physically abusive. (p. 1454)
In any case, incompetence in parenting seems to be stress-inducing and to lead to continuity of abuse. Thus, intergenerational parallelism in physical abuse was moderated by parental consistency of discipline (Pears & Capaldi, 2001): Parents who experienced high levels of abusive acts and injuries but who were consistent in their discipline were less abusive than abused parents who were inconsistent disciplinarians. This is congruent with prior research suggesting that abusive parents may represent an extreme group of unskilled and ineffective parents (Burgess and Youngblade, 1988; Greenwald et al., 1997; Knutson and Bower, 1994; Zaidi et al., 1989). A parent who has very poor discipline skills is likely to experience a great deal of stress and frustration in dealing with their children. When this is coupled with a history of having received severe physical punishment, it is not surprising that the outcome might be the transmission of abusive treatment from one generation to the next. (p. 1454)
Not surprisingly, intergenerational continuity of abusive parenting tends to be moderated by the quality of attachment relationships with caregivers such that under conditions of poorer quality attachment
84
Chapter 4
relationships during childhood, the probability of transmission of abusive parenting across generations is increased (Zuravin et al. 1996). Further, continuity in cultural attitudes favoring abusive parenting practices might be expected to increase the degree of intergenerational continuity of experiences of abusive parenting (Cashmore & de Hass, 1995). The continuity might be the consequence of stability of social-class-related subcultural patterns of child rearing and, perhaps, social-class-linked stressors (Simons et al., 1995). Just as moderators specify conditions under which intergenerational continuity of abuse occurs, so also the moderators specify contingencies for intergenerational discontinuity. One review of the literature suggests that two major factors distinguish between nonrepeaters and repeaters of parental abuse (Putallaz et al., 1998): The first factor is that mothers who described themselves as victims of childhood abuse but were not abusing their own children reported involvement in an emotionally supportive relationship with a nonabusing adult during childhood or with a therapist at any point in their lives.... Nonrepeaters were also more likely to be receiving emotional support currently from the father of the baby and/or of the family members...or were involved in more satisfying interpersonal relationships...presumably these supportive relationships functioned to modify therapeutically the feelings and expectations of these parents....The second factor is that nonrepeating parents were aware of their abuse and spoke of it in a coherent, integrated fashion with much emotion and detail and further verbalized how they wanted to raise their children. In contrast, the parents who abused their children spoke of their abusive history in vague generalities without emotion, were unable to recall specific experiences, and provided idealized descriptions of parents.... (p. 392)
As a specific example, Egeland (1988) reported with regard to a study of the intergenerational transmission of abusive parenting: Our attempts to account for the exceptions to continuity from one generation to the next yielded some interesting findings which made intuitive as well as theoretical sense.... The most compelling findings were in the area of relationships, where we found that mothers who broke the cycle of abuse were as children more likely to have an emotionally supportive relationship with another adult and were as adults more likely to have an emotionally supportive husband or boyfriend.... These emotionally supportive relationships provided the parents with an alternative working model of relationships which is different from the abusive relationships they experienced as a child. Parents who experienced abuse and other forms of maltreatment as a child carry a set of expectations regarding self and others into adulthood. Without an alternative model, parents abused as children develop
Moderators of Intergenerational Parallelism
85
expectations that care and nurturants are not available and that they are not a lovable, valuable person. Carrying the model of poor quality relationships experienced in childhood into adulthood is a likely explanation for continuity of parental maltreatment. (p. 92)
Numerous other investigators confirm that these or similar variables moderate intergenerational continuity in parenting patterns, and they suggest other factors that function in this way: Social support has been suggested as a moderator of the intergenerational transmission of abuse (Hunter & Kilstrom, 1979); intergenerational continuity is weakened when parents are less demanding of their children (Caliso & Milner, 1994); religious affiliation has also been observed to moderate intergenerational continuity of experiences of abuse (Kaufman & Zigler, 1993); and the intergenerational transmission of abusive parenting experiences can be moderated by interventions that enhance self-esteem and provide more acceptable techniques for adapting to adverse life circumstances (Youngblade & Belsky, 1990). As in the case of other intergenerational patterns, some moderators are more distal in their effects, operating through more proximal contingencies. Cases in point are the genders of each generational representative and the congruity of gender across generations. A good deal of evidence exists that the gender of the earlier generation moderates intergenerational continuity. Thus, Simons and his associates (1991) reported gender differences concerning the strength of the direct paths between parenting of two successive generations. They reported that the association between the generations was much stronger for the mothers than for the fathers, a finding that might be accounted for by a culture that assigns mothers the primary supporting role and fathers the supporting role in the socialization process. They also reported that harsh discipline by the grandmother, but not by the grandfather, had an effect on the mothers’ severe discipline of boys and girls and on the fathers’ harsh parenting of boys. Similarly, Van Ijzendoorn (1995) reported on the basis of a meta-analysis that there was a stronger association between attachment representations of mothers and their children than between fathers and their children possibly because fathers ordinarily play a lesser role in child care during the early years. However, the father’s influence might increase during the child’s later years. Finally, intergenerational relationships between parents’ and childrens’ recollection of parental responses are observed to be moderated by the gender of the parent and offspring Lundberg, Perris, Schlette, and Adolfson (2000) found that:
86
Chapter 4 ...the correlation concerning emotional warmth is stronger in the father-son than in the father-daughter constellation. Also, correlations concerning emotional warmth are nearly nil in the mother-son constellation and much less pronounced than those concerning the fathers and the mother-daughter constellation. The experience of overprotection appears to be modestly correlated in the father-son constellation and in both the mother-son and the mother-daughter combinations. (p. 874)
Deviant Behaviors
As is the case with putative causes or correlates of deviant behavior, intergenerational continuity or discontinuity in deviant patterns is influenced by a number of contingencies. Some observers perseverate on and suggest possible moderators of intergenerational discontinuity of deviant behavior (Smith & Farrington, 2004). Discontinuities are potentially more important than continuities, not least because they shed light on processes by which risk may be interrupted (Rutter et al., 1998; Rutter, 1998). Since data suggest that continuing antisocial behavior, pairing with antisocial mates, and high levels of marital conflict are implicated in continuity, it might be important to look further at the G3 children of G2 antisocial males who have harmonious relationships with a non-antisocial spouse to see if such a relationship moderated the risk for the child. To the extent that some of the most antisocial fathers may not be living with their children, this may also be protective to the extent that these fathers may parent poorly and their relationships with partners may be highly disruptive. Although further investigations of this issue is beyond the scope of this paper, several studies have suggested that partnerships with prosocial mates represent turning points and discontinuity within the life span in antisocial trajectories, even for those with genetic risk factors (Farrington and West, 1995; Sampson and Laub, 1993). (p. 244)
A number of these suggestions find support in the research literature, but the effects are complex in nature. Some evidence exists that the degree to which an antisocial parent is present in the home moderates the degree of intergenerational continuity in antisocial behavior. Thus, when fathers with criminal backgrounds were not absent for periods greater than 6 months before the child’s 17th birthday, the male child was more likely to be arrested than the children of fathers with criminal backgrounds who were absent for at least 6 months. For noncriminal fathers, criminality was associated with the absence of the fathers (McCord, 1991). However, Tapscott and associates (1996) found that the degree of exposure to the
Moderators of Intergenerational Parallelism
87
antisocial father did not moderate intergenerational continuity between paternal antisocial personality disorder and child conduct disorder. The presence/absence of an antisocial parent might or might not be associated with identification with the parent. In any case, the degree of identification with a parent moderates intergenerational continuity in deviant behavior. Thus, subjects who had been exposed to aggression and identified strongly with their fathers had significantly higher scores on current relationship aggression than did subjects who had a low identification with their fathers (MacEwen, 1994). Identification might be facilitated by warmth; thus, intergenerational (dis)continuity in deviance might be moderated by parental warmth. However, the hypothesis that a more positive relationship with the problem-drinking parent will lead to greater risk of excessive drinking in the second-generation subjects finds some support only for women with fathers with drinking problems. The prediction that the offspring most at risk have had more negative relationships with parents finds support in the subgroup of men with mothers with drinking problems. In this subgroup, offspring were more at risk for drinking if they had poor relationships with their fathers (Orford & Velleman, 1991). Imitation of abstention or lowvolume alcohol-use parents tends to occur under conditions where the dominant parent shows relatively strong religious beliefs and has no signs of problem drinking and where the offspring feels strong affection for both parents, and they hold relatively strong religious beliefs and attend church regularly (Harburg, et al., 1990). Where social changes indicate, for example, a relaxation of many informal and formal social controls, one might expect individuals in successive generations measured at the same developmental stage to differ in performance of deviant behavior. Straus and Gelles (1988) reported nationwide decreases in the incidence of child and spouse abuse between 1975 and 1985. These data were compatible with explanations in terms of “changes in American society that took place during or immediately before the decade of this study, including: changes in the family, in the economy, in the social acceptability of family violence, in alternatives available to women, in the social control processes, and in the availability of treatment and prevention services” (p. 475). Thus, social changes could disrupt the continuity of intergenerational transmission of related phenomena. As with correlates of deviant behavior (which might or might not be interpretable as reflections of deviance), intergenerational parallelisms
88
Chapter 4
in deviance have been reported to be moderated by distal variables such as gender and gender congruity. Thus, Jankowski and her associates (1999) reported that subjects who observed their same-sex parent exclusively manifest physical aggression against their spouse were increasingly likely to manifest dating aggression, whereas subjects who exclusively witnessed their opposite-sex parent display marital aggression were not at greater risk for displaying dating aggression. These researchers also cited other studies indicating that same-sex parents are prominent models for their children (Bussey & Bandura, 1984; Perry & Bussey, 1979). In another study (Doumas, Margolin, & John, 1994), exposure to aggression predicted aggressive behavior across three generations for males, but not for females. For females, however, marital aggression in the first generation predicted husband-to-wife marital aggression in the second generation. Gender also appears to moderate intergenerational parallelism in substance use as well. Thus, parental substance use is more strongly related to female adolescent substance use than to male adolescent use (Clayton, 1991; Thompson & Wilsnack, 1984). Smoking by the mother is more strongly related than smoking by the father to children’s smoking, particularly daughter’s smoking (Kandel & Wu, 1995). More imitation in drinking level occurs between offspring and same-sex parent, while nonimitation is associated with cross-sex parent drinking, particularly for the only female single child displaying a negative correlation with father’s drinking level (Harburg, Davis, & Caplan, 1982). Specifications of distal moderators by themselves are not very informative. It is only when “mediational moderators” are offered that we begin to approach an understanding of the moderating conditions for intergenerational (dis)continuity of deviant behavior. For example, Stith and her associates (2000) observed gender effects in their meta-analysis of the intergenerational transmission of spouse abuse and offered possible explanations for the effect: There are a variety of explanations for why growing up in a violent home may have a stronger relationship to victimization for female offspring than to perpetuation for male offspring. One possible explanation may be that children are modeling the behavior of their same-sex parent. Differential socialization practices may also help to explain these findings...Boys are reinforced more often for being aggressive, whereas girls are reinforced for being passive. Thus cultural socialization practices may interact with modeling of same-sex parent behavior, leading to differential effects for boys and girls growing up in violent homes. (p. 648)
Moderators of Intergenerational Parallelism
89
Summary The literature offers a plethora of instances in which specific variables are observed to moderate degree of intergenerational (dis)continuity in patterns of deviant behavior or their correlates. At the very least, it might be concluded that these exemplars of intergenerational parallelism are conditional on a number of factors. It might even be possible to suggest that some of the moderators are particularly meaningful as contingencies for observing intergenerational (dis)continuity. Certainly variables such as changes in normative definition, identification, family conflict, parental warmth, gender, and related variables surface again and again. However, in the absence of a general theoretical structure, it is not possible to organize these variables in a way that contributes to an understanding of intergenerational continuity. Are these variables causally related to each other or alternative specifications of the same theoretical construct? How should any one of these variables be interpreted theoretically—does absence of the father from the home signify lack of a role model, attenuated social support, an occasion for stigmatization, loss of income, weakened social control, or any combination or all of these? It is not likely that any interpretations will be credible unless it is viewed in the context of a general theory/nomological network. Although these questions will not be definitively answered even with being guided by a theoretical framework, certainly the attempt to explain intergenerational discontinuity will proceed in an orderly fashion if it is guided by such a theory. Therefore, informed by a general theory, a number of models were estimated that specified conditions under which intergenerational (dis)continuity in deviant behavior and causally related constructs would be observed.
Estimating Moderators of Intergenerational Parallelism From the review of the literature, it is apparent that intergenerational continuities in deviant behavior and their hypothetical causally related constructs are contingent on a variety of moderating circumstances. Instances of intergenerational parallelism are variable in the nature of the circumstances that moderate the relationship. As expected, each of the three focal constructs under consideration (deviant behavior, disposition to engage in deviance, and negative self-feelings) has their own theoretically indicated moderators.
90
Chapter 4
Deviant Behavior According to the general theory of deviance that guides the analyses presented in this chapter, the continuity of deviant behavior depends on at least two general factors. First, the first-generation subjects (G1) must necessarily be alienated from the conventional world. Under conditions of such alienation, conventional influences are less likely to stimulate changes in deviant behavior over the first-generation subject’s life course. If changes in deviance had occurred over the first-generation subject’s life course, they would be less likely to transmit deviant attitudes to this next generation and would be less likely, as well, to model such behavior for the next generation (G2). Second, intergenerational continuity in deviant behavior is more likely to occur under conditions where an environment conducive to deviant behavior existed for the second-generation (G2) subject. If the environment was less than conducive to or supportive of deviant behavior that might otherwise have been transmitted by the first-generation subject to the second-generation youth, there is a lesser likelihood that parallelism in deviant behavior across generations would be observed. To test these hypotheses, the structural equation models estimating the degree of intergenerational parallelism were tested and compared under mutually exclusive conditions. For multi-item indexes, the groups were defined as being above or below the median value of the score. For single-item indexes, the groups were defined by having affirmed or denied the item. Where there were more than one eligible G2 child in the family, the scores for the G2 youths were aggregated. The latent construct for deviant behavior was modeled as described in chapter 3. All variables were selected to reflect alienation from the conventional world and contexts conducive to continuity of deviant behavior, respectively. For each of four variables (see Tables 4.1–4.4), it was hypothesized that subjects that were high on the variable indicating alienation from the conventional world or an environment conducive to support for deviant behavior, the degree of intergenerational parallelism noted in deviant behavior would be significantly greater than that observed for individuals who were low on the variable in question. Three variables were taken to reflect G1 alienation from the conventional environment and one variable was taken to reflect environments conducive to the continuity of deviance in the second generation.
91
Moderators of Intergenerational Parallelism
Alienation from Conventional Others
One indicator of identification with, or alienation from, the conventional world is the extent to which values held by one’s peer group are endorsed by the subject. Failure of G1 youths to endorse values of their age cohort is taken as an indicator of alienation form the cohort. The values in question relate to appearance (“important to be good looking”), athletics (“...good at sports”), and popularity (“...liked by kids of the appropriate sex”, “...have a lot of friends”). That these values are reflective of broad acceptance is suggested by the large majority endorsing them in each case. The comparison of the measurement models and structural effects for intergenerational parallelism of deviance under the two mutually exclusive conditions is summarized in Table 4.1. Referring to the unstandardized (b) coefficients, as expected, for individuals who judged these values as “not important,” and therefore appeared to be alienated from the value system of their age cohort, the degree of intergenerational continuity in deviant behavior (b = 0.316) was appreciably higher than for youths who endorsed the values and therefore were more likely to be subject to conventional internal and external evaluative judgments (b = 0.195). It was reasoned that alienated youths who rejected conventional values would be less subject to pressures to forego continued deviant activities and so would be more likely to continue the deviant activity into Table 4.1. Intergenerational Parallelism of Deviant Behavior by Endorsement or Rejection of Conventional Peer Values by G1 Youths Endorse Conventional Peer Values (N = 1,041) b G1 deviant behavior Theft Violence Drugs G2 deviant behavior Theft Violence Drugs Intergenerational parallelism effect a
Fixed parameter.
B a
Significance
Reject Conventional Peer Values (N = 617) b
B
Significance
1.000 0.785 1.454
0.627 0.492 0.911
0.000 14.063 3.397
1.000 0.756 1.297
0.658 0.497 0.853
0.000 8.652 4.175
1.000 0.850 0.472
0.734 0.717 0.611
0.000 17.249 14.551
1.000 0.865 0.517
0.747 0.698 0.625
0.000 12.747 10.975
0.195
0.188
4.113
0.316
0.330
4.920
92
Chapter 4
adulthood, and so serve as deviant models for G2 youths, or (through deviant parenting) would induce stressful life experiences and maladaptive deviant coping patterns to assuage the distress experienced in conjunction with the life experience. For individuals who were low on the score measuring rejection (“not important”) of the salience of peer values, the intergenerational continuity of deviant behavior was appreciably and significantly less (albeit still significant) (b = 0.195). Presumably, for individuals who endorsed conventional peer values, those G1 youths who engaged in deviant behavior would be subject to pressures to discontinue deviant behavior. The conventional peers would serve as a positive reference group and the expectations of conventional groups would deter (to some extent) continuity of deviant behavior throughout the G1 life span, and so the occasion to serve as models or stimuli for deviant adaptations in G2. The significant (but lower) G1-G2 correlation under the condition of endorsing conventional values would be expected because of consequences (negative sanctions) of G1 deviance that would to some degree confirm deviant disposition throughout the G1 life course. By the same reasoning that informed the hypothesis that rejection of conventional peer values would moderate the intergenerational parallelism of deviance, it was hypothesized that the perception of rejection by conventional peers would moderate the degree of intergenerational parallelism that was observed. Perceived rejection by the kids at school by G1 youths was measured by a two-item index indicating that the kids at school did not like them very much and that they were not very good at the kinds of things that the kids at school think are important. The contrasting structural equation models estimating the effect of G1 deviant behavior on G2 deviant behavior for G1 youths who were high and low respectively on perceived rejection by school peers are summarized in Table 4.2. As hypothesized, intergenerational continuity of deviant behavior was moderated by G1 perceptions of rejection by kids at school. For G1 youths who asserted that the kids at school did not like them very much and/or that they were not very good at the kinds of things that the kids at school think are important, the degree of intergenerational parallelism in deviance was appreciably greater (b = 0.319) than for G1 youths in the mutually exclusive category (b = 0.156). Indeed, the coefficient was twice as strong in the former group than in the latter group. The sense of alienation from the conventional peer group due to perceived failure and rejection leads to an attenuation of identification with the conventional normative framework and a need to seek alternative
93
Moderators of Intergenerational Parallelism Table 4.2. Intergenerational Parallelism of Deviant Behavior by G1 Youth’s Perceived Rejection by Conventional Peers
G1 deviant behavior Theft Violence Drugs G2 deviant behavior Theft Violence Drugs Intergenerational parallelism effect
High G1 Perceived Rejection by Conventional Peers (N = 897)
Low G1 Perceived Rejection by Conventional Peers (N = 761)
b
Significance
b
B a
B
Significance
1.000 0.679 1.459
0.618 0.419 0.901
0.000 9.357 4.286
1.000 0.905 1.164
0.713 0.645 0.830
0.000 14.946 3.212
1.000 0.804 0.453
0.781 0.713 0.627
0.000 16.696 13.659
1.000 0.905 0.522
0.697 0.701 0.575
0.000 13.292 11.955
0.319
0.283
5.621
0.156
0.188
3.058
a
Fixed parameter.
deviant adaptations. Under these conditions, the G1 subjects who engaged in deviance would be likely to continue the deviant responses throughout the life course and either (1) serve as socialization models for G2 deviance or (2) (through deviant parenting) motivate deviant adaptations while depriving the G2 youths of models for conventional coping patterns. Continuity of deviance depends in part on the (in)effectiveness of social controls. In the presence of strong social controls, youths who engage in deviance (for whatever reason) are less likely to continue deviant behavior throughout the life course. Where social controls are weak, youths are less likely to be deterred from, or are more motivated to continue, deviance. One indication of strength of social controls is the salience of attitudes of conventional others. It is to be expected that among youths who care what conventional others think of them, there would be less continuity (greater discontinuity) of deviant behavior than among youths who do not care what conventional others think of them (i.e., for youths who are not uncomfortable with their deviance). Reasoning that youths who were not uncomfortable with their deviance (or were comfortable with their deviance) would say that they were not bothered very much when their parents, or teachers, or kids at school disapproved of their behavior, whereas youths who were uncomfortable with their deviance would admit to being bothered by disapproval by these (apparently significant) others, an additive score was contrived
94
Chapter 4
consisting of three items: “When I do something my parents (teachers, the kids at school) dislike it bothers me very much.” The distribution was divided into those G1 youths who were more and less bothered by disapproval (of their behavior) by parents, teachers, and the kids at school. It was expected that for those who cared less about others’ disapproval, a greater degree of intergenerational parallelism in deviance would be observed in comparison with the degree that was observed for youths who cared more about others’ opinions. The contrasting structural equation models estimating the effects of G1 deviant behavior on G2 deviant behavior for G1 youths who cared more or less about conventional others’ opinion of their behavior are summarized in Table 4.3. As expected, intergenerational parallelism in deviance was significantly greater for youths who were less bothered when significant others disliked their behavior (b = 0.288) than when they were more bothered (b = 0.176). The greater continuity in deviance for the “less bothered” group is accounted for, in part, by the relative inefficiency of any sanctions that might have been administered by significant others in response to G1 deviant behavior. If you do not care what others think, then you are less likely to modify your behavior in response to their expressed attitudes. The lesser (but still statistically significant) intergenerational continuity in deviant behavior observed for the “more bothered” group is accounted for, in part, by G1 modification of deviant behavior in
Table 4.3. Intergenerational Parallelism of Deviant Behavior by G1 Youth’s Emotional Investment in the Evaluations of Conventional Others Bothered More by Evaluations of Conventional Others (N = 941) b G1 deviant behavior Theft Violence Drugs G2 deviant behavior Theft Violence Drugs Intergenerational parallelism effect a
Fixed parameter.
Bothered Less by Evaluations of Conventional Others (N = 717)
B
Significance
b
1.000 0.846 1.812
0.547 0.462 0.990
0.000 12.305 2.296
1.000 0.718 1.158
0.708 0.508 0.820
0.000 10.608 4.576
1.000 0.880 0.453
0.725 0.717 0.565
0.000 15.674 13.837
1.000 0.811 0.510
0.762 0.695 0.658
0.000 13.741 11.566
0.176
0.161
3.338
0.288
0.288
4.870
a
B
Significance
Moderators of Intergenerational Parallelism
95
response to perceived disapproving attitudes on the part of significant others. That some degree of continuity is still observed for G1 youths who care less about others’ evaluations is explainable in terms of consequences of G1 adolescent deviant behavior (alienation producing, negative social sanction, G1 deviant parenting patterns, G1 adult deviant modeling) that have implications for G2 deviant behavior. Indeed, not caring might be a defensive reaction to earlier negative sanctions by significant others to G1 deviance. In order to assuage the distressful self-feelings consequent on G1 deviant behavior, the G1 youths denigrate the sources (and therefore the legitimacy) of the negative sanctions. In sum, G1 alienation from the conventional world moderates the degree to which G1 deviant behavior is associated with G2 deviant behavior. G1 alienation is reflected in the rejection of conventional values or normative standards, perceptions of being rejected by conventional others, and emotional withdrawal from conventional others who might have served as agents of social control. Presumably, G1 alienation increases intergenerational parallelism through facilitating G1 intragenerational continuity of deviant behavior into adulthood. Adult deviance influences G2 deviance through (1) modeling deviant adaptations, (2) inducing stressful life circumstances and concomitant negative self-feelings that, in turn, evoke deviant adaptations by G2 youths, (3) failing to provide conventional social controls, and (4) facilitating access to supportive deviant environments. Indeed, this last consequence reflects a second class of moderators of intergenerational parallelism of deviant behavior. Supportive G2 Social Contexts
In addition to G1 circumstances reflecting alienation from the conventional world that moderate the intergenerational continuity of deviant behavior, the social contexts of G2 youths would be expected to moderate the degree of intergenerational continuity. The likelihood that G1 adolescent deviant behavior would influence G2 deviant behavior via adult deviant behavior, and under conditions of weakened social controls, would be magnified when the G2 social context is supportive of G2 deviant behavior. An important such context would be the G2 (perceived) peer environment at school. Presumably, where deviant behavior is prevalent among peers and, therefore, is to a degree normalized, there would be less resistance to
96
Chapter 4
engaging in deviant behavior given the disposition to do so (as the object of dysfunctional parenting patterns or as the observer of deviant parental role models). Further, the prevalence of deviance among one’s peers would offer models for learning deviant patterns and occasions to practice deviant behaviors. In short, where deviant behaviors are prevalent, youths would be more likely to perceive social support for, learn, and have opportunities to engage in deviant behaviors. Thus, where deviant behaviors were perceived as more prevalent, youths would be more likely to enact deviant behaviors, and where deviant behaviors were viewed as less prevalent, youths would be less likely to act out deviant dispositions. To test for the moderating influence of perceived prevalence of peer deviance, the G2 youths were divided into two groups as relatively high or low on a four-item index indicating whether “many of the kids at school” were perceived by G2 youths as engaging in deviant acts: purposely damage or destroy public or private property that does not belong to them; carry razors, switchblades, or knives as weapons; take little things worth less than $2 that do not belong to them; beat up people who have not done anything to them. For each of the two groups, a structural equations model was estimated that specified intergenerational parallelism in deviant behavior. The results of the analysis are summarized in Table 4.4. As hypothesized,
Table 4.4. Intergenerational Parallelism of Deviant Behavior by G2 Youth’s Perception of Peer Deviance at School Lower Prevalence of Peer Deviance (N = 1,130) b G1 deviant behavior Theft Violence Drugs G2 deviant behavior Theft Violence Drugs Intergenerational parallelism effect a
Fixed parameter.
B a
Significance
Higher Prevalence of Peer Deviance (N = 528) b
B
Significance
1.000 0.818 1.512
0.610 0.499 0.922
0.000 14.158 3.541
1.000 0.704 1.410
0.646 0.454 0.910
0.000 8.862 3.554
1.000 0.910 0.502
0.672 0.680 0.545
0.000 16.194 14.989
1.000 0.783 0.483
0.773 0.679 0.659
0.000 11.160 8.765
0.163
0.203
4.315
0.350
0.279
4.256
Moderators of Intergenerational Parallelism
97
magnitude of intergenerational continuity was appreciably greater (more than twice the size) for youths who were higher (b = 0.350) than for the youths who were lower (b = 0.163) on the index of perceived deviance by “many of the kids at school.” Thus, the findings support the hypothesis that the degree of intergenerational parallelism is moderated by G2 perceptions of the prevalence of deviance among school peers. In sum, the findings reported thus far suggest that intergenerational continuity in deviant behavior is moderated by the following: degree of (lack of) endorsement of conventional peer values; perceived failure according to peer standards and consequent rejection by conventional peers; (not) caring about the attitudes of “significant” others toward the behavior of the G1 youths; and G2 social contexts that provide support for and opportunities to learn and enact deviant patterns. The findings are congruent with a theoretical framework that specifies intragenerational continuity of G1 deviant behavior over the life course to be a function of alienation from conventional normative systems and those who adhere to those systems and the sequelae of such alienation: (1) weakened social controls that otherwise have forestalled continuity of deviant behavior and (2) motivation to seek alternative (deviant) adaptations to achieve gratifications that were frustrated in the context of the conventional normative system. The influence of the G1 life course on G2 continuity of deviance is presumed to be effected through (1) the G2 youth being part of a stigmatized familial network and correlated G1 deviant parenting patterns that alienate the G2 youth and lead to deviant adaptations to assuage emotional distress that is secondary to the G1 parent’s stigmatizations, abuse, and indifference and (2) modeling of G1 deviant behaviors that are imitated by G2 youth who simultaneously are deprived of models for more efficacious and conventional modes of coping. These effects of G1 adults on G2 deviance are expected to be magnified under conditions where the G2 youths operate in the context of a deviant peer environment that provides emotional support, models, occasions, and opportunities for acting out deviant dispositions. As will be noted further in chapter 6, intergenerational parallelism in deviance is explainable, in part, in terms of continuity of causes of deviance in each generation. For example, in both G1 and G2, deviant attitudes would be expected to increase the likelihood of acting out deviant behaviors; also, intergenerational parallelism in deviance would be accounted for by continuity in deviant attitudes across generations. The deviant attitudes in G1
98
Chapter 4
would have consequences that increase the likelihood of G2 deviant attitudes and, so, intergenerational parallelism in deviance might be explainable in terms of the intergenerational continuity of common intragenerational causes of deviant attitudes in each generation. In any case, instances of intergenerational continuity of putative common intragenerational causes of deviant behavior have their own moderators. Two instances of intergenerational continuity, with regard to their moderators, will be considered because of their predominate theoretical significance as intragenerational causes of deviance: moderators of (1) intergenerational continuity of deviant dispositions and (2) intergenerational continuity of negative self-feelings.
Disposition to Deviance Intergenerational stability in disposition to deviance (and, indeed, in any disposition) is contingent on circumstances that facilitate (1) intragenerational stability in G1 deviant dispositions between youth and adulthood and (2) intergenerational influences of G1 adults on G2 youths. Intragenerational Stability
In general, deviant dispositions would be expected to be continuous throughout the life course (as would any dispositions) to the extent that youths did not experience discomfort in their social relationships whether that is reflected in rejection by others, new expectations incumbent upon one, thwarted aspirations, or simply disquietude regarding meeting others’ expectations. Youths who were comfortable with who they were and their current station in life would have no need to change their disposition to behave; nor would they be moved to change if significant others were comfortable with (or were perceived as comfortable with) the youths and their deviant behavior. However, discomfort with oneself, whether caused by others’ discomfort with the youth’s behavior or by frustrated aspirations or by new expectations, would increase the need to change one’s attitudes. Discomfort with one’s self-evaluations would instigate changes in reference groups and correlated attitudinal changes. Both the experience of (dis)comfort and the stability or change in reference groups are caused by or reflected in one’s experiences in social relationships or groups or in changing circumstances that influence such experiences.
99
Moderators of Intergenerational Parallelism
Youths characterized by deviant attitudes who, nevertheless, were not negatively sanctioned as adults but, rather were made to feel comfortable in their social networks would be more likely to continue their attitudes throughout the life course, and so communicate these attitudes to their children. Conversely, youths characterized by deviant attitudes who were negatively sanctioned and for this or other reasons were made to feel uncomfortable in their peer groups would be less likely to continue their deviant dispositions throughout the life course, and so would be less likely to communicate these attitudes to their children. In order to test this reasoning, structural equation models specifying intergenerational parallelism in disposition to deviance were estimated for two groups: those who affirmed and those who denied as young adults, that they felt comfortable with their friends. In this and all other multigroup models dealing with moderators of disposition to deviance, the latter construct was modeled as described in chapter 3. The comparison of the models is summarized in Table 4.5. As hypothesized, the magnitude of intergenerational parallelism in deviant attitudes was observed to be appreciably greater for G1 young adults who asserted that they felt comfortable with their friends (b = 0.404) than for those who asserted that they were not comfortable with their friends (b = −0.012, not significant).
Table 4.5. Intergenerational Parallelism of Disposition to Deviance by G1 Affirmation/Denial as Young Adults That They Feel Comfortable with Their Friends G1 Young Adult Does Not Feel Comfortable with Friends (N = 313) b G1 deviant behavior Theft Violence Drugs G2 deviant behavior Theft Violence Drugs Intergenerational parallelism effect a
Fixed parameter.
B a
G1 Young Adult Feels Comfortable with Friends (N = 1,345)
Significance
b
B
Significance
1.000 1.119 1.317
0.545 0.611 0.719
0.000 5.597 5.094
1.000 1.289 1.400
0.452 0.583 0.633
0.000 8.353 7.509
1.000 0.400 0.646
0.614 0.409 0.518
0.000 3.975 3.343
1.000 0.405 0.649
0.694 0.491 0.587
0.000 12.243 12.438
−0.012
−0.010
−0.102
0.404
0.264
4.902
100
Chapter 4
Presumably, where the youth does not (or ceases to) find gratification in groups that are supportive of deviant attitudes and where the youth develops affection for other positive reference groups, the motivation to maintain the same behavioral disposition is attenuated. Because the G1 dispositions to behave are not stable, any influence on G2 disposition to replicate the G1 youth’s attitudes would similarly be weakened. Consistent with the previous reasoning, it was expected that G1 young adults who perceive themselves as rejected by their friends would feel alienated and change their attitudes that were previously endorsed by the group either as a reflection of alienation from the group or in order to feel accepted by an alternative positive reference group that eschewed those antisocial attitudes. In either case, intragenerational continuity of antisocial attitudes over the G1 life course would be diminished, as would the antisocial influence of G1 adults on G2 antisocial attitudes whether by (1) providing antisocial models or (2) inducing stressful experiences and consequent deviant adaptations on the part of G2 youths. Conversely, G1 subjects characterized by youthful antisocial dispositions, but who as young adults do not feel unwanted, would be more likely to continue to hold such attitudes throughout the life course within a social (peer) context that either endorses or tolerates such attitudes. Continuing to hold such attitudes as adults increases the likelihood of transmitting them to the next generation as role models or by eliciting G2 adaptations in response to distressful correlates of the antisocial attitudes. In order to test this reasoning, structural equation models specifying intergenerational parallelism in disposition to deviance were estimated for two groups of subjects differentiated according to whether the G1 young adults did or did not feel unwanted by their friends. The results are summarized in Table 4.6. Presuming that the G1 youths who were higher on disposition to deviance but as young adults did not feel unwanted by their friends indicated an interpersonal network that was supportive of deviance, it was expected that intergenerational parallelism of disposition to deviance would be greater where G1 subjects (as young adults) indicated that they did not feel unwanted by their friends than where they indicated that they felt unwanted by their friends. As hypothesized, for G1 youths who do not feel unwanted by their friends, the degree of intergenerational parallelism in deviant attitudes (dispositions) was appreciably greater (b = 0.390) than for youths who assert that they do feel unwanted (b = −0.068, not significant). To have held antisocial attitudes (characterized as having had a deviant disposition)
101
Moderators of Intergenerational Parallelism Table 4.6 Intergenerational Parallelism of Disposition to Deviance by G1 Affirmation/Denial as Young Adults of Feeling Unwanted by Their Friends G1 Young Adults Do Not Feel Unwanted by Friends (N = 1475)
G1 Disposition to Deviance Antisocial Coping Disrespect Others Rejection of Norms G2 Disposition to Deviance Antisocial Coping Disrespect Others Rejection of Norms Intergenerational Parallelism Effect
b
B
1.000 1.261 1.387
0.458 0.577 0.635
0.000 8.701 7.915
1.000 1.053 1.246
0.622 0.655 0.776
0.000 6.271 5.646
1.000 0.396 0.650
0.668 0.468 0.577
0.000 11.861 11.770
1.000 0.498 0.664
0.741 0.545 0.560
0.000 5.160 5.032
0.390
0.263
5.056
−0.068
−0.058
−0.505
*
Significance
G1 Young Adults Feel Unwanted by Friends (N = 183) b
B
Significance
*
fixed parameter
while asserting that one does not feel rejected by one’s friends suggests membership in a deviant group or the presence of a supportive deviant subculture. In such a case, G1 intragenerational continuity in deviant dispositions is facilitated, as is adult G1 communication of such attitudes to G2 youths. Similar results should be obtained, and like interpretations would be warranted when using another indicator of comfort with adult peers. It was hypothesized that for G1 young adults who indicated that they do not worry about their relationships with their friends, intergenerational parallelism in disposition to deviance would be appreciably greater than for G2 young adults who indicated that they do worry about their relationships with their friends. The hypothesis was tested by estimating structural equations models specifying intergenerational continuity in disposition to deviance separately for G1 young adults who do and do not, respectively, worry about their relationships with their adult peers. The results of the analysis are summarized in Table 4.7. As hypothesized, intergenerational parallelism in disposition to deviance was appreciably (about five times) greater for G1 subjects who indicated as young adults that they did not worry about their relationships with friends (b = 0.398) than for G1 subjects who indicated as young adults that they did worry about their
102
Chapter 4
Table 4.7. Intergenerational Parallelism of Disposition to Deviance by G1 Affirmation/Denial as Young Adults of Worrying About Their Relationships with Friends G1 Young Adults Does Not Worry About Relationship (N = 1,228) b G1 disposition to deviance Antisocial coping Disrespect others Rejection of norms G2 disposition to deviance Antisocial coping Disrespect others Rejection of norms Intergenerational parallelism effect
B
a
G1 Young Adult Worries About Relationship (N = 430)
Significance
b
B
Significance
1.000 1.119 1.286
0.486 0.543 0.625
0.000 7.889 7.222
1.000 1.354 1.506
0.481 0.652 0.725
0.000 6.065 5.864
1.000 0.354 0.640
0.697 0.447 0.586
0.000 10.823 10.965
1.000 0.571 0.680
0.621 0.553 0.541
0.000 6.873 6.351
0.390
0.263
4.871
0.078
0.059
0.669
a
Fixed parameter.
relationships with their friends (b = 0.078, not significant). Presumably, for the former group, a level of psychological comfort with friends as young adults indicates a supportive environment for continuity of deviant dispositions displayed by the G1 adolescent youths and, thus, opportunities to transmit such attitudes to (or to provide problematic/distressful situations that evoke deviant attitudes by) G2 youth. For the latter group, worrying about one’s peer relationships suggests that deviant dispositions are problematic for the G1 subjects’ friendship groups, thus rendering these attitudes as amenable to change prior to G1 parenthood; and the absence of G1 intragenerational stability in deviant attitudes diminishes the likelihood of transmitting the deviant dispositions to G2 youths. The intragenerational stability of deviant attitudes should depend on the stability of one’s membership/reference groups. If one is comfortable with one’s friendship groups, even when holding deviant dispositions, the likelihood of intragenerational stability and cross-generational transmission of deviant behavior should be facilitated. In fact, we have seen that this is the case when focusing on G1 young adult relationships with their friends. Similarly, if the G1 youth who holds deviant attitudes as a youth remains comfortable with his socioeconomic station as a young adult, it
103
Moderators of Intergenerational Parallelism
might be assumed that such attitudes are compatible with those endorsed in the G1 adult’s social class (assuming intragenerational stability). Thus, the G1 adult might be more likely to display and model behaviors reflecting those attitudes for the G2 youth. However, if the person is uncomfortable with his or her social class, it might appear that the G1 adult has changed reference groups and correlated expectations of, or aspirations for, oneself—aspirations or expectations that are incompatible with youthful deviant dispositions; thus, the adult would be less likely to communicate deviant dispositions (or occasion distressful correlates of such attitudes that might evoke G2 deviant adaptations) to the G2 youths. Informed by these considerations, separate models were estimated that specified intergenerational parallelism in disposition to deviance for G1 subjects who as young adults denied and affirmed, respectively, that they felt unsure about themselves when thinking about their social class. It was expected that for those who did not feel unsure of themselves, intergenerational continuity would be appreciably greater than for those who were unsure of themselves. The results of the analysis are summarized in Table 4.8.
Table 4.8. Intergenerational Parallelism of Disposition to Deviance by G1 Affirmation/Denial as Young Adults of Feeling Unsure When Thinking About Their Social Class Do Not Feel Unsure About Themselves (N = 1,380) b G1 disposition to deviance Antisocial coping Disrespect others Rejection of norms G2 disposition to deviance Antisocial coping Disrespect others Rejection of norms Intergenerational parallelism effect a
Fixed parameter.
B
a
Feel Unsure About Themselves (N = 278)
Significance
b
B
Significance
1.000 1.257 1.490
0.451 0.567 0.672
0.000 8.610 7.548
1.000 0.973 1.126
0.593 0.578 0.668
0.000 4.808 4.854
1.000 0.369 0.606
0.698 0.469 0.564
0.000 11.441 11.721
1.000 0.596 0.957
0.562 0.484 0.641
0.000 5.035 4.747
0.400
0.256
4.924
0.006
0.006
0.063
104
Chapter 4
Consistent with expectations, intergenerational continuity in deviant dispositions between G1 and G2 youths was appreciably greater when the G1 youths as young adults indicated that they did not feel unsure of themselves when thinking about their social class (b = 0.400) than when the G1 youths as young adults indicated that they did feel unsure of themselves when thinking about their social class (b = 0.006, not significant). For the former group, comfort with one’s social class would suggest continuing compatibility of youthful antisocial dispositions with adult social class (again, assuming intragenerational stability in deviant dispositions and socioeconomic stratum); thus, the transgenerational communication of antisocial attitudes would be facilitated. For the latter group, signification by the G1 youths as adults that they were unsure of themselves when thinking about their social class suggests changing perceptions about the normative standards of one’s social class or changing aspirations about one’s desired social class. In either case, reevaluation of the acceptability of youthful antisocial disposition is likely to impede intergenerational transmission of deviant dispositions from the G1 adult to the G2 youth. First-generation youths whose life circumstances remain stable are more likely to manifest intragenration stability in attitudes (including disposition to deviance) as well and so are more likely to transmit these attitudes to G2 youths. Conversely, a change in life circumstances would be expected to disrupt the viability of stable attitudes whether over the life course or intergenerationally. A case in point is a life event such as being promoted at work. Youths with deviant dispositions who reported as adults that they were promoted at work would be obliged to meet new expectations, would be further immersed in a conventional network, and would be more subject to conventional social controls. Further, by virtue of the increased gratifications in their conventional roles, they would be less disposed to maintain their deviant attitudes: thus, they would be less likely to communicate the deviant attitudes of their youth to their own children. Youths with deviant attitudes who did not report such a life event would be more likely to manifest stability in other life circumstances, including those that continued to be supportive of preexisting deviant attitudes. Based on the assumptions, it was hypothesized that G1 subjects who reported as young adults that they were not promoted at work would manifest appreciably greater intergenerational parallelism in disposition to deviance than G1 subjects who reported as young adults that they were promoted at work.
105
Moderators of Intergenerational Parallelism
To test this hypothesis, structural equation models were estimated that specified intergenerational parallelism in disposition to deviance for G1 subjects who as young adults reported that they were not promoted at work and for G1 subjects who reported that they were promoted at work. The results of the analysis are summarized in Table 4.9. As hypothesized, persons who reported as young adults that they were not recently promoted at work manifested an appreciably greater degree of intergenerational continuity in adolescent deviant attitudes (b = 0.633) than individuals who reported that they had been recently promoted at work (b = 0.181). To some extent, these interpretations replicate those offered for feeling unsure when thinking about their social class. Being unsure suggests changing standards and expectations and, therefore, decreased perceptions of the acceptability of youthful antisocial dispositions. Decreased perceptions of the acceptability of antisocial dispositions on the part of the G1 adult decreases the likelihood of communication of these attitudes to the G2 youth. In addition, the possibility exists that not being promoted at work might reflect an occasion for exacerbation of alienation from the conventional world, assuming that the young adult anticipated being promoted and judged not being promoted as unjust.
Table 4.9. Intergenerational Parallelism of Disposition to Deviance by G1 Young Adult Affirmation/Denial of Recently Being Promoted at Work Not Promoted at Work (N = 313) b G1 disposition to deviance Antisocial coping Disrespect others Rejection of norms G2 disposition to deviance Antisocial coping Disrespect others Rejection of norms Intergenerational parallelism effect a
Fixed parameter.
B
a
Significance
Promoted at Work (N = 1,085) b
B
Significance
1.000 1.269 1.943
0.389 0.494 0.756
0.000 5.023 4.317
1.000 1.220 1.166
0.522 0.637 0.608
0.000 8.360 8.290
1.000 0.414 0.703
0.648 0.446 0.586
0.000 7.269 6.800
1.000 0.412 0.661
0.676 0.492 0.586
0.000 10.875 11.036
0.633
0.372
3.909
0.181
0.139
2.547
106
Chapter 4
Intergenerational Influences
Given intragenerational stability between early adolescence and adulthood, intergenerational parallelism of deviant dispositions depends on circumstances that facilitate or impede communications of the antisocial disposition from the G1 adult to the G2 youth. Among the more salient circumstances that moderate the degree of communication of antisocial dispositions from the G1 adult to the G2 youth is the nature of the relationship between the two. Presumably, the G2 youth would be more likely to identify with and adopt the parents’ antisocial attitudes if the youth had more, rather than less, favorable attitudes toward the parents. On the basis of this reasoning, it was hypothesized that the degree of intergenerational continuity of deviant dispositions would be greater for G2 youths who find it easy to discuss problems with their parents than for the youths who do not find it easy to discuss problems with parents. This would suggest that here the G2 dispositions are the product of social learning rather than the alternative outcome of alienated adaptations to parent-induced stress. In order to test the hypothesis, two structural equation models were estimated that specified intergenerational parallelism in disposition to deviance for G2 youths who did and did not, respectively, find it easy to discuss problems with their parents. The results of the analysis are summarized in Table 4.10. As predicted, the degree of intergenerational continuity in deviant dispositions was substantially greater when the G2 youths indicated close relations with their parents (b = 0.470) than when the G2 youths indicated more strained relations with their parents (b = 0.113, not significant). In sum, the results taken as a whole suggest that intragenerational continuity in deviant dispositions is contingent on three kinds of moderator: social support for antisocial dispositions (suggested by the friendship moderators reported by the G1 youths as adults); stability in life circumstances (reflected in not being promoted at work, and not feeling unsure of oneself when thinking about one’s social class reported by G1 subjects as young adults); and close relations with parents (reflected in G2 reports of finding it easy to discuss problems with parents). Under conditions where peer support for deviant attitudes appears to exist (suggesting a deviant subculture), intragenerational stability in life circumstances exists, and G1 parents and G2 youths have a good relationship (suggesting G1 social support as well as G2 influencibility), the degree of intergenerational
107
Moderators of Intergenerational Parallelism
Table 4.10. Intergenerational Parallelism of Disposition to Deviance by G2 Youth Affirmation/Denial That It Is Easy to Discuss Problems with Parents Not Easy to Discuss Problems with Parents (N = 876) b G1 disposition to deviance Antisocial coping Disrespect others Rejection of norms G2 Disposition to deviance Antisocial coping Disrespect others Rejection of norms Intergenerational parallelism effect
B
a
Significance
Easy to Discuss Problems with Parents (N = 782) b
B
Significance
1.000 1.242 1.343
0.484 0.601 0.650
0.000 7.104 6.310
1.000 1.298 1.466
0.452 0.587 0.662
0.000 6.704 6.427
1.000 0.333 0.686
0.596 0.427 0.555
0.000 8.875 9.320
1.000 0.467 0.710
0.680 0.468 0.593
0.000 8.790 8.189
0.113
0.094
1.485
0.470
0.312
4.204
a
Fixed parameter.
continuity in deviant dispositions tends to be appreciably stronger. Where peer support for deviant dispositions appears to be lacking, changes in life circumstances or reference groups occur, and parent-child relations are strained, intergenerational parallelism of deviant dispositions is diminished.
Negative Self-Feelings As noted earlier, within each generation, negative self-feelings influences the adoption of deviant patterns toward the goal of reducing these distressful feelings. The transgenerational continuity in negative selffeelings accounts, in part, for the transgenerational continuity in deviant attitudes and behavior. However, the intergenerational continuity of negative self-feelings is contingent on a number of circumstances. Some of these moderators refer to G1 intragenerational continuity of negative selffeelings between early adolescence and adulthood, whereas other contingencies refer to the influence of G1 parental negative self-feelings on G2 youthful negative self-feelings.
108
Chapter 4
Moderators of Intragenerational Continuity
It was expected that intergenerational parallelism in negative selffeelings would be greater under conditions in which the G1 youths reported higher levels of various deviant patterns. Presumably deviant patterns would evoke negative sanctions and other distressing consequences that would stabilize distress over the life course. Additionally, engaging in deviant patterns might reflect the absence of efficacious conventional coping mechanisms that might have assuaged or forestalled negative selffeelings. In either case, manifestations of negative self-feelings and the modeling of ineffective coping mechanisms by G1 subjects as parents would increase the probability that the G2 youths would manifest similar negative self-feelings. In accordance with this reasoning, structural equation models specifying intergenerational parallelism were estimated separately for G1 youths who scored above and below the median, respectively, on a measure of engaging in theft-related activity as youths (using the measure described in chapter 3). It was hypothesized that for G1 youths who scored above the median on the theft score in early adolescence, the degree of intergenerational parallelism in negative self-feelings would be appreciably greater than for G1 youths who scored below the median on the theft score in early adolescence. The results of the analysis are summarized in Table 4.11. Consistent with this reasoning, it was observed that for G1 youths who scored higher on a scale indicating involvement in theft as adolescents, the degree of intergenerational parallelism in negative self-feelings between G1 and G2 adolescents was appreciably greater (b = 0.537) than for G1 youths who scored lower on the measure of theft activity (b = 0.279). It was also hypothesized that engaging in violent acts as youths would moderate the degree of intergenerational continuity of negative selffeelings. As expected (see Table 4.12) for G1 subjects who scored higher on a measure of engaging in violent behavior as adolescents, the degree of intergenerational continuity in negative self-feelings was appreciably greater (b = 0.687) than for G1 subjects who scored lower on the measure (b = 0.266). Again, it is presumed that engaging in deviant behavior will be associated with distressing sequelae such as rejection by conventional others and deprivation of resources needed to satisfy one’s needs. It will be noted that the degree of intergenerational parallelism in psychological distress was still statistically significant for G1 adolescent
109
Moderators of Intergenerational Parallelism Table 4.11. Intergenerational Parallelism of Negative Self-feelings by G1 Involvement in Theft During Early Adolescence Low G1 Involvement in Theft (N = 959) b G1 negative selffeelings Anxiety Self-derogation Depression G2 negative selffeelings Anxiety Self-derogation Depression Intergenerational parallelism effect
B
a
Significance
High G1 Involvement in Theft (N = 699) b
B
Significance
1.000 1.112 1.352
0.564 0.627 0.762
0.000 13.666 12.245
1.000 1.155 1.527
0.515 0.595 0.787
0.000 10.265 9.208
1.000 1.029 0.770
0.602 0.794 0.785
0.000 16.857 15.861
1.000 1.135 0.905
0.572 0.762 0.775
0.000 11.996 13.033
0.279
0.159
3.611
0.537
0.306
5.110
a
Fixed parameter.
Table 4.12. Intergenerational Parallelism of Negative Self-feelings by G1 Involvement in Violence During Early Adolescence Low G1 Involvement in Violence (N = 1,203) b G1 negative selffeelings Anxiety Self-derogation Depression G2 negative selffeelings Anxiety Self-derogation Depression Intergenerational parallelism effect a
Fixed parameter.
B
a
High G1 Involvement in Violence (N = 455)
Significance
b
B
Significance
1.000 1.065 1.281
0.576 0.613 0.737
0.000 14.161 13.117
1.000 1.264 1.792
0.470 0.594 0.842
0.000 8.653 7.239
1.000 1.018 0.782
0.610 0.786 0.782
0.000 18.341 18.535
1.000 1.152 0.894
0.560 0.756 0.766
0.000 8.945 9.066
0.266
0.157
3.877
0.687
0.347
4.715
110
Chapter 4
subjects who were below the median in theft and interpersonal violence (although the intergenerational continuity was substantially less than for the mutually exclusive groups). Apparently, psychological distress remains stable across the G1 life course and continues to provide models for ineffective or maladaptive coping practices for the G2 youths. The concurrence of deviant behavior with negative self-feelings for G1 youths suggests that the deviant behavior does not so much reflect a subcultural endorsement of the behavior as it would appear to reflect maladaptive or ineffective coping responses that would likely facilitate G1 intragenerational continuity of psychological distress and transgenerational communications of ineffective/maladaptive coping patterns that would fail to forestall or assuage G2 experiences of negative self-feelings. Consistent with this reasoning, it was expected that alienation from conventional sources of social support would moderate the observation of intergenerational parallelism of negative self-feelings. Alienation, in the sense of emotional divorcement, from conventional others (parents, teachers, friends) deprives the person of important sources of social support. Negative self-feelings are more likely to continue in the absence of social support than in its presence. Social support provides opportunities to resolve stress-inducing problems and offers the potential for esteem-inducing responses. In the presence of social support, the experience of negative self-feelings is more likely to be discontinuous because the sources of distress might be obviated through the problem-solving/affective support mechanisms offered by significant others. Conversely, in the absence of support, the initial G1 experience of psychological distress (reflecting as it does the inability to forestall occasions for or assuage negative self-feelings) is likely to continue over the G1 life course, and the G2 youths are more likely to be deprived of effective coping patterns for dealing with distressful life circumstances by modeling G1 parental responses. In order to test this reasoning, the G1 population was divided into two groups according to whether the youths were above or below the median on a score indicating whether the youths thought it was very important what their parents, teachers, and the kids at school thought of them. A higher score indicated that the G1 youths tended not to evaluate these opinions as important and that, therefore, the distressed youth was alienated from what might otherwise have been important sources of social support. It was hypothesized that for youths scoring higher on the index, the observed degree of intergenerational parallelism in
111
Moderators of Intergenerational Parallelism
negative self-feelings would be substantially greater than for youths scoring lower on the index. A comparison of the structural equations models estimated for the two groups is summarized in Table 4.13. As hypothesized, intergenerational continuity in negative self-feelings was appreciably greater for G1 youths who indicated that the opinions others (parents, teachers, kids at school) hold of them were not important (b = 0.514) than for G1 youths who indicated that the opinions held of them by significant others were important (b = 0.293). For the latter group, presumably, the negative selffeelings experienced had the potential of being assuaged by recourse to sources of social support; thus, the experience of negative self-feelings was less likely to be continuous throughout the life course. Nor, in this instance, was communication of ineffective coping strategies as likely to be communicated to G2 youths in the absence of G1 distress continuing into parenthood. If it is assumed that reports of negative self-feelings by G1 youths signify the learning of ineffective or maladaptive coping patterns, it is to be expected that consistent socializing responses by parents would increase the likelihood of maintaining these ineffective responses throughout the life course. The absence of effective coping patterns, and the Table 4.13. Intergenerational Parallelism of Negative Self-feelings by G1 Alienation from Conventional Others During Early Adolescence Important What Parents, Teachers, Kids at School Think of G1 Youth (N = 934) b G1 negative selffeelings Anxiety Self-derogation Depression G2 negative selffeelings Anxiety Self-derogation Depression Intergenerational parallelism effect a
Fixed parameter.
B
a
Significance
Not Important What Parents, Teachers, Kids at School Think of G1 Youth (N = 724) b
B
Significance
1.000 1.129 1.450
0.560 0.633 0.813
0.000 14.553 12.578
1.000 0.847 1.091
0.651 0.552 0.711
0.000 6.765 6.670
1.000 1.059 0.848
0.592 0.779 0.797
0.000 15.431 14.792
1.000 0.950 0.782
0.584 0.669 0.815
0.000 6.611 7.860
0.293
0.172
3.772
0.514
0.308
6.216
112
Chapter 4
increased probability of experiences of psychological distress, then would be communicated by the G1 parents to the G2 youths. By this reasoning, it was expected that for G1 youths who indicated in early adolescence that their “parents pretty much agree about how they should be raised,” the degree of intergenerational parallelism in negative self-feelings would be substantially greater than for G1 youths who denied that their parents agreed on how they should be observed. The hypothesis was supported. Reference to Table 4.14 will indicate that, for the former group, the observed degree of intergenerational continuity in negative self-feelings was appreciably greater (b = 0.464) than for the G1 youths who denied that their parents reached consensus on how they should be raised (b = 0.029, not significant). For this latter group, the lack of consensus decreases the probability of continuity in maladaptive or ineffective mechanisms over the life course and, therefore, their communication to the next generation. Individuals who are disposed to experience negative self-feelings as youths apparently do not posses the kinds of effective coping mechanism that are capable of forestalling or assuaging distressful experiences. Therefore, in the presence of such experiences later in life, they would be likely, once more, to experience negative self-feelings, assuming that they Table 4.14. Intergenerational Parallelism of Negative Self-feelings by G1 Youth Reports of Parental Agreement on Child Rearing G1 Youth Asserts That Parents Agree on How the Youth Should Be Raised (N = 1,424) b G1 negative selffeelings Anxiety Self-derogation Depression G2 negative selffeelings Anxiety Self-derogation Depression Intergenerational parallelism effect a
Fixed parameter.
G1 Youth Asserts That Parents Do Not Agree on How Youth Should Be Raised (N = 234)
B
Significance
b
1.000 1.155 1.464
0.538 0.621 0.787
0.000 15.583 14.013
1.000 1.155 1.527
0.515 0.595 0.787
0.000 10.265 9.208
1.000 1.094 0.855
0.584 0.792 0.789
0.000 18.820 18.382
1.000 1.135 0.905
0.572 0.762 0.775
0.000 11.996 13.033
0.464
0.263
6.853
0.029
0.021
0.256
a
B
Significance
113
Moderators of Intergenerational Parallelism
have not developed efficacious coping mechanisms in the interim; as parents, they would be likely to model and communicate ineffective or maladaptive coping mechanisms to their G2 children. By this reasoning, it is to be expected that intergenerational continuity in negative self-feelings (via G1 intragenerational continuity) would be appreciably greater under conditions whereby the G1 subjects as adults experienced life events that (1) required adaptation to new normative requirements that might challenge essentially inadequate coping patterns (suggested by preexisting psychological distress), (2) disrupted existing coping patterns, or (3) were intrinsically distressful. Accordingly, it was hypothesized that for G1 youths who (as young adults) indicated that they “got married or started living with someone of the opposite sex” the degree of intergenerational continuity in negative self-feelings observed would be appreciably greater than for G1 youths who indicated (as young adults) that they did not enter into a marital or cohabitating relationship. The hypothesis was tested by comparing structural equations models specifying intergenerational parallelism of negative self-feelings for G1 subjects who did and did not report as young adults that they had entered into a marital or cohabitating relationship. The results of the analysis are summarized in Table 4.15. As hypothesized, the degree of intergeneraTable 4.15. Intergenerational Parallelism of Negative Self-feelings by G1 Reports as Young Adults Entering into a Marital or Cohabitating Relationship G1 Young Adult Did Not Enter into Marital or Cohabitating Relationship (N = 270) b G1 negative selffeelings Anxiety Self-derogation Depression G2 negative selffeelings Anxiety Self-Derogation Depression Intergenerational parallelism effect a
Fixed parameter.
B
a
Significance
G1 Young Adult Did Enter into Marital or Cohabitating Relationship (N = 1,388) b
B
Significance
1.000 1.252 1.364
0.549 0.687 0.748
0.000 7.216 6.945
1.000 1.083 1.378
0.565 0.612 0.778
0.000 16.345 14.757
1.000 0.809 0.565
0.630 0.721 0.736
0.000 8.481 8.642
1.000 1.131 0.883
0.579 0.787 0.787
0.000 18.259 18.381
0.082
0.043
0.510
0.405
0.247
6.371
114
Chapter 4
tional continuity in negative self-feelings observed between G1 and G2 was substantially larger (b = 0.405) for the group indicating entering into marital or cohabitating relations as young adults than for the mutually exclusive group (b = 0.082, not significant). This finding is compatible with the conclusion that for G1 youths who initially experienced negative self-feelings, entering into marital or cohabitating relations (particularly in early adolescence) provides a stress-invoking experience for those who apparently early on had not developed effective coping mechanisms for dealing with life stress. By similar reasoning, it was hypothesized that for G1 youths who indicated as young adults that someone important to them died, the degree of intergenerational parallelism in negative self-feelings would be substantially greater than for the mutually exclusive group. The loss of someone important is an intrinsically noxious event and, by definition, reflects the loss of a salient source of socioemotional support that might ordinarily assuage the distress associated with adverse life experiences. Further, to the extent that the G1 subject was functionally interdependent with the object of bereavement, he or she might be required to take on the role obligations of the deceased—a circumstance that is also potentially stress-inducing. The combination of youthful (G1) inability to forestall or assuage distress (evidenced by G1 negative self-feelings) and the presumed continuity of this into adulthood, and the challenges faced by young adults, via distressful life circumstances increase the probability that G2 ineffective coping with life stress will be learned by the G2 children and occasions for experiencing negative self-feelings will be provided to the G2 children. As in the case of entering into marital or cohabitating relationships, the hypotheses were tested by estimating separate structural equations models specifying intergenerational parallelism in negative self-feelings for G1 subjects who did and did not, respectively, report as young adults having had someone important to them die. The results of the analysis are summarized in Table 4.16. Consistent with these expectations, the magnitude of observed intergenerational parallelism in negative self-feelings was observed to be substantially greater for G1 youths who, as young adults, reported the recent loss of an important person in their life (b = 0.731) than for the mutually exclusive category of G1 youths (b = 0.274).
115
Moderators of Intergenerational Parallelism
Table 4.16. Intergenerational Parallelism of Negative Self-feelings by G1 Reports as Young Adults That Someone Important to Them Died G1 Young Adult Does Not Report That Someone Important to Them Died (N = 1,249) b G1 negative selffeelings Anxiety Self-derogation Depression G2 negative selffeelings Anxiety Self-derogation Depression Intergenerational parallelism effect a
B
a
G1 Young Adult Does Report That Someone Important to Them Died (N = 409)
Significance
b
B
Significance
1.000 1.167 1.394
0.552 0.645 0.770
0.000 15.241 13.514
1.000 0.956 1.336
0.595 0.568 0.795
0.000 9.632 9.578
1.000 1.167 0.890
0.559 0.791 0.785
0.000 16.846 17.177
1.000 0.885 0.724
0.654 0.751 0.795
0.000 113.574 10.540
0.274
0.170
4.263
0.731
0.396
5.335
Fixed parameter.
Moderators of Transgenerational Continuity
Intergenerational parallelism in negative self-feelings depends, in part, on G1 continuity of youthful psychological distress into parenthood, at which time the ineffective coping mechanisms that failed to forestall or assuage psychological distress are modeled for, and communicated to, the children of the next generation. The moderators just considered relate primarily to the facilitation of intragenerational continuity of G1 negative self-feelings. Other moderators influence the extent to which G1 parental psychological distress influences G2 distress. The transgenerational continuity of negative self-feelings depends, in part, on the G2 youth’s experience of rejection and failure. If the G2 youths have such experiences, in the absence of socialization into the use of effective coping mechanisms, they are likely to experience higher levels of psychological distress, as did their parents. As an index of social rejection, the G2 respondents were asked to respond to a three-item index indicating whether they were liked by peers. As an indicator of failure, the G2 youths responded to three items signifying whether they, for example, doubted if they would get ahead in life as far as they would really like.
116
Chapter 4
Regarding G2 experiences of rejection as a moderator of intergenerational continuity of negative self-feelings, it was hypothesized that the magnitude of intergenerational continuity would be greater where the G2 youths reported higher levels (above the median) of peer rejection during early adolescence than where the G2 youths reported lower levels (below the median) of peer rejection on a three-item index based on responses to the following: “Do you have a lot of friends?” “Do you have a close friend right now” “Are you liked by kids of the opposite sex?”. The hypothesis was tested by estimating separate structural equation models specifying intergenerational parallelism of negative self-feelings for G2 youths who were above and below, respectively, the median score on the measure of peer rejection. The results of the analysis are summarized in Table 4.17. As expected, under conditions whereby G2 youths perceived themselves as not liked by peers, the degree of intergenerational continuity of negative self-feelings observed was far greater (b = 0.522) than that observed for G2 youths who affirmed that they were liked by peers (b = 0.280). Where G2 youths are rejected by peers, the presumed inadequate coping patterns (suggested by G1 negative self-feelings) transmitted by the G1 parents would preclude G2 youths forestalling or assuaging the distressful self-feelings that ordinarily accompany peer rejection.
Table 4.17. Intergenerational Parallelism of Negative Self-feelings by G2 Reports in Early Adolescence of Peer Rejection Low G2 Peer Rejection (N = 1,089) b G1 negative selffeelings Anxiety Self-derogation Depression G2 negative selffeelings Anxiety Self-derogation Depression Intergenerational parallelism effect a
Fixed parameter.
B
a
Significance
High G2 Peer Rejection (N = 569) b
B
Significance
1.000 1.093 1.342
0.569 0.621 0.763
0.000 14.432 12.931
1.000 1.128 1.451
0.556 0.627 0.807
0.000 10.801 10.189
1.000 1.012 0.771
0.583 0.763 0.752
0.000 16.199 16.464
1.000 1.180 0.937
0.581 0.796 0.823
0.000 11.896 11.592
0.280
0.168
3.844
0.522
0.314
5.312
117
Moderators of Intergenerational Parallelism
Hence, G2 youths under such conditions would be prone to parallel the experiences of negative self-feelings manifested by the G1 youths. In the absence of such provocative circumstances and through the use of avoidant (and other) defense mechanisms, it might be possible for G2 youths to forestall repetition of the G1 youths’ higher levels of psychological distress. Regarding G2 self-perceptions of failure as a moderator of intergenerational continuity of negative self-feelings, it was expected that for G2 youths who anticipated failure in life, the likelihood of replicating G1 negative self-feelings would be greater than for G2 youths who did not anticipate failure. A comparison of structural equation models specifying intergenerational parallelism of negative self-feelings for G2 who scored variously high and low on a measure of expectations of failure is summarized in Table 4.18. As hypothesized, intergenerational parallelism in negative self-feelings was moderated by an index of anticipated failure on the part of G2 youths. For G2 youths who anticipated failure in life (above median scores on a three-item index: I doubt if I will get ahead in life as far as I would really like. There isn’t much chance that a kid from my neighborhood will get ahead; I would do a lot better in life if society didn’t have the cards stacked against me), the observed degree of intergenerational
Table 4.18. Intergenerational Parallelism of Negative Self-feelings by G2 Youth Expectations of Failure Low Expectation of Failure (N = 860) b G1 negative selffeelings Anxiety Self-derogation Depression G2 negative selffeelings Anxiety Self-derogation Depression Intergenerational parallelism effect a
Fixed parameter.
High Expectation of Failure (N = 789)
B
Significance
b
1.000 1.133 1.298
0.585 0.663 0.759
0.000 13.708 13.011
1.000 1.113 1.526
0.523 0.583 0.799
0.000 11.474 9.908
1.000 0.969 0.760
0.503 0.691 0.691
0.000 9.731 10.902
1.000 1.246 0.924
0.558 0.796 0.786
0.000 13.499 13.380
0.191
0.156
2.882
0.447
0.249
4.839
a
B
Significance
118
Chapter 4
parallelism in negative self-feelings was appreciably greater (b = 0.447) than for G2 youths who scored below the median on the index (b = 0.191). Of course, it should be noted that any index of perceived (anticipated) failure at the same time reflects the perceived inability to forestall these circumstances through personal effort. In any case, where G2 youths anticipate the inability to forestall failure, they are more likely to parallel the subjective distress (ordinarily associated with anticipated failure) manifested by G1 youths. Experiences of social rejection and failure by G2 youths are likely to exacerbate and, so, replicate the negative self-feelings experienced by their parents at a comparable developmental stage. However, the experience of negative self-feelings in the face of failure and rejection is most likely to occur in the absence of social support that might have assuaged these predictable distressful sequelae. Among the more salient potential sources of social support available to the child are the parents. Under benign circumstances, parents would be available to help the child either forestall the occasions for negative selffeelings or to soften the impact of such distressful feelings. Thus, supportive parents would lessen the replication of any negative self-feelings experienced by the parents themselves as youths. However, in the absence of parental support, the G2 youth would be more likely to parallel the negative self-feelings experienced by their parent(s) at comparable developmental stages. This is the case not only because the child is deprived of coping resources but also because the perception by the child of parents’ failure to provide instrumental and socioemotional support is intrinsically distressing insofar as it reflects badly on the G2 youths’ sense of worth. The absence of parental support is reflected in a number or ways: absence of supervision, the child not being disposed to display affection or feel close to the parents, and not receiving positive reinforcement for desirable behaviors—to name a few. For each of the indicators of absence of social support, it would be expected that G2 youths who were higher on the index would be more likely to recapitulate the negative selffeelings of the parents at the same developmental stage than G2 youths who were lower on the index. These expectations were tested by comparing the structural equations models specifying intergenerational continuity of negative self-feelings for G2 youths who were high and low, respectively, on each index of G2 lack of parental support. One indicator of lack of parental support is neglecting—that is, not caring enough about the child to exercise supervision, as is reflected in a
119
Moderators of Intergenerational Parallelism
four-item scale indicating allowing the child to leave the house without telling the parent, not knowing where the child is and whom they are with, and not knowing the child’s best friends and the parents of the child’s best friends. It was hypothesized that the magnitude of intergenerational parallelism in negative self-feelings would be appreciably higher where the G1 parent manifested scores above the median on the index of absence of parental supervision than where the G1 parent manifested scores below the median on absence of parental supervision. The results of the analysis are summarized in Table 4.19. As hypothesized, the degree of intergenerational parallelism in negative self-feelings was appreciably greater where the G2 youth perceived the parents exercising low parental supervision (b = 0.509) than where the G2 youth perceived the parent as exercising more supervision (b = 0.268). In the absence of parental support, the G2 child is less likely to have the resources (psychological or otherwise) to forestall the experience of stressful life circumstances or to assuage the concomitant distressful self-feelings. Thus, the G2 youth is more likely to replicate the experiences of negative selffeelings reported by the G1 parents as adolescents. This is in contrast to the situation where the G2 youths have more access to supportive parents
Table 4.19. Intergenerational Parallelism of Negative Self-feelings by G2 Reports of Parental Supervision G2 Youth Reports High Parental Supervision (N = 1,019) b G1 negative selffeelings Anxiety Self-derogation Depression G2 negative selffeelings Anxiety Self-derogation Depression Intergenerational parallelism effect a
Fixed parameter.
B
a
Significance
G2 Youth Reports Low Parental Supervision (N = 639) b
B
Significance
1.000 1.058 1.334
0.567 0.600 0.757
0.000 13.568 11.981
1.000 1.191 1.489
0.547 0.651 0.815
0.000 11.442 11.167
1.000 1.077 0.760
0.596 0.828 0.762
0.000 17.387 17.071
1.000 1.123 1.004
0.550 0.705 0.822
0.000 10.909 11.232
0.268
0.157
3.629
0.509
0.322
5.286
120
Chapter 4
who might help forestall the experience of distress (thus altering the degree of intergenerational parallelism of negative self-feelings). The attenuation of positive affective ties between parent and child also reflects a relative paucity of social support. Thus, it was hypothesized that for G2 youths who deny that they openly show affection to their parents, the degree of intergenerational parallelism in psychological distress would be greater than for G2 youths who affirm that they openly show affection to their parents. (See Table 4.20.) As hypothesized, for the G2 youths who denied that they openly show affection to their parents, the magnitude of intergenerational parallelism in psychological distress was appreciably greater (b = 0.532) than for the G2 youths who affirmed that they show affection to their parents (b = 0.302). For the latter group, the probability was somewhat greater that the affectional ties with parents would be somewhat instrumental in allowing the G2 youths to cope emotionally with distressful life circumstances; thus, the G2 youths would be less likely to replicate the experience of negative self-feelings by the G1 youth at the same developmental stage. Not surprisingly, in view of the preceding findings, examination of Table 4.21. supports the expectation that a report of (not) feeling close to one’s parents moderates the magnitude of intergenerational parallelism in negative self-feelings. For G2 youths who deemed that they do not feel Table 4.20. Intergenerational Parallelism of Negative Self-feelings by G2 Reports of Showing Affection to Parents G2 Youth Shows Affection to Parents (N = 1,229) b G1 negative selffeelings Anxiety Self-derogation Depression G2 negative selffeelings Anxiety Self-derogation Depression Intergenerational parallelism effect a
Fixed parameter.
B
a
Significance
G2 Youth Does Not Show Affection to Parents (N = 429) b
B
Significance
1.000 1.151 1.495
0.541 0.623 0.809
0.000 15.901 14.076
1.000 1.000 1.161
0.609 0.609 0.707
0.000 8.275 8.064
1.000 1.080 0.780
0.564 0.790 0.767
0.000 16.377 18.028
1.000 1.028 0.916
0.646 0.736 0.804
0.000 10.893 9.557
0.302
0.178
4.426
0.532
0.322
4.281
121
Moderators of Intergenerational Parallelism
Table 4.21. Intergenerational Parallelism of Negative Self-feelings by G2 Youth Assertions of Feeling Close to Parents G2 Youth Feels Close to Parents (N = 1,462) b G1 negative selffeelings Anxiety Self-derogation Depression G2 negative selffeelings Anxiety Self-derogation Depression Intergenerational parallelism effect a
G2 Youth Does Not Feel Close to Parents (N = 196)
B
Significance
1.000 1.121 1.429
0.553 0.620 0.791
0.000 16.986 15.295
1.000 0.988 1.126
0.611 0.604 0.688
0.000 5.579 5.834
1.000 0.922 0.742
0.577 0.744 0.754
0.000 17.211 18.372
1.000 1.330 1.088
0.583 0.813 0.797
0.000 6.325 6.036
0.300
0.182
4.898
0.536
0.329
3.084
a
b
B
Significance
Fixed parameter.
close to their parents, the observed degree of intergenerational parallelism in psychological distress was substantially greater (b = 0.536) than for G2 youths who affirmed that they feel close to their parents (b = 0.300). Again, emotional identification with parents suggests the presence of resources that might facilitate effective coping with negative self-feelings, and the presence of resources that forestall or assuage distress lessens the likelihood of the G2 youth replicating the experience of negative selffeelings reported by the G1 youth. The absence of social support is reflected also in the inability to evoke rewarding responses from parents when conforming to their expectations. The ability to evoke rewarding responses from parents is reflected in responses to the question addressed to G2 youths: “When you have been especially good, how often (often, sometimes, hardly ever or never) does your parent praise you...kiss or hug you,...let you have something special to eat,...buy you something special,...let you have extra privileges?” A high score indicates that rewards are hardly ever or never forthcoming. It was hypothesized that for G2 youths who hardly ever or never receive rewards (above the median scores), the degree of intergenerational parallelism in negative self-feelings would be greater than for G2 youths who more often receive rewards (below the median scores). The results
122
Chapter 4
summarized in Table 4.22 are congruent with the hypothesis. For G2 youths who are less likely to receive parental rewards for good behavior, the degree of intergenerational parallelism observed was greater (b = 0.458) than for G2 youths who were more likely to receive parental rewards (b = 0.279). Presumably, the failure to evoke rewards for conventional behavior reflects the attenuation of ties with a potentially valuable tension-reducing resource. At the same time, the lack of rewards fails to reinforce the behaviors that are likely to evoke rewarding/tensionreducing responses from others in the G2 youth’s environment. Hence, the G2 youths who are not the objects of appropriate rewards are more likely to experience the kinds of distress that parallel those of the G1 youths at comparable developmental stages. However, youths who maintain ties with sources of support who reinforce adaptive patterns are less likely to recapitulate negative self-feelings experienced by the G1 youths. The extent to which the G2 youth experiences negative self-feelings and so repeats the experiences of negative self-feelings reported by G1 youths depends, in part, on the degree to which the G2 youth learns from his parents effective and conventional, as opposed to ineffective and maladaptive, coping mechanisms for dealing with distressful feelings.
Table 4.22. Intergenerational Parallelism of Negative Self-feelings by G2 Youth Assertions of Receiving Parental Rewards for Good Behavior G2 Youth Does Receive Rewards for Good Behavior (N = 851) b G1 negative selffeelings Anxiety Self-derogation Depression G2 negative selffeelings Anxiety Self-Derogation Depression Intergenerational parallelism effect a
Fixed parameter.
B
a
G2 Youth Does Not Receive Rewards for Good Behavior (N = 807)
Significance
b
B
Significance
1.000 1.111 1.391
0.558 0.620 0.776
0.000 13.125 11.563
1.000 1.109 1.378
0.565 0.626 0.778
0.000 12.291 11.538
1.000 0.948 0.714
0.603 0.770 0.765
0.000 14.859 14.615
1.000 1.196 0.949
0.570 0.784 0.803
0.000 13.675 13.803
0.279
0.159
3.320
0.458
0.282
5.414
123
Moderators of Intergenerational Parallelism
Presumably, the exposure to maladaptive coping mechanisms would simultaneously preclude forestalling occasions for, and assuaging, negative self-feelings and indeed would constitute another source of stress. A case in point is the G2 observation of regular and prolonged parent use of alcohol or marijuana over a long period of time. It was hypothesized that the degree of intergenerational parallelism in negative self-feelings observed would be substantially greater for G2 youths who observed regular and long-term use of alcohol or marijuana by the parent than for G2 youths who did not indicate such patterns. The results summarized in Table 4.23 support the hypothesis. The magnitude of intergenerational parallelism was substantially greater for the G2 youths who reported regular and prolonged parental substance use (b = 0.671) than for G2 youths who did not report this pattern (b = 0.309). This result was to be expected for several reasons. First, as suggested earlier, these patterns over the long term are ineffective in reducing negative self-feelings. Second, the learning of maladaptive coping mechanisms often precludes the learning of conventional and efficacious patterns. Third, the parental substance use patterns have concomitants that represent sources of negative self-feelings for the G2 youths, including abusive and neglectful parenting patterns, stigmatization of the G2 youth as the Table 4.23. Intergenerational Parallelism of Negative Self-feelings by G2 Youth Assertions of Regular and Prolonged Use of Alcohol or Marijuana by Parent G2 Youth Does Not Report Regular and Prolonged Parental Use (N = 1,446) b G1 negative selffeelings Anxiety Self-derogation Depression G2 negative selffeelings Anxiety Self-derogation Depression Intergenerational parallelism effect a
Fixed parameter.
B
a
Significance
G2 Youth Does Report Regular and Prolonged Parental Use (N = 212) b
B
Significance
1.000 1.133 1.390
0.554 0.627 0.769
0.000 15.916 14.804
1.000 1.011 1.339
0.600 0.606 0.803
0.000 7.965 6.898
1.000 1.112 0.833
0.571 0.791 0.778
0.000 19.112 19.345
1.000 1.035 0.832
0.635 0.758 0.759
0.000 6.691 6.982
0.309
0.186
5.039
0.671
0.406
3.682
124
Chapter 4
child of substance-abusing parents, and the necessity of the child frequently prematurely taking on a caretaker role and the burden of performing duties that the parent is no longer (if he or she ever was) willing and able to perform. Thus, under circumstances of parental substance use, the likelihood is increased that the G2 youth would manifest levels of psychological distress that parallel those reported by the G1 youth at a comparable developmental stage. In sum, the degree of intergenerational parallelism in negative self-feelings that is observed is contingent on circumstances that facilitate the continuity of such feelings between youth and parenthood for the first generation and that facilitate the influence of the first-generation on second-generation experiences that either reflect on or have implications for the level of negative self-feelings. The continuity of negative self-feelings between early adolescence and parenthood for the G1 subjects is facilitated by the early performance of stress-inducing behavior and its concomitants (deviant behaviors and consequent negative social sanctions), alienation from potential sources of social support that might forestall stress or assuage distress (not important what parents, teachers, kids at school think of me), consistent parental socialization in inadequate coping patterns (the combination of higher levels of psychological distress and parental agreement on patterns of child rearing), and the experience of late life events that challenge inadequate coping patterns (the combination of higher levels of distress with life events as young adults such as marriage/cohabitation and the death of a significant other). The intergenerational transfer of psychological distress from G1 parents is conditional on the G2 youth’s experiences of rejection (by peers) and failure (no expectations of upward socioeconomic mobility), absence of supportive parents who might have helped forestall stress and assuage distressful feelings (G2 youth does not show affection or feel close to parents and does not receive positive reinforcement for conventional coping), and G2 perceptions of G1 parental use of ineffective and maladaptive coping mechanisms (substance use).
Part III DECOMPOSING INTERGENERATIONAL PARALLELISM
In Part III, the relevant literature on the decomposition of intergenerational (dis)continuities in deviant behavior and its correlates are reviewed. Informed by the guiding general theory of deviant behavior and the literature, models are estimated that, in part or totally, statistically decompose the intergenerational continuities in deviant behaviors and its selected salient correlates observed in a multigenerational longitudinal study. Intergenerational continuities are accounted for in two ways: First, firstgeneration deviance (or related phenomena) is assumed to have consequences that, in turn, influence second-generation deviance; that is, part of the explanation for observed intergenerational continuities is in terms of intervening causal processes. The specification of the intervening (indirect) effect attenuates the direct association observed between the firstgeneration phenomenon and the second-generation phenomenon. Second, intergenerational continuity in deviance (or its selected salient correlates) are explained (partly or in toto) in terms of common antecedents of deviance (or the common correlates of deviance) in each generation. Specification of a particular cause of deviance in each generation and specification of intergenerational continuity for the hypothetical cause, at least in part, explain (statistically decompose) the originally observed intergenerational parallelism in deviance. The intergenerational continuity in the putative cause of deviance is also explainable either in terms of causal consequences of the first-generation cause on the secondgeneration cause (intervening processes) or in terms of intergenerational stability of intragenerational antecedents. 125
126
Part III
In chapter 5, we review the literature on intervening processes in observed intergenerational continuities in deviant behavior and related phenomena. In the context of this literature and guided by a general theory of deviant behavior, models are estimated that specify intervening processes in intergenerational parallelism of deviance or in selected hypothetical causes of deviant behavior (disposition to deviate and negative selffeelings). In chapter 6, models are estimated that specify common causal antecedents and intergenerational continuity of these causes of deviant behavior, disposition to deviance, and negative self-feelings. As in the case of intervening processes, the models are specified and discussed in the context of the relevant literatures and the guiding theoretical framework.
5 Intervening Processes
In chapter 4, it was argued that the degree of intergenerational (dis)continuity that is observed (whether in deviance, correlates of deviance, or unrelated phenomena) is contingent on a number of variables. Under some circumstances, intergenerational parallelism is observed to exist and to be relatively strong, whereas under other circumstances, the degree of intergenerational continuity is not statistically significant or is relatively weak. In any case, where a statistically significant intergenerational parallelism is observed, whether for the population of interest at large or for a segment of the population that is defined by some moderating value, the question remains as to how to explain the significant intergenerational parallelism. One kind of explanation is in terms of mediating variables; that is, parallelism might occur because the first generation sets in motion a chain of reactions according to which intervening processes mediate the causal relationship between the first- and second-generation phenomena. If these intervening processes are modeled, the originally observed intergenerational relationship is weakened or reduced to nonsignificance, suggesting that the putative mediating variables indeed offer part or all of the explanation of the observed intergenerational parallelism. The effects of intervening processes on decomposing instances of intergenerational parallelism have been hypothesized and observed widely, and certain of these intervening circumstances have been modeled successfully in the present study. Following a review of the literature on intervening processes, exemplars of models that specify mediating variables are described and discussed in the context of this literature and the guiding theoretical framework. 127
128
Chapter 5
Literature on Intervening Processes Numerous instances of the decomposition of intergenerational parallelism have been presented in the literature. In some cases, the models have reflected what are interpretable as intergenerational parallelism in correlates of deviant behavior. In other instances, the models reflect intervening processes in intergenerational parallelisms in deviant behavior itself. The literature for each are considered in turn.
Correlates of Deviant Behavior In some instances it is difficult to determine whether certain putative causes or correlates of deviant behavior should be treated as deviant behaviors themselves. Should abusive parenting practices, for example, be treated as causes of other forms of deviant behavior within each generation, or should abusive parenting be treated as manifestations of deviant responses/impulses in their own right? In any case, for present purposes, we will treat such variables as correlates rather than reflections of deviant behavior. A broad range of intergenerational patterns have been explained in terms of intervening variables. For example, the transgenerational continuity in school-related failure and rejection was decomposed, in part, by a number of intervening variables, including “family structure, mothers’ educational attainment, and level of mothers’ involvement in their children’s school activities and interest in their progress at school” (Kaplan, Kaplan, & Liu, 2000, p. 235). Similarly, a number of factors are implicated in the continuity or discontinuity of the intergenerational cycle of poverty (Harper, Marcus, & Moore, 2003): A multitude of causal factors—economic, political, environmental and social—are involved in life course and intergenerational poverty transmission. Negative impacts of, for example, indebtedness, unemployment, conflict, ecological stress, or cultural norms, to name a few, can result in harm in childhood that impacts over a life course or between generations. Some critical examples include poor nutrition and chronic ill-health, low educational achievement, psychological harm and low aspirations. What is pertinent is how and whether the real and felt negative effects can be overcome over a life course and/or between generations and, if not, what it is that prevents positive outcomes. There are clearly potential opportunities to disrupt negative poverty cycles, opportunities which themselves are mediated
Intervening Processes
129
through the same economic, social, political and environmental drivers that have the potential to harm (Moore 2001). (p. 536)
Among the more frequently considered intergenerational patterns that are considered with regard to their intervening patterns are early child bearing, divorce, and (particularly abusive) parenting patterns. Early Childbearing
Alternative explanations based on causally related intervening mechanisms, as opposed to intergenerationally continuous intragenerational causal mechanisms, have been discussed with regard to, for example, intergenerational transmission of age at first birth (Barber, 2001b): Bengtson (1975) suggests that parents’ and children’s attitudes, values, and preferences may be similar because of their shared social positions, background, and experiences. Similarly, parents’ and children’s childbearing behavior may be similar because of their shared social positions, background, and experiences. This is consistent with the idea that the socialization of children is only spuriously related to their mother’s age at first birth; children may behave like their mothers simply because their mothers’ and their own opportunities and constraints were shaped by the same social forces... An important alternative hypothesis is that the first-generation mother’s childbearing behavior itself produces the intergenerational similarity in firstbirth timing. This hypothesis also draws on the socialization perspective. One plausible explanation is that a direct effect of mother’s age at first birth on her child’s age at first birth is through mother’s attitude change. Mothers who experience early first births are likely to form a more positive attitude toward early childbearing as a result, and they transmit those attitudes to their children through socialization processes. These attitudes may lead to earlier childbearing (Barber, 2001)[a]. Even if mothers do not transmit those values to their children, children respond directly to their mother’s attitudes toward childbearing behavior, regardless of their own attitudes (Barber 2000). Another plausible explanation is that children model their own behavior directly on their mother’s behavior. (p. 222)
Others have suggested attitudinal transmission or modeling as mediating variables. Intergenerational continuity in teenage childbearing (Manlove, 1997) was mediated in part by preferences for age of childbearing. Teen mothers tended to produce daughters who reported earlier ideal ages for childbearing, which, in turn, was associated with increased likelihood of teenage childbearing by the daughter. Hardy and her associates (1998) suggested that
130
Chapter 5 ...normative expectations of timing of childbearing, neighborhood effects, and parenting behavior during the childhood or adolescent years (i.e., coercive parenting cycles occurring during the elementary school years and low supervision and monitoring of behavior and peers during the high school years) are potential pathways for such intergenerational continuity (Capaldi, Crosby, & Stoolmiller, 1996; Patterson, De Baryshe, & Ramsey, 1989; Serbin, Moskowitz, Schwartzman, & Ledingham, 1991). (p. 1229)
Others suggested that intergenerational parallelism in teenage motherhood is accounted for, in part, by the intervening influence of the daughter’s educational performance. First-generation teen mothers had daughters who achieved less well and were less likely to attend schools that were academically oriented. These variables, in turn, were associated with increased risk of teenage births (Manlove, 1997). Divorce
Intergenerational continuity in divorce is explainable, in part, by intervening variables relating to the consequences of the parental divorce for the child that might persist into adulthood. These include the experience of emotional distress and impaired ability to participate in interpersonal relationships (Amato, 1996; Amato & Booth, 1997; Amato & Rogers, 1997; Cherlin, Chase-Lansdale, & McRae, 1998; Ross & Mirowsky, 1999), educational attainment (Sandefur & Wells, 1999), and time of marriage (Bumpass et al., 1991; South, 1995). Thus, intergenerational transmission of divorce migth be mediated by adverse psychological outcomes on children that persist over the life course (Amato & Keith, 1991; Cherlin et al., & McRae, 1998). As Wolfinger (1999) observed: Perhaps as a result, the adult children of divorce more often report low levels of marital satisfaction than do people from intact families (Amato & Booth 1991; Glenn & Kramer, 1985). Recent evidence suggests that impaired interpersonal skills play a strong role in explaining the intergenerational transmission of divorce (Amato 1996). (p. 415)
Regarding the mediating effect of interpersonal behavior (Amato, 1996): Offspring whose parents divorced, compared with those whose parents remained continuously married, are more likely to have an interpersonal style marked by problematic behavior (problems with anger, jealousy, hurt feelings, communication, infidelity, and so on), and these interpersonal problems, in turn, increase the risk of divorce. Furthermore, lifecourse and socioeconomic variables mediate little of the association between parental divorce and
Intervening Processes
131
interpersonal behavior problems. These findings suggest that parental divorce has direct impact on these interpersonal behaviors. These findings are consistent with the notion that adult children from divorced families are exposed to poor models of dyadic behavior and may not learn the skills and attitudes that facilitate successful functioning within marital roles. Similarly, children of divorce may be predisposed to develop traits (such as lack of trust or an inability to commit) that exacerbate relationship tension. (p. 638)
The mediating role of age at marriage and cohabitation in the intergenerational transmission of divorce might be for any of a number of reasons (Amato, 1996). With regard to age at marriage, parental divorces might motivate early marriage so that the children can escape from an unhappy household and seek alternative solutions to their unresolved needs; also, early age at marriage by the children might predict divorce because they impulsively select less-than-suitable partners and have failed to experience successful anticipatory socialization during which they could learn necessary coping skills to deal with interpersonal relationships successfully. Regarding cohabitation, it is possible that the offspring of divorced parents have learned unconventional adaptations to life stress or because they have an inhibition against committing to long-term relationships. Cohabitation, in turn, might increase the probability of divorce because it reflects a preexisting disposition to escape from, rather than commit to, interpersonal relationships that are at all stress-inducing. Uncommitted individuals experience less normative constraint against marital dissolution when the cohabiting individuals eventually marry.
Abusive Parenting or Experiences of Abuse
The practices of abusive parenting and the experience of abuse have both been observed to be intergenerationally continuous, and both have been the object of analyses or speculations that attempt to explain these continuities in terms of intervening processes. Among the related mediating mechanisms that have been offered are the following: social learning, antisocial attitudes, mate selection, attitudes of insecurity, inadequate coping mechanisms, and neurodevelopmental factors. The intergenerational transmission of harsh parenting, where the use of corporal punishment is replicated by the children of those parents when they become parents, appears to be mediated by the social learning of aggressive behavior from the physically punitive parents (Muller et al., 1995);
132
Chapter 5
that is, intergenerational parallelism in physically punitive parenting is mediated by aggressive behaviors learned by the children of punitive parents in the course of the socialization process. Consistent with this formulation are numerous studies that indicate a relationship between being the object of corporal punishment and aggressive responses to others (McCord, 1988; Muller, Fitzgerald, Sullivan, & Zucker, 1994; Simons et al., 1991). Considering possible mediating factors in the observed intergenerational transmission of risk for child sexual abuse, McCloskey and Bailey (2000) speculated that attachment theory might provide a mediating mechanism accounting for intergenerational risk for sexual abuse: It is possible that women who have been sexually abused develop an ‘internal working model’ of sexual relationships that encompasses exploitive, coercive, and domineering behavior among men. If such a relationship ‘template’ results from early exposure to sexual abuse, then these women might be more tolerant of men either in their relationships or their social spheres who are potential abusers of their daughters. (p. 1032)
Dissociation has been suggested as a mediating variable intervening between the experience of child abuse and the exercise of abuse as adults (Narang & Contreras, 2000): Importantly, level of dissociation significantly mediated the observed association between physical abuse history and physical abuse potential. Thus, over-reliance on dissociation appeared to function as a mechanism explaining the relation between abuse history and abuse potential. (p. 660)
Parental attachment has been suggested as one mediator of the intergenerational continuity of abusive behavior such that children who are the objects of abusive parental behavior experience insecurity in their relationship with their parents, which, in turn, increases the probability that, as parents, these insecure individuals will manifest similar rejecting/abusive behavior that they experienced as children toward their own children (Main & Goldwyn, 1984). Recent neurodevelopmental perspectives implicate the development of the brain as mediating intergenerational continuity of parenting patterns. Thus, Chen and Kaplan (2001) citing Perry (1997) pointed out that researchers in the neurodevelopment field argue ...that childhood experiences affect the development of the brain, which in turn influences emotional, behavioral, and cognitive development. The
Intervening Processes
133
development of the brain is highly use-dependent. Growing up in a violent or neglectful family environment tends to impair the development of the higher and more complex part of the brain, which controls primitive impulses and promotes capabilities of empathy and attachment, while facilitating the development of the lower and simpler portion of the brain, which prompts adaptive survival functions, such as a ‘fight or flight’ state. Conversely, early experiences of good parenting that assist the development of the higher portion of the brain should promote empathetic capabilities for better interpersonal relations. Although brain research focuses on the impact of early experiences, assuming continuity of human behavior, parental behavior assessed during adolescence should in part reflect substantial early experiences. (p. 20)
Simons and his associates (1995,) reported that intergenerational continuity in harsh parenting was mediated by an antisocial orientation on the part of the parents: Harsh treatment as a child was associated with a general antisocial orientation, which in turn, predicted chronic aggression toward one’s children. Further, parents with this antisocial behavior trait were likely to engage in violence toward their spouse. Physical aggression toward the marital partner was related to violence toward children but this relationship was eliminated when the effect of antisocial behavior trait was controlled, suggesting that the two phenomena are correlated because each is a reflection of an underlying antisocial orientation. This pattern of findings is consistent with criminological theories that view persistent aggression toward family members as an expression of a general antisocial syndrome that has its roots in inept parenting (e.g., Gottfredson & Hirschi, 1990; Patterson et al., 1992). (p. 162)
Many of these same mediators are specified as mediators in the observed instances of intergenerational continuities in deviance.
Deviant Behavior Both the theoretical and empirical literatures suggest a number of conceptually related variables that are hypothesized or observed to mediate (in whole or in part) observed instances of intergenerational parallelism in deviance. The mediating constructs include G1 deviant behavior observed at a later stage in the life course, G1 adult experiences of family conflict, G1 psychological distress subsequent to G1 deviance, G1 adult parenting patterns, and socioeconomic states.
134
Chapter 5
Adult Deviant Behavior
The observation of intergenerational parallelism in deviance between youths in one generation and their children at the same developmental stage is explained often in terms of the intervening influence of adult deviance on the part of the first-generation subject. In effect, as Smith and Farrington (2004, p. 242) observed, “an important mediating mechanism for intergenerational continuity of antisocial behavior may be the intra-generational continuity of antisocial behavior into early adulthood.…” Intragenerational stability in deviant behavior between childhood and adulthood is observed frequently (Huesmann, Eron, Lefkowitz, & Walder, 1984b), and adult deviance is associated with deviant behavior by the children of those adults (Farrington, 2002; Stern & Smith, 1995). The intervening influence of G1 adult deviance on G2 deviance might reflect the operation of any of a number of mechanisms. Adult deviance by G1 subjects effectively models and legitimates such behavior for their children. Such behavior also might induce distressful circumstances that the child adapts to through the adoption of deviant coping mechanisms, particularly in the absence of models for effective conventional coping mechanisms. Distress-inducing circumstances for the G2 youth include the interrelated phenomena of family conflict, G1 adult psychological distress, and inadequate or maleficent parenting patterns.
Family Conflict
Intergenerational continuity of childhood behavior might be mediated through selection of spouse and marital conflict that leads to an environment conducive to behavioral problems in the next generation (Belsky, Youngblade, & Pensky, 1989; Caspi & Elder, 1988). A large body of the literature in fact suggests that the intergenerational parallelism noted for various forms of deviant behavior is mediated by marital conflict (Offord, Allen, & Abrams, 1978; West & Prinz, 1987). Deviant behavior on the part of parents is related to marital conflict (Billings, Kessler, Gomberg, & Weiner, 1979; Hooley, Richters, Weintraub, & Neale, 1987; Jacob & Krahn, 1988; Ruscher & Gotlib, 1988), and marital conflict among parents is associated with deviant adaptations on the part of their children (Amato & Keith, 1991; Grych & Fincham, 1990; Loeber & Stouthamer-Loeber, 1986).
Intervening Processes
135
A case in point is the work of Caspi and Elder (1988) as described by Putallaz and her associates (1998): Caspi and Elder found strong correlations among personality, marital, and parenting difficulties within each generation, correlations which held in each succeeding generation, that is, they were reproduced intergenerationally. Their analyses suggested that from one generation to the next, unstable personalities (involving explosive behavior and irritability) developed through socialization within family contexts characterized by marital conflict and poor parenting. Children with unstable personalities or problem behaviors then carry this instability into adulthood, increasing the likelihood that they will establish a marital relationship which is high in conflict. This conflictual relationship then creates a context more conducive to non-optimal parenting, which results in problem behaviors in their children, the next generation. Thus, within a generation, problem behaviors lead to problem relationships; across generations, problem relationships create a socialization environment which results in offspring developing problem behaviors, and the cycle continues. (p. 410)
The mediating role of parental conflict in the frequently observed association between some form of deviant behavior in the first generation and deviant behavior in the second generation might be accounted for by a number of mechanisms. Parental conflict might be regarded as a stressor in the lives of the children that is adapted to through the use of deviant mechanisms. Parental conflict might reflect the failure of the parents to employ effective coping mechanisms in dealing with stress and, therefore, the failure to transmit such coping mechanisms to their children, who use maladaptive mechanisms in the absence of more effective coping patterns. Parental conflict might also reflect the necessity of focusing energy on the marital relationship to the exclusion of oversight of their children’s activities, thus permitting the performance of deviant acts that might otherwise be forestalled by parental monitoring.
Psychological Distress
As an example of the mediating influence of psychological distress, Kaplan and Liu (1999) reported that a substantial intergenerational parallelism in antisocial character for a sample of mother-daughter dyads tested at the same developmental stage (early adolescence) was partially decomposed by the mediating influence of the mother’s psychological distress. The psychological distress was, in part, influenced by poor child-rearing practices that were an outcome of antisocial character in the first generation.
136
Chapter 5
However, child-rearing patterns had no independent effect on antisocial character, but, rather, they have an indirect effect via its association with psychological distress. Apparently, first-generation female antisocial dispositions lead to ineffective parenting skills that contribute to the emotional distress of a mother. The inability to control the behavior of her daughter and being the object of the daughter’s hostile responses to the inadequate parenting style presumably leads to psychological dysfunction. The mother’s psychological distress, in turn, precludes the transmission of adequate coping styles, a sense of self-efficacy, and self- acceptance to the daughter who adapts to her circumstances with an antisocial coping style. Consistent with these speculations are a number of related findings in the literature. Thus, Conger and his associates (1994) reported that adolescent externalizing and internalizing symptoms are the outcome of parental hostility secondary to parent-adolescent conflict, marital conflict, parent’s depressed mood, and economic pressure. These dynamics are compatible with the observation that the mid-life mother is experiencing the physical, psychological, and social changes associated with this stage in the life cycle and with Rossi’s (1980) suggestion (cited by La Sorsa and Fodor, 1990) that the parent undergoing mid-life developmental stress might influence low morale and family strain by complicating the parent’s ability to cope with the adolescent child or by contributing to the adolescent’s experience of stress. Perhaps the strain associated with the sequelae of early antisocial involvement on the part of the mid-life parents, exacerbated by the experience of stresses associated with mid-life crises, intensified the disruption of adequate parenting practices and the experience of psychological distress (to which the inadequate parenting also contributes). The intense distress experienced by the parent precludes the transmission of adequate coping resources to the adolescent child and, in effect, invites the child to adopt antisocial responses in more or less effective attempts to reduce the adolescent’s own experiences of psychological distress. Findings relating to the mediating influence of psychological distress is consistent also with the observation by Gest and his associates (1993) that the experience of negative life events (presumably associated with subjective distress) involving the parents of the subjects was associated with the subjects’ increases in conduct problems. Effective parenting was associated with fewer parent-related negative events during the subjects’ adolescence. The effective parenting did not compensate
Intervening Processes
137
for the negative effects of the adversity and it did not moderate the effects of adversity. These results are also interpretable (along with those of Kaplan and Liu, 1999) as indicating that effective parenting is an epiphenomenon of the paucity of subjectively distressful experiences. It is further possible that parental psychological dysfunction reflects both the prior inability to cope with children, in particular, and to cope more generally. If this is the case, then parental psychological dysfunction might reflect the inability to transmit adequate coping patterns to the next generation and, therefore, the need to adopt deviant adaptations to deal with life stress. Consistent with this, Holloway and Machida (1991) reported that the use of coping strategies involving distancing, escape/avoidance, and social support was associated with symptoms of distress. However, the use of active behavioral and cognitive coping strategies was associated with feelings of control in child-rearing situations and with authoritative parenting. Child-Rearing Patterns
Whether regarded as antecedents, consequences, or independent correlates of other mediating variables, arguably, parenting practices have excited the most interest among students of the intergenerational continuity of deviant behavior. Intergenerational parallelism in antisocial behavior is, in part, accounted for by the mediating influence of parenting practices such that antisocial behavior leads to less-than-salutary parenting, which, in turn, influences antisocial behavior in the next generation (Patterson & Dishion, 1988). First-generation parenting patterns have been observed to mediate the relationship between a variety of first-generation and second-generation deviant patterns (Chassin, Pillow, Curran, Molina, & Barrera, 1993; Conger, Patterson, & Ge, 1995; Roosa, Tien, Groppenbacher, Michaels, & Dumka, 1993). Influence of parental drinking on drinking by the adolescent appears to be mediated by family management practices and prohibitions against involvement of children in alcohol abuse by other family members (Peterson et al., 1994). First-generation deviant patterns influence the development of less-than-acceptable parenting techniques. The families of substance abusers tend to display poor parenting skills (Kandel, 1990) and increasingly apply punitive responses to children (Smyth, Miller, Janicki, & Mudar, 1995). Sons of alcoholic fathers experience less effective discipline than sons of control fathers. In general,
138
Chapter 5
quality of parental discipline was worse in the substance-abusing families (Tarter, Blackson, Martin, Loeber, & Moss, 1993). Poor parenting styles, in turn, influence second-generation antisocial outcomes (Sansbury & Wahler, 1992; Simons, Johnson, & Conger, 1994). The research literature consistently supports hypothesized associations between parenting patterns and any of a variety of deviant outcomes (Barnes & Farrell, 1992; Baumrind, 1991; Roosa et al., 1993; Sampson & Laub, 1993), including substance abuse patterns (Tarter et al., 1993), internalizing disorders (Straus & Kantor, 1994), and antisocial behavior (Sansbury & Wahler, 1992). Simons and his associates (1995) indicated that aggression toward family members might be a particular expression of a general antisocial syndrome that has its roots in inappropriate parenting. A range of specific parenting patterns have been implicated in second-generation antisocial outcomes. Parenting practices that increase the probability of any of a range of deviant outcomes include inadequate supervision, ineffective disciplinary practices, and rejecting, distrusting, and abusive parental responses toward the child (Barber, Olsen, & Shagle, 1994; Kumpfer, 1995). Loeber and Stouthamer-Loeber (1986, pp. 29–30) conducted a meta-analysis of concurrent and longitudinal studies of the relation of family factors to juvenile conduct problems and delinquency and concluded: “Analyses of longitudinal data show that socialization variables, such as lack of parental supervision, parental rejection, and parent-child involvement are among the most powerful predictors of juvenile conduct problems and delinquency.” Late delinquency was anticipated by child-rearing patterns characterized by authoritarian approaches, absence of sensitive communication and negotiation between parent and child, and the use of physical punishment (Newson, Newson, & Adams, 1993), which weaken the ties between parent and child. Parenting patterns that attenuate the relationship between the parent and child, in turn, influence the child’s deviant behavior (Loeber & Stouthamer-Loeber, 1986; Mann & MacKenzie, 1996; Patterson & Stouthamer-Loeber, 1984; Wooton, Frick, Shelton, & Silverthorn, 1997). One of the more robust findings in the literature is the association between parental deviance and parenting practices that attenuate the relationship between the parent and the child. The weakening of the parent-child bond refers both to the absence of positive affect expressed between the parent and child and the absence of oversight by the parent of the child’s behaviors. For example, depressed parents are less responsive to their children
Intervening Processes
139
(Goodman & Brumley, 1990) and alcoholic fathers are less directive toward their children than nonalcoholic fathers (Jacob, Ritchey, Cvitkovic, & Blane, 1981). The mechanisms through which first-generation child-rearing patterns lead to second-generation antisocial outcomes, again, appear to be highly variable. Ineffective parenting patterns impede internalization of conventional normative prescriptions and proscriptions (Brook, Brook, Gordon, Whiteman, & Cohen, 1990), which would contravene deviant acts through the exercise of self-control (Giever, 1995). Being raised by generally aggressive parents increases the probability that aggressive behavior will be regarded as normative. Children reared in homes in which the parents are punitive tend to become egocentric. Both unbridled expressiveness and egocentricity increase the likelihood of antisocial behavior (McCord, 1988). Deviant outcomes by children in response to dysfunctional family patterns, in turn, evoke rejecting or avoidant patterns from the parents that exacerbate dispositions to engage in deviant outcomes on the part of the children (Tarter et al., 1993). The weakening of parental ties might attenuate social controls that might have forestalled acting out of deviant impulses. Parenting patterns that reflect overly aggressive or withdrawing behavior might be modeled to deviant extremes by the children in place of available conventional coping mechanisms. In sum, at least in part, the relationship observed between antisocial behavior in successive generations might be accounted for by the intervening influence of poor parenting practices that are the outcome of the deviant behavior by the earlier generation and that, in turn, increase the likelihood of antisocial behavior in the next generation (Patterson & Dishion, 1988). Whether the effects of G1 deviance on G1 parenting practices and the effects of G1 parenting practices on G2 deviance are mediated by other intervening mechanisms (and the nature of these mechanisms) remains to be determined. Socioeconomic Status
Variables related to socioeconomic status (educational level, income, occupational status) might reasonably be hypothesized to intervene in the process of intergenerational continuity of deviance. First-generation deviance, by virtue of negative social sanctions or their sequelae, might foreclose opportunities for upward social mobility, and frustration associated with foreclosed opportunities and their correlates occasion deviant
140
Chapter 5
adaptations. The literature supporting such speculation is not abundant. However, some studies do suggest the intervening influence of socioeconomic-status-related variables in observed instances of intergenerational continuity of deviance. For example, mothers’ education levels, in part, mediated the intergenerational association between mothers’ childhood aggression and withdrawal, on the one hand, and their children’s functioning, on the other hand (Serbin et al., 1998). Further, teenage parenthood observed in the first generation decreases parental encouragement of the daughter remaining in school and staying in school is inversely related to teenage pregnancy (Manlove, 1997). However, the precise mechanisms through which G1 deviance affects G1 or G2 socioeconomic status and socioeconomic status affects G2 deviance, remain to be determined.
Structural Equation Models The magnitude of intergenerational parallelism that is observed, whether for the total sample or for a subgroup defined by a moderating value, is explainable in either or both of two ways: in terms of mediating/intervening processes or in terms of intergenerational continuity of intragenerational causes. In this chapter, theoretically informed models are estimated that specify mediating mechanisms that, in part, account for observed intergenerational continuity (for the full sample) of deviant behavior and of two salient causes of deviant behavior: disposition to deviance and psychological distress. The models specify that G1 deviant behavior (or its putative causes) have statistically significant consequences that, in turn, have significant consequences for G2 deviance (or its putative causes). The effect of specifying the intervening variable is to attenuate the original observed intergenerational effect. The structural equation models specifying each of several intervening (mediating) constructs for each of the intergenerational patterns (of deviant behavior, disposition to deviance, and negative self-feelings) are summarized in Tables 5.1, 5.2, and 5.3, respectively. Column 1 in each table specifies the intervening construct. Column 2 recalls the baseline intergenerational parallelism coefficient (the effect of G1 on G2). Column 3 indicates the residual intergenerational effect after specifying the mediating construct. The difference between columns 3 and 4 signifies the degree to which the intergenerational effect was accounted for by the mediating construct. Column 4 in each table indicates the effect that the G1
Table 5.1. Unstandardized (b)/Completely Standardized (B) Effects in Structural Equation Models Specifying Intervening Constructs in Intergenerational Parallelism of Deviant Behavior (1)
(2)
(3)
(4)
(5)
Intervening Constructs
Baseline Intergenerational Effect for Deviant Behavior (Prior to Specifying Intervening Construct)
Residual Intergenerational Effect for Deviant Behavior (After Specifying Intervening Construct)
Effect of G1 Deviant Behavior on Intervening Construct
Effect of Intervening Construct on G2 Deviant Behavior
G1 young adult socioeconomic status G1 young adult distressful emotions G2 youth perception of parental religiosity
0.240/0.239
0.161/0.117
−0.295/−0.214
−0.194/−0.293
0.240/0.239
0.164/0.182
0.150/0.236
0.339/0.240
0.240/0.239
0.163/0.184
0.565/0.287
0.091/0.203
Table 5.2. Unstandardized (b)/Completely Standardized (B) Effects in Structural Equation Models Specifying Intervening Constructs in Intergenerational Parallelism of Disposition to Deviance (1) Intervening Constructs
G1 young adult Socioeconomi status G1 parental socioeconomic status G1 family conflict reported by G1 parent of G2 youth
(2) Baseline Intergenerational Effect for Deviant Behavior (Prior to Specifying Intervening Construct)
(3)
(4)
(5)
Residual Intergenerational Effect for Deviant Behavior (After Specifying Intervening Construct)
Effect of G1 Deviant Behavior on Intervening Construct
Effect of Intervening Construct on G2 Disposition to Deviance
0.311/0.216
0.078/0.056*
−0.593/−0.388
−0.382/−0.421
0.311/0.216
0.101/0.068*
−0.449/−0.347
−0.498/−0.435
0.311/0.216
0.153/0.109*
0.708/0.349
0.219/0.316
Note. An asterisk indicates not significant.
142
Chapter 5
Table 5.3. Unstandardized (b)/Completely Standardized (B) Effects in Structural Equation Models Specifying Intervening Constructs in Intergenerational Parallelism of Negative Self-Feelings (1)
(2)
(3)
(4)
(5)
Intervening Constructs
Baseline Intergenerational Effect for Deviant Behavior (Prior to Specifying Intervening Construct)
Residual Intergenerational Effect for Deviant Behavior (After Specifying Intervening Construct)
Effect of G1 Deviant Behavior on Intervening Construct
Effect of Intervening Construct on Negative Self-Feelings
G1 young adult stressful emotions G1 family conflict reported by G1 parent of G2 youth
0.385/0.228
0.214/0.128
0.331/0.423
0.503/0.236
0.385/0.228
0.230/0.136
0.657/0.352
0.235/0.260
construct (e.g., deviant behavior) had on the intervening construct, and column 5 indicates the effect of the intervening construct on the G2 construct (e.g., deviant behavior).
Deviant Behavior Three theoretically informed constructs were specified as, in part, accounting for intergenerational continuity in deviant behavior: G1 young adult socioeconomic status, G1 young adult distressful emotions, and the G2 youth’s perception of the G1 parent’s religiosity. These intervening constructs variously relate to the continuity of G1 deviant behavior over the life course and to the influence of the G1 parents on G2 youths with regard to G2 deviant behavior. Socioeconomic Status
The performance of deviant behavior by G1 youths has important structural consequences for the vertical social mobility of the G1 youths and these consequences, in turn, influence the likelihood of G2 youths
Intervening Processes
143
engaging in deviant behavior. Deviant acts by G1 youths have the consequence of evoking negative social sanctions that stigmatize and deprive the G1 youth of resources that are necessary for the youth’s need satisfactions. These facts (further) alienate the G1 youth from conventional society; thus, he or she loses motivation to achieve according to conventional standards and, rather, opts for deviant values or deviant adaptations for the satisfaction of conventional values. Among the sanctions that are employed by conventional society is suspension from school, and among the deviant adaptations opted for by the alienated youth, is dropping out of school. This outcome along with the concomitant difficulties of prior stigmatization decrease the likelihood of upward mobility in the social structure and increase the likelihood of downward social mobility by young adulthood for the G1 subject. Among the consequences of these outcomes for the G2 youth are experiences of stigmatization, a paucity of social resources, the failure to learn conventional adaptations from alienated parents, and the learning of ineffective maladaptive outcomes. By this reasoning, it is to be expected that the position in the conventional hierarchical structure would mediate intergenerational parallelism in deviant behavior. To test this expectation, a structural equations model was estimated that specified an inverse effect of G1 deviant behavior on the G1 young adult position in the social structure and an inverse effect of the G1 young adult socioeconomic position on G2 deviant behavior. The structural effects are reported in the first row of Table 5.1. The latent constructs for G1 and G2 deviance are specified as described earlier in chapter 3. The socioeconomic position of the G1 subject as a young adult was specified as a latent construct reflected in two measurement variables: years of formal schooling, measured on an 11point scale ranging from “no formal schooling” to “postgraduate degree;” and a 6-item scale ranging from “lower class” to “upper class” in response to the followingquestion: “Judging by the prestige or the respect people have for the occupation, education, income, family, and group memberships of you and your family members, do you think you are in the ….” The items were administered when the G1 subjects were young adults. Higher scores indicated higher placement in the socioeconomic structure. As hypothesized, G1 young adult socioeconomic status partially decomposed intergenerational parallelism in deviance. Referring to the completely standardized (bottom) coefficients in Table 5.1, G1 deviant
144
Chapter 5
behavior was inversely related to G1 young adult socioeconomic status (B = −0.214) and G1 young adult socioeconomic status was inversely related to G2 deviance (B = −0.293). Although the residual degree of intergenerational parallelism in deviance remained significant (B = 0.177), the magnitude of intergeneration parallelism in deviance was reduced from the baseline model prior to specifying the mediating influence (B = 0.239). Thus, the mediating influence of G1 adult socioeconomic status, in part, explained the observed intergenerational parallelism of deviance. Although these results will not be reported here in detail, it should be noted that similar results were observed for the intervening influence of G1 socioeconomic status when the G1 adult was 35–39 years of age using the same two indicators as reported above along with an additional indicator of occupational prestige. Distressful Emotions
The performance of deviant acts on the part of G1 youths for whatever reason is expected to have a range of consequences that would adversely affect the life course of the youths into adulthood. The experience of being the object of consequent negative sanctions, concomitant stigmatization, and social rejection by conventional others, loss of social support and of the instrumental resources ordinarily provided by conventional others, and the learning of deviant and inefficacious coping patterns rather than more conventional and effective adaptive responses increases the likelihood that the person will experience feelings of alienation from the conventional world. These feelings would be exacerbated as the G1 adult, bereft of resources and conventional effective coping responses, inevitably experiences failure and rejection in the range of social roles. The experience of chronic dysphoria on the part of G1 adults, in turn, has implications for the G1 adults. The experience of distress influences the inadequate performance of social roles—most notably, the parental role. The neglectful or abusive parenting and the concomitant failure of G1 parents to transmit effective conventional coping parents to G2 youths, along with stressful social contexts that are the lot of the G1 deviant adult, influence feelings of alienation and dysphoria and consequent deviant adaptations on the part of the G2 youth. These theoretically informed linkages that are compatible with the above-reviewed empirical literature suggest that dysphoric emotions should mediate and partially decompose intergenerational continuity in deviant
Intervening Processes
145
behavior. To test this formulation, a model was established that specified a positive effect of G1 deviance during early adolescence on G1 distressful emotions measured during early adulthood and a positive effect of G1 young adult distressful emotions on G2 adolescent deviance. The estimation of this structural equation model is summarized in the second row of Table 5.1. The latent constructs for G1 and G2 deviance are measured as described in chapter 3. The latent construct for Distressful Emotions is reflected in four measurement variables, each of which reflects responses by the G1 subject during early adulthood. In all cases, higher scores indicate greater experience of distressful emotions. The first measure is based on cumulative responses to two items: “Can you think of any time during the past several years, say since you started junior high school until now, when you were very unhappy of when you were under great stress or pressure, almost more than you could bear?” “…when you were extremely happy, so happy you could hardly be more pleased?” (reverse coded). The second measurement variable is a cumulative score based on the number of “unhappy” (as opposed to either “happy” or “neither happy nor unhappy”) responses that the G1 young adult indicated in response to the question: “Looking back over the years since you started junior high school until now, in general how happy have you been about: the number or kinds of friends you had; your relations with your parents; your family’s or your income; your experiences at school; your jobs or work experience; your marriages; your relationship with males/females (opposite sex); being accepted and included by other people or groups; your race, religion, or ethnic group; your physical appearance?” The third measure is a cumulative score based on “yes” responses to the following question: “Can you think of any time between when you were in the seventh grade and now that: you felt alone and without anyone to help you solve your problems; you didn’t know who to go to for help; your situation was so different you had to learn everything all over again?” The fourth measurement variable is a cumulative score based on the frequency (often, sometimes, hardly ever, never) with which they experienced dysphoric emotions thinking about their life as a whole or about specific kinds of relations (friendships, on the job, unemployment, marital status, parenthood, extended family). The emotions include upset, worried, satisfied (reverse coded), happy (reverse coded), unsure of yourself, bothered, angry, ashamed, bored, dissatisfied, and lonely,
146
Chapter 5
As hypothesized, G1 experiences of distressful emotions as young adults in fact mediated the observed intergenerational parallelism in deviant behavior between G1 and G2 youths: G1 deviant behavior had a positive effect on G1 young adult distressful emotions (B = 0.236) and G1 young adult distressful emotions was positively associated with G2 youthful deviant behavior (B = 0.240). The residual direct effect of G1 youthful deviance on G2 youthful deviance (b = 0.182) remained statistically significant. However, this effect was reduced from that observed in the baseline model (b = 0.239) prior to specifying G1 young adult distressful emotions as a mediating construct. Parental Religiosity
Both on theoretical grounds and congruent with the theoretical literature, it is to be expected that G1 youthful deviance would have the effect of alienating the G1 youth from conventional social institutions and motivate seeking alternative deviant adaptations through which he or she might gain an acceptable self-image. The negative social sanctions that are the consequences of G1 deviant behavior are both reflected in and influence social rejection and deprivation of resources that are precursors to social failure. In response to the experience of rejection and failure that are secondary to the stigmatization associated with being the object of negative social sanctions, the G1 subject comes to reject the validity of conventional values and becomes more amenable to the adoption of alternative deviant standards that he or she might more easily approximate and so facilitate a more acceptable self-image. The alienation from conventional institutions on the part of G1 adults, in turn, has implications for the deviant behavior of G2 youths. The alienation from conventional social institutions and values is transmitted to the G2 youth, and the G2 youth inherits the fruits of the infelicitous sequelae of the G1 parent’s youthful and possibly continuing deviance. The combination of the learned negative attitudes toward conventional social institutions and the attenuation of social controls that is implicit in such attitudes, along with the stressful circumstances associated with a less-than-savory family reputation and concomitant negative sanctions (alienation from conventional institutions is itself an occasion for being the object of negative social sanctions) increases the likelihood of the G2 youth engaging in deviant activities. Although G1 rejection of conventional institutions is to some extent implicit in downward social mobility (insofar as this implicates an
Intervening Processes
147
amotivational syndrome), arguably a more germane indicator of rejection of conventional social institutions is nonparticipation in religious institutions. Not participating in organized religion at the same time implies attenuation of ties to conventional values and social networks, nonresponsivity to social controls that might otherwise constrain deviant impulses, and vulnerability to negative social sanctions by virtue of violating norms prescribing participation in religious institutions. Accordingly, a structural equations model is estimated that specifies the mediating role of G1 adult religiosity in the intergenerational continuity of deviant behavior. The model specifies a positive effect of G1 youthful deviance on G1 adult nonparticipation in religious institutions (as this is perceived by the G2 youth) and a positive effect of G1 adult nonparticipation on G2 youthful deviant behavior. The results of estimating the structural equation model are summarized in the third row of Table 5.1. The latent constructs for G1 and G2 deviance are reflected in the same measurement variables as described in chapter 3. The degree of nonparticipation in religious institutions was reflected in four measurement variables: the reports by G2 youths of the frequency (a 5-point scale ranging from “once a week or more” to a high score of “hardly ever or never”) of attendance at religious services by the youth’s mother or father and of the importance of religion (along a 4-point scale ranging from “very important” to a high score of “not at all important”) in the life of the G2 youth’s mother and father. Referring to the completely standardized (bottom) coefficients in Table 5.1, it will be observed that, as expected, G1 youthful deviance had a significant positive effect on the G1 parent’s nonparticipation in religion, when the G1 parent was 35–39 years of age (B = 0.287), and the G1 parents nonparticipation in religion had a positive impact on the G2 youth’s deviant behavior (B = 0.203). The residual direct intergenerational effect on G1 youth deviance on G2 youth deviance remained statistically significant (B = 0.184), but the magnitude of the effect decreased from that observed in the baseline model (B = 0.239) before specifying the intervening construct. In summary, part of the explanation of observed intergenerational parallelism in deviance is accounted for by specification of intervening constructs that are modeled as consequences of G1 youths’ deviance and as antecedents of G2 youths’ deviance. The mediating constructs are interpretable in terms of (1) the adverse consequences of G1 youth deviance that alienate the G1 youth from conventional social institutions over the life course via stigmatization and deprivation of resources that are secondary to the negative sanctions induced by the G1 youth deviance and (2)
148
Chapter 5
the transmission by alienated G1 adults of: deviant coping patterns, circumstances that are likely to be stress-inducing, thus motivating deviant adaptations, and attenuated social controls that (had they been effective) would have forestalled the acting out of deviant adaptations. It will be noted that the mediating influences of the intervening constructs are modest in their explanatory value. This is the case, in part, because the effects of G1 deviance on intervening constructs, and the effects of the intervening constructs on G2 deviance are themselves contingent on numerous specifiable circumstances. The continuity and transmission of deviant behavior are more likely to evoke contingent consequences because of their visibility, unlike disposition to deviance and negative self-feelings—intrapsychic phenomena that need not be as visible to others (except for their behavioral consequences). In any case, as will be observed shortly, the explanatory value of intervening constructs in accounting for observed intergenerational parallelism in disposition to deviance and negative self-feelings (salient hypothetical causes of deviant behavior) is greater than in accounting for continuity of deviant behavior. Rather (as will be observed in chapter 6), continuity in deviant behavior is better explained in terms of intergenerational continuity of intragenerational causes of deviant behavior (such as disposition to deviance, negative social sanctions, and deviant peers).
Disposition to Deviance Because disposition to deviance is known to be a salient cause of deviant behavior, it is essential to understand the intergenerational continuity of this construct as well. In part, such continuity is explained in terms of the consequences of G1 disposition to deviance over the G1 life course that, in turn, have consequences for G2 deviance. Three constructs were estimated to be mediating constructs, two of which relate to structural disadvantage and the third of which relates to familial relations. The results of the three structural equation models that specify constructs as intervening in the intergenerational parallelism of disposition to deviance are summarized in Table 5.2. Unlike the models specifying intervening constructs in the relationship between G1 and G2 deviant behavior, the specification of the intervening constructs in the relationship between G1 and G2 disposition to deviance reduces the relationship to nonsignificance.
Intervening Processes
149
Structural Disadvantage
As in the case of intergenerational continuity of deviant behavior, socioeconomic status was expected to mediate the observed intergenerational parallelism in disposition to deviance. The development of disposition to deviance as it becomes apparent in behavioral expressions is likely to evoke sanctions in the forms of social rejection and deprivation of resources that further cause alienation of the G1 youth from the conventional normative structure. The alienation is expressed in the form of loss of motivation to conform to and the genesis of motivation to deviate from the normative structure. The loss of motivation to pursue conventional values together with the impediments posed to upward social mobility that accompany or reflect negative social sanctions increase the likelihood that G1 youth deviance would be inversely related to the G1 socioeconomic status of the G1 young adult. Undesirable placement in the conventional system of stratification, in turn, has consequences such as the experience of psychological stress associated with a paucity of material and interpersonal resources, alienative attitudes, and (perhaps) social-class-related child-rearing practices that together increase the inability of the G2 youth to forestall stress, gain social acceptance, and achieve success. The alienation of the G2 youth that is consequent to these outcomes increases the disposition to seek alternative deviant adaptations, lacking effective conventional coping resources, through which the G2 youth can satisfy his needs. Based on this reasoning, two structural equations models were estimated that specify an inverse effect of G1 youth disposition to deviance on G1 socioeconomic states and an inverse effect of G1 socioeconomic status on G2 youth disposition to deviance. The first of these specifies G1 young adult socioeconomic status and the second model specifies G1 socioeconomic status when the G1 subject was the parent of the G2 adolescent. The two models are summarized in the first two rows of Table 5.2. The measurement variables for the G1 young adult socioeconomic status latent construct are as described earlier in this chapter in connection with the mediating influence of G1 young adult socioeconomic status in the intergenerational continuity of deviant behavior. For the socioeconomic status constructs measured when the G1 subject was a parent of the G2 adolescent, a third measurement variable was included that reflected occupational prestige, gauged by a 14-point scale ranging from laborer to professional with a doctoral degree or the equivalent.
150
Chapter 5
Regarding the specification of G1 young adult socioeconomic status as an intervening construct, reference to the completely standardized values (following the slashes) in the first row of Table 5.2, will indicate that, as hypothesized, G1 youth disposition to deviance had an inverse effect on G1 young adult socioeconomic status (B = −0.388) and G1 young adult socioeconomic status had an inverse effect on G2 youth disposition to deviate (B = −0.421). The specification of the intervening construct had the effect of reducing the baseline intergenerational coefficient (B = 0.216) to a nonsignificant level (B = 0.056). Thus, the explanatory value of the mediating construct was appreciably greater for disposition to deviance than for deviant behavior. Although the results do not demonstrate the validity of the theoretical linkages that informed the analysis, they are, indeed, congruent with that reasoning. Reference to the second row in Table 5.2 will indicate that virtually identical results were obtained when the model specified G1 socioeconomic status when the G1 parents were 35–39 years of age as an intervening variable. G1 disposition to deviance was inversely related to G1 socioeconomic status (B = −0.347) and G1 socioeconomic status was inversely related to G2 youth disposition to deviance (B = −0.435). The residual direct effect of G1 disposition to deviate on G2 disposition to deviate following specification of the intervening parental socioeconomic status (B = 0.068) was not significant and reflected an appreciable reduction from the baseline intergenerational coefficient (b = 0.216) observed before specifying the intervening construct. Familial Relations
Another construct specified as intervening in the relationship between G1 and G2 disposition to deviance reflects intrafamilial relations. The construct reflects conflictual relations between G1 spouses and between G1 parents and G2 children. On theoretical grounds, family conflict is expected to be both an outcome of G1 youth disposition to deviance and an antecedent of G2 youth disposition to deviance. G1 disposition to deviance, composed of both loss of motivation to conform to and motivation to deviate from conventional norms, along with the behavioral manifestations of these antisocial attitudes, would be expected to eventuate in selecting deviant social networks, learning maladaptive coping mechanisms, and being deprived of learning how (and motivation) to conform to conventional role definitions, including the roles of spouse and parent. The conflictual family relations
Intervening Processes
151
that result from this situation throughout the life course of the G1 respondent have important consequences for the G2 youth’s own disposition to engage in conflictful relations. Familial conflict serves to model for the G2 youth maladaptive and antisocial attitudes. In addition, the familial conflicts (marital and parent-child) are intrinsically distressing situations that evoke coping, defensive, or adaptive responses that represent attempts to assuage, if not forestall, the distress. Because the G2 youth has not been socialized to utilize conventional and effective coping patterns and often shares the parents’ alienative attitudes, he or she will be disposed to engage in deviant patterns. As this is expressed in a hypothetical structural equations model, G1 youth disposition to deviance is expected to have a positive effect on familial conflict when the G1 subject is 35–39 years of age. G1 familial conflict as an adult parent of the G2 youth, in turn, is expected to have a positive effect on G2 youth disposition to engage in deviant activities. The estimation of the structural effects and measurement models is summarized in the third row of Table 5.2. Familial conflict was reflected in terms of three measurement variables reported by the G1 respondent at age 35–39 years. The first measure is a cumulative score consisting of eight items indicating that: the marriage/partnership is characterized by lack of reciprocity, lack of openness, unreliability of the spouse, unapproachability, lack of mutual affection, lack of spousal appreciation, and unsatisfactory sexual relations. The second measure is a cumulative score based on three items indicating that the G1 adult insulted, swore at, threatened to hurt, or physically hurt the spouse/partner. The third measure concerns the parent-child relationship and is a 4-item cumulative index indicating that when the (G2) child does something wrong, the (G1) parent often ridicules the child, acts cold or unfriendly, physically punishes the child, and expresses anger. Referring to the completely standardized coefficients in the third row of Table 5.2 (following the slash), it will be observed that familial conflict indeed mediated the observed intergenerational parallelism in disposition to deviance. G1 youth disposition to deviance was positively related to familial conflict (B = 0.349) and familial conflict, as reported by the G1 respondent at age 35–39, anticipated G2 youth disposition to deviance (B = 0.316). By interpolating the intervening construct, the baseline intergenerational parallelism (B = 0.216) was reduced to nonsignificance (B = 0.109). The findings were consistent with the assumption that the observed intergenerational parallelism reflected, at least in part, a causal relationship that was mediated by a familial conflict construct or its correlates.
152
Chapter 5
Negative Self-Feelings Negative self-feelings on theoretical grounds is expected to have, within each generation, important effects on deviant disposition and, less directly, on deviant behavior. This expectation appears to be warranted in view of the demonstration of the observed temporal relationship using longitudinal data (Kaplan and Johnson, 2001). Indeed, as will be demonstrated in chapter 6, intergenerational continuity in negative self-feelings, together with the G1 and G2 intragenerational effects of this construct, in part explain instances of observed intergenerational parallelism in disposition to deviance. Consequently, it is necessary to understand and explain intergenerational parallelism of negative self-feelings in terms of mediating mechanisms as well as (in chapter 6) common intragenerational antecedents that manifest intergenerational continuity. Negative selffeelings is continuous across generations both because it has continuity across the G1 life course and because later in the G1 life course, it directly affects the attitudes, behaviors, and outcomes of the G2 youths. Intragenerational stability in G1 psychological distress is observed, in part, because G1 youth negative self-feelings reflects a stable, changeresistant trait as well as a momentary state. The G1 youth negative selffeelings also reflects the absence of adequate coping mechanisms that might have permitted successful performance and the evocation of positive attitudes from significant others and the forestalling of stressful experiences and the assuagement of distressful emotions. Across generations, the G1 adult transmits inadequate coping, defensive, or adaptive patterns and occasions for the G2 youth experiencing unbearable distress. These linkages are illustrated by two intervening constructs. One of these reflects the inability of the G1 youth to forestall negative self-feelings over the life course. The other intervening construct reflects the genesis of stressful circumstances and the modeling of maladaptive coping mechanisms by the G1 parent for the G2 youth. G1 Young Adult Stressful Emotions
The stability of negative self-feelings throughout the G1 subjects’ life course is prerequisite to the construct exercising an effect on G2 negative self-feelings. Implicit in the expectation of intragenerational continuity of negative self-feelings is the ongoing experience of unhappy life circumstances, including unmet obligations and unsatisfactory relationships,
Intervening Processes
153
absence of social support to facilitate meeting one’s obligations and to assuage distress, and concomitant distressful affects evoked in all of one’s social relationships. This constellation of outcomes is captured in the construct described earlier under the rubric of distressful emotions as a mediator of intergenerational continuity of deviance in connection with the model summarized in the second row of Table 5.1. Early G1 youth negative self-feelings is expected to anticipate as well as reflect ongoing G1 distressful life circumstances throughout the life course, as well as the lack of adaptive coping patterns or social support that might have forestalled or rendered distressful life circumstances tolerable. These G1 circumstances in the lives of the G1 young adult subjects, in turn, would be expected to increase the likelihood of G2 youth negative self-feelings insofar as (1) the inability of G1 adults to perform adequately in their social roles (including the parental and breadwinner roles) would increase the likelihood of G2 youths encountering stressful life circumstances and (2) the absence of G1 adult conventional and adaptive mechanisms would preclude the transmission of effective and conventional coping mechanisms that might have forestalled the unabated experience of negative self-feelings. Hence, a structural equations model was estimated that specified the intervening influence of distressful emotions in the intergenerational continuity of negative self-feelings. Whereas negative self-feelings reflected the experience of negative affect, the construct of distressful emotions reflected more the social context in which distressful emotions were experienced and the absence of resources that might have forestalled or assuaged the experience of distress. Both G1 and G2 negative self-feelings were measured as described in chapter 3 and G1 young adult distressful emotions was measured as described in connection with the estimation of the model summarized in the second row of Table 5.1. These measures, it will be recalled, reflected the experience of distressful life circumstances, the consequent experience of dysphoric affects, and the absence of intrapsychic and interpersonal resources that might have forestalled or assuaged the experience of negative self-feelings. The estimation of the model specifying G1 young adult distressful emotions as intervening in intergenerational parallelism of negative selffeelings is summarized in the first row of Table 5.3. As hypothesized, referring to the completely standardized coefficients (following the slash), youth negative self-feelings has a direct positive effect on G1 young adult distressful emotions (B = 0.423) and G1 young adult distressful emotions
154
Chapter 5
has a direct positive effect on G2 youth negative self-feelings (B = 0.236). The residual effect of G1 youth negative self-feelings on G2 youth negative self-feelings (B = 0.128) is reduced appreciably from the baseline intergenerational coefficient (B = 0.228) observed prior to specifying the intervening construct; that is, the intergenerational continuity in negative self-feelings is explained (decomposed) in part by the intervening distressful emotions construct: G1 youth negative self-feelings affects G1 young adult distressful emotions and G1 young adult distressful emotions affects G2 youth negative self-feelings. Family Conflict
On theoretical and empirical grounds, it is to be expected that G1 youth negative self-feelings, reflecting a lack of psychological and social resources to forestall or assuage negative self-feelings and to negotiate interpersonal relations in mutually satisfying ways, would eventuate in conflictful family relationships during later adulthood. Familial conflict would model maladaptive rather than efficacious conventional coping patterns, which would engender other stressful circumstances, leading to further negative self-feelings for the G2 youths. These linkages suggest that familial conflict reported by the G1 adult would mediate the observed intergenerational parallelism in negative selffeelings. To test this presumption, a structural equations model was estimated that specifies a positive association between G1 youth negative self-feelings and G1 parental reports of familial conflict and between reports of familial conflict and G2 negative self-feelings. Both G1 and G2 youth negative self-feelings were measured as described in chapter 3. Familial conflict was measured as described earlier in this chapter in connection with the decomposition of observed intergenerational parallelism of disposition to deviance, a structural equations model that was summarized in the third row of Table 5.2. The construct involved marital dissatisfaction and spousal abuse by the G1 adult (when 35–39 years of age) and parenting patterns involving ridicule, cold and unfriendly attitudes, expressions of anger, and physical punishment. The estimation of the model is summarized in the second row of Table 5.3. As hypothesized, familial conflict mediated intergenerational parallelism in negative self-feelings: G1 youth negative self-feelings had a positive effect on G1 adult (parent) reports of familial conflict
Intervening Processes
155
(B = 0.352) and G1 adult reports of familial conflict had a positive effect on G2 youth negative self-feelings (B = 0.260). The baseline coefficient for intergenerational parallelism of negative self-feelings (B = 0.228) was attenuated to B = 0.136 as a result of specifying the intervening construct.
Summary The observed intergenerational parallelism in deviant behavior and two putative causes of deviant behavior is accounted for, in part, by intervening mechanisms that are consequences of the G1 phenomenon and antecedents to the G2 phenomenon. The mediating function of the putative intervening construct is suggested by the combination of its temporal mediation and the effect that it has on attenuating the magnitude of the baseline intergenerational continuity after specifying the intervening constructs. The mediating mechanisms implicate both intragenerational stability of circumstances over the life course of the G1 subject between youth and parenthood, and the proximal impact of the G1 parent on the G2 youth. For intergenerational parallelism of deviance, the (modest) mediating effects that were hypothesized and observed suggested that G1 youth deviance had consequences that alienated the G1 subject from conventional social institutions over the life course. The alienation presumably resulted from the deprivation of social resources and the stigmatization that were secondary to being the object of negative social sanctions. The proximal impact of G1 adult intervening constructs on G2 youth deviance are presumed to operate through the G1 transmission of stress-inducing maladaptive coping mechanisms to the G2 youth that occasion G2 deviant adaptations and through the attenuation of G1 social controls exercised by the G1 adult. Disposition to deviance, a salient cause of deviant behavior in each generation, has been observed to exercise intergenerational continuity through a variety of mediating mechanisms. Disposition to deviance both reflects and has consequences that exacerbate G1 alienation from the conventional normative system throughout the G1 life course. The mediating variables are said to reflect G1 experiences of rejection and failure, concomitant dysphoria, alienative dispositions directed toward the conventional socionormative system, the adoption of ineffective and deviant coping mechanisms over the period between youth and parenthood, and
156
Chapter 5
engagement in dysfunctional role performance in a variety of social spheres, including the family. These constructs, in turn, occasion the G2 experience of stress, alienation from the conventional social structure, and the disposition to adopt deviant adaptations to life stress. Negative self-feelings, an antecedent of both disposition to deviance and deviant behavior, also has consequences for the life course of the G1 subject that, in turn, influence the negative self-feelings of the G2 youth. The mediating variables that, in part, explain intergenerational parallelism in psychological distress are said to reflect, or incur at later stages in the life course, the experience of distressful life circumstances of failure and rejection, the lack of effective coping mechanisms, and dysfunctional role performance in a variety of social relational contexts. These outcomes, in turn, either constitute stressful life circumstances for the G2 youth or reflect the transmission of ineffective coping patterns to the G2 youth that might otherwise have served to forestall or assuage the negative selffeelings associated with threatening life circumstances. The foregoing models suggest that part of the explanation of intergenerational parallelism lies in the causal effects of the G1 phenomenon on intervening constructs that, in turn, have consequences for the G2 phenomenon. However, another facet of the explanation of intergenerational parallelism implicates intergenerational continuity of variables that have common influences on the phenomenon of interest in each generation. It is to this explanatory mode that attention is now directed.
6 Intergenerational Continuity of Intragenerational Causes of Deviant Behavior and Its Correlates
In chapter 5, it was observed that intergenerational parallelism in deviance, disposition to deviance, and psychological distress were, in part, accounted for by intervening processes whereby the consequences of these constructs, in turn, had causal implications for the same constructs in the next generation. The modeling of the intervening construct had the effect of attenuating the original intergenerational parallelism, thus demonstrating the explanatory significance of the intervening construct. In the present chapter, the decomposition of the intergenerational effects implicates intergenerational continuity of common antecedents of deviance (or disposition to deviance or psychological distress) in each generation. For example, in both the first and second generation, deviance is influenced by dysfunctional parenting patterns, and dysfunctional parenting practices are continuous across generations. The continuity of the common intragenerational cause of deviance (here, dysfunctional parenting) might be decomposed (explained) by intervening processes or by intergenerationally continuous variables that are common antecedents of the intragenerational causes. In the first section of this chapter, literature will be reviewed that comment on or demonstrate that intergenerational continuities in deviance or its salient correlates might be accounted for by intergenerational continuities of phenomena that are taken to be causally implicated in the onset of deviance-related phenomena in each generation. In the second section, 157
158
Chapter 6
a number of theoretically informed models will be estimated that specify intergenerationally continuous constructs that influence deviance (or its correlates) in each generation and (by so doing) account in large measure for the initial observation of intergenerational parallelism of deviant behavior (or its salient correlates).
Literature on Intergenerational Continuities of Intragenerational Causes An abundant, but largely speculative, literature exists that addresses the explanation of intergenerational parallelism in terms of common antecedents. The literature is said to be speculative in the sense that one or more elements of the explanatory model is missing in the empirical estimates. In addition to a baseline model that specifies an association between the dependent variable in one generation with the same variable in the second generation, the model would necessarily specify an effect of the G1 explanatory variable on the G1-dependent variable, an effect of the G2 explanatory variable on the G2-dependent variable, an intergenerational effect of the G1 explanatory variable on the G2 explanatory variable, and an intergenerational effect of the G1-dependent variable on the G2-dependent variable. Insofar as the baseline intergenerational effect for the dependent variable is attenuated by the specification of the intragenerational and cross-generational effects of the explanatory variable, to that extent is the original intergenerational effect explained by the intergenerational continuity of a putative common intragenerational antecedent. Although systematic tests of the assumptions are rare, numerous observers presume that intergenerational parallelism in deviant behavior or other phenomena need not be explained by mediating processes alone (Markowitz, 2001): The relationship between experiencing violence as a child and engaging in violence against children and spouses as an adult may be due, in part, to other factors such as stable differences in personality, genetic similarities, or parents’ and children’s exposure to similar adverse social environments (Tedeschi & Felson, 1994). (p. 215)
Similarly, Huesmann and his associates (1984b) observed: Quite probably the impressive stability of aggressive behavior across time and generations is a product both of the continuity of constitutional factors
Intergenerational Continuity of Intragenerational Causes
159
and the continuity of environmental factors. Certainly, constitutional characteristics including genetic factors (Lagerspetz & Lagerspetz, 1974; Christiansen, 1977), hormonal factors (Kreuz & Rose, 1972; Rada, Kellner, & Winslow, 1976), and neurological trauma (Mark & Ervin, 1970) play some role in aggressive behavior. Just as certainly, a person’s environment and learning history strongly influences his or her aggressive tendencies (Bandura, 1973; Berkowitz, 1962; Eron et al., 1971; Huesmann et al., 1984[a]). The relative importance of these factors is arguable. What is not arguable is that aggressive behavior, however engendered, once established, remains remarkably stable across time, situation, and even generations within a family. (p. 1133)
Independent of the explanation of intergenerational parallelism in terms of mediating causal linkages, it has been noted also that continuity might be stimulated by sharing of genes between generations and by sharing the same physical and social circumstances (Van Ijzendoorn, 1992): Living in the same neighborhood, and even in the same family house, may constitute factors stimulating intergenerational continuity. This is an example of cumulative continuity in which an individual’s environment reinforces a certain interactional style, thereby sustaining the behavior pattern across the life course, and maybe even across generations (Caspi, Bem, and Elder, 1989). Therefore, the strength of intergenerational parenting will be inflated if genetic transmission of parenting determinants and contextual stability influencing the continuity of parenting attitudes and behaviors are not taken into account. (pp. 77–78)
For example, the relationship observed among male subjects between having punitive parents and being punitive parents, although significant in bivariate relationship, was obviated when controlling for father’s occupation and subject’s IQ. Thus, intergenerational parallelism might be accounted for by common antecedents and exercise their influence intergenerationally. It is possible that subcultural values relating to the expression of aggressive impulses might account for the parallelism. It is also possible that genetically transmitted IQ influenced the ability to use alternative forms of social control (Lefkowitz et al., 1978). Further, in the case of experience of sexual abuse, it is possible that this heightened risk of sexually abused women’s daughters could be explained through extended family relationships that include the mothers’ family members. That is, women might sustain contacts with their family of origin despite prior sexual abuse experiences perpetrated by family members, and this continued contact places their daughters at the same risk to which they were once subjected. (McCloskey & Bailey, 2000, p. 1032)
160
Chapter 6
Although numerous common antecedents of deviant behavior (and its salient correlates) might be specified, two of the more frequently cited common antecedents of deviant behavior that manifest intergenerational continuity are presented for purposes of illustration: parenting patterns and socioeconomic status. Each of these examples is reported to influence outcomes of interest (deviant behavior and its correlates) in each generation and each of these is reported to display intergenerational continuity. Further, each of these examples of intergenerational parallelism might be explained in terms of intervening causal processes, common intragenerational antecedents that display intergenerational continuity, or moderating circumstances.
Parenting Patterns It has been well established that intragenerational experiences of, for example, abusive parenting (both physical and verbal) influence deviant behavior in that generation. Research has further suggested that observations of intergenerational parallelism in deviant behavior might be accounted for by the intergenerational transmission of causes of deviant behavior such as abusive parenting (Frias-Armenta, 2002). Regarding the intragenerational effects of parenting patterns, abusive parenting patterns lead to aggressive behavior on the part of the child throughout the life span (Luntz & Widom, 1994; Malinosky-Rummell & Hansen, 1993). Other reports have suggested a linkage between abusive parenting patterns and the subsequent alcohol abuse of the children suffering parental abuse (Brown & Anderson, 1991; Downs, Miller, Testa, & Panek, 1992). Victims of parental abuse also experienced greater emotional problems, including self-derogation and depression (Carlin et al., 1994; Silverman, Reinherz, & Giaconia, 1996; Styron & JanoffBulman, 1997). Intragenerational influences of parenting practices on children’s deviant behavior are nicely summarized by Smith and Farrington (2004): Consistent evidence shows that various aspects of maladaptive parenting by one generation influence the unfolding of antisocial behavior in the children of the next generation. Aspects of parenting that have been implicated include poor supervision, inconsistent discipline, parental conflict, and lack of affection and support (for example, Farrington 1995; McCord, 1991; Patterson, Reid, & Dishion, 1992; Simons, Wu, Johnson, & Conger, 1995; Stern & Smith, 1999).
Intergenerational Continuity of Intragenerational Causes
161
There is also evidence that parental attitudes and parenting styles, particularly authoritarianism and permissiveness, have a negative impact on children’s development and behavior (Baumrind, 1966, 1978; Simons, Whitbeck, Conger, & Chyi-In [sic, Wu], 1991; Steinberg, 2000). Intergenerational continuities as well as negative behavioral impact have perhaps been most clearly documented in the case of very harsh or abusive parenting (Egeland, Jacobvitz, & Sroufe, 1988; Kaufman & Zigler, 1987; Rutter, Quinton, & Liddle, 1983). (p. 232)
These influences on deviance in each generation are themselves continuous across the generation’s behavior. Thus, poor parental supervision, which within-generation is observed to be associated with child conduct behaviors, is observed to be continuous across generations (Smith & Farrington, 2004); that is, those who experience poor supervision by their parents tended to be poor supervisors as parents. In general, poor parenting practices in each of successive generations are transmitted from one generation to the next (Patterson & Dishion, 1988). Evidence has been reported of intergenerational transmission of perceived parenting. For example, correlations have been reported between parents’ and offspring’s early memories of parental rearing, particularly with regard to the experience of emotional warmth as recollected by fathers and their offspring (Lundberg et al., 2000). The somewhat weaker correlations between mothers’ and offspring’s recollections of early parental rearing referred primarily to dimensions of parenting relating to overprotection and rejection. Putallaz and her associates (1998) citing the work of Main and associates (1985) noted: ...strong evidence for intergenerational continuity in attachment status as the parents’ mental representations of the security of their own attachment was related to the measures of their children’s attachment security taken during infancy, with the relations being particularly strong for mothers relative to fathers... (p. 7).
Thus the characteristics of family context tend to be reproduced across generations. Offer and his associates (2003) contrasted groups of adolescents who were followed up 27 years later: The Continuous growth group, our early research showed, had no adolescent turmoil and sailed through adolescence and young adulthood. Twenty-seven years later, this group still stands out from the other two developmental groups. They have more traditional families and interests, reflecting the families of their childhood and adolescence.
162
Chapter 6 The Tumultuous group, by contrast, was more likely to come from disrupted or disturbing backgrounds, to be disturbed as teenagers, and less focused on family and traditional institutions as adults. The Surgent group was between these two poles, less well adjusted than the Continuous group and better adjusted than the Tumultuous group in adolescence and continuing this pattern into their middle-age years. (p. 7)
The intergenerational parallelism in patterns of parenting or (more broadly) family context is explainable (as is deviance, disposition to deviance, and negative self-feelings) in terms or moderating variables, intergenerational continuity of intragenerational effects, intervening processes, and (perhaps) residual direct effects reflecting social modeling mechanisms.
Moderating Variables
The degree of continuity across generations in intergenerational continuity of parenting patterns is contingent on a number of circumstances, one of the more salient of which is cultural change or stability in attitudes toward child-rearing patterns. For example, general population surveys tend to indicate a decreasing use of harsh discipline (Straus & Gelles, 1988; Straus et al., 1980). Consistent with these surveys, Simons and his associates (1991) reported an appreciable decrease in harsh parenting over the course of a single generation. Insofar as the generations differ in the degree to which there is social support for certain behaviors, to that extent a lesser degree of intergenerational continuity in the behavior will be observed.
Intergenerational Continuities of Intragenerational Effects
The explanation of intergenerational continuities in parenting patterns often might be explained in terms of constructs that have effects on parenting practices in each generation and themselves are intergenerationally continuous. Cases in point are parental beliefs about parenting, socioeconomic status, and depression. In each generation, beliefs about parenting affect actual parenting practices. These beliefs, via their influences on parenting practices, are transmitted intergenerationally in addition to the transmission of the parenting behaviors themselves (Simons et al., 1992):
Intergenerational Continuity of Intragenerational Causes
163
Overall, the results indicated that mothers and fathers convey their beliefs to their adolescent children via their parenting practices. Parents who endorse corporal punishment tend to engage in harsh discipline, and this style of discipline was in turn associated with adolescent children subscribing to corporal punishment as an effective approach to discipline. Similarly, parents who believe that parenting has a major impact on child development were likely to be involved and supportive, and their adolescent children, in turn, endorsed the idea that quality of parenting influences child outcomes. (p. 833)
Evidence exists that socioeconomic status might represent a contextual variable, the continuity of which accounts, in part, for intergenerational parallelism in any of a variety of phenomena. For example, it has been observed that parents of low education relative to highly educated parents are more likely to use harsh discipline in raising boys (Simons et al., 1991). Both mothers’ and fathers’ education were correlated with education and economic hardship in their respective families of origin. These findings are consistent with a conclusion that intergenerational continuity in socioeconomic status is associated with harsh parenting of boys within the respective generations and that these associations might attenuate any observed intergenerational continuity in harsh parenting of boys. The intergenerational continuity of parenting patterns might be explained, in part, by the effects of depression on parenting in each generation and the intergenerational continuity of depression (mediated by first generation adult parenting) (Whitbeck et al., 1992): In summary, this research replicated retrospective and prospective studies on parental depression and the parent-child relationship by combining both methodologies into an intergenerational model. The model provides evidence for the negative consequences of parental depressed mood on parenting behaviors and suggests a cycle of intergenerational continuity of depressed affect and rejecting parenting. These findings contribute to a growing body of evidence that suggest that childrearing practices are transmitted across generations both indirectly through the development of personality traits that affect parenting and directly through modeling of parenting behaviors. (p. 1044)
Intervening Processes
Frequently, the observed intergenerational continuity in family processes, including parenting patterns, is accounted for by the consequences of the G1 family processes that, in turn, influence G2 family
164
Chapter 6
processes (i.e., in terms of intervening mechanisms). The intergenerational continuity of abusive or constructive parenting has been observed to be mediated by a number of variables. Chen and Kaplan (2001), for example, have observed that the intergenerational continuity of constructive parenting is decomposed, in part, by the intervening influence of interpersonal relations and social participation. More often than not, the intervening variables are taken to be G1 deviant responses to their parents’ parenting patterns. In this connection, Brook, Whiteman, and Zheng (2002) reported: This investigation provides evidence for a process of intergenerational influence whereby the effects of the grandmother-parent relationship are transmitted across two generations through their association with the parent-toddler relationship. A more distant grandmother-parent (G1/G2) relationship when the parent was an adolescent predicted unconventional behavior and psychological problems in the parent as the young adult (G2), which then were associated with his or her having a less warm relationship with his/her own child (G2/G3). This, in turn, predicted problem behavior in the toddler (G3). There appears to be a parallel between the path linking the grandmother-parent relationship and parent personality and the one linking the parent-toddler relationship and toddler behavior. These relationship variables seem particularly strong in view of the fact that the G1/G2 was assessed when the parent was an adolescent and the G2/G3 relationship was assessed when they were parents of toddlers. (p. 73)
Although West and Farrington’s (1977) data might be interpreted as fitting a model by which socialization experiences mediate the influence of first-generation deviant behavior on secondgeneration deviant behavior, the data might also be taken to fit a model of intergenerational continuity of criminogenic environments that produce deviant patterns within each generation. The continuity of criminogenic environments is mediated by the deviant behavior of the firstgeneration subject: [a] constellation of adverse features of family background (including poverty, too many children, marital disharmony, and inappropriate childrearing methods) among which parental criminality is likely to be one element, leads to a constellation of antisocial features when sons reach the age of eighteen, among which criminality is again likely to be one element. Young delinquent adults, by their irresponsibly hedonistic attitudes and ineffectual methods of coping with social demands, tend to recreate for their own children the same undesirable family environments, thus perpetuating from one generation to the next a range of social problems of which delinquency is but one symptom. (p. 161)
Intergenerational Continuity of Intragenerational Causes
165
Social Modeling
The intergenerational continuity in parenting patterns that is observed frequently is, in part, explained as a direct effect reflecting socializationrelated processes. Stability effects across generations that remain after hypothesizing stability of intragenerational causes and intervening variables are compatible with explanations involving socialization/ role-modeling processes. However, it is rare that studies are able to test these explanations of intergenerational stability. As Kahn and Anderson (1992) observed: Our results are consistent with the hypothesis that teen births are repeated across generations because mothers socialize their daughters with attitudes and preferences regarding the appropriate time and way to start a family. Unfortunately we cannot measure these underlying family-building preferences for either the mother or her daughter.... (p. 50)
Socioeconomic Status The intergenerational parallelism of deviant behavior (or its correlates) is explainable frequently in terms of the intragenerational effects of socioeconomic status on deviance (or its correlates) in each generation and the intergenerational continuity of socioeconomic level. A case in point is the role of stable socioeconomic environments on intergenerational patterns of teenage fertility (Kahn & Anderson, 1992): Early first births also may be reproduced across generations because of their indirect impact on the socioeconomic and family environment in which children grow up. Mothers with early first births are more likely than other mothers to raise their children in impoverished conditions (Garfinkel & McLanahan 1986). This is the case in part because they are more likely to have come from a poor background, but also in part because their early pregnancy may have cut short their education as well as the development of marketable job skills (Hofferth & Moore 1979; Moore & Burt 1982). In addition, teenage mothers are more likely to raise their children as single parents, either because they never married or because their teen marriage placed them at a higher risk of divorce (Furstenberg 1976; Moore et al. 1981). Growing up in these kind of socioeconomic (e.g., with limited financial resources and few opportunities for upward mobility) has been shown to place daughters at a higher risk of teenage pregnancy. This pattern may be due to the low opportunity costs for early pregnancies for poor youth in view of their relatively low aspirations and the reduced alternatives that they foresee (Hofferth, Kahn, & Baldwin 1987; McLanahan & Bumpass 1988). Also, the local neighborhood and peer group may play important roles either
166
Chapter 6 by encouraging risky behavior or by providing role models of other teens who become pregnant outside marriage (Hogan & Kitagawa 1985). Finally, teens growing up in poorer environments also may face higher risks of pregnancy because they have less access to effective contraception (Kahn, Rindfuss, & Guilkey 1990). (pp. 41–42)
Continuity in socioeconomic level across generations has also been implicated in intergenerational parallelism of putative causes of deviant behavior such as parenting patterns. Thus, Simons and his associates (1995) observed, with regard to the role of social-class influences in accounting for intergenerational transmission of harsh parenting: Given this association, there are at least two ways in which social-class influences might account for the correlation between parenting practices across generations. First, the linkages that have been reported between generations may merely represent the tendency of adult children to replicate the lower social-class status of their parents with its accompanying stressors and life-style, a life-style that may promote irritability and increase the likelihood of harsh parenting (Burgess and Youngblade, 1988)...., growing up in a lower class family may influence people’s approach to child-rearing or their child-rearing values regardless of the socioeconomic level that they are able to achieve. This possibility requires taking into account that the socioeconomic status of the parents during the years that the adult children was growing up. (pp. 159–160)
The intergenerational effects of socioeconomic status on deviance and its correlates are mediated and moderated by a number of constructs. Socioeconomic status has indirect effects on deviant outcomes in a number of ways. Thus, lower income parents are more likely to exert powerassertive discipline particularly in circumstances where they feel the environment is out of control; lower socioeconomic status (SES) parents are less likely to provide support, to respond to the child’s needs, and to express affection; lower SES tends to stimulate marital conflict and abusive and neglectful responses to children and to attribute willfulness to their children; lower SES parents manifest negative affect expressed as depression, feelings of helplessness, alienation, and related affects in response to stressful life circumstances (Baumrind, 1994). As Pears and Capaldi (2001) observed: Analyses indicated that the effect of SES on parents’ abusive behavior toward their children was mediated by the parents’ own experiences of abuse. This is consistent with other studies showing that the effects of contextual factors like SES on behaviors may be mediated by more proximal factors, such as parenting skills (DeGarmo, Forgatch, & Martinez, 1999). (p. 1455)
Intergenerational Continuity of Intragenerational Causes
167
These adverse outcomes of socioeconomic status are moderated by environmental factors (Baumrind, 1994): Community support lessens poor parents’ tendency to engage in punitive and coercive discipline, in part by monitoring the caregiving of mothers in distress (Cochran & Brassard, 1979; Wilson, 1989). Well-organized neighborhoods with caring neighbors who show respect for each other appear to buffer the effects of poverty (Garbarino & Sherman, 1980). The presence of a stable partner or a compatible grandmother has been found to be an important source of support for poor families (Furstenberg, Brooks-Gunn, & Morgan, 1987; Kellam, Ensminger, & Turner, 1977). Emotional support from their own mothers has been found to decrease adolescent mothers’ hostility and indifference toward their children (Colletta, 1981). (p. 362)
That intergenerational continuity in socioeconomic-related phenomena exists does not appear to be in doubt. With regard to education and occupation, some degree of intergenerational continuity has been observed. Second-generation subjects who have low educational and occupational positions tend to have parents who also have low educational and occupational positions (Shlonsky, 1984). Regarding being black, Rutter and Madge (1976) observed for England: Continuities may... be evident within socio-cultural groups. During the two generations or so that black people have been in this country, they have continued to be disadvantaged in educational attainment, employment, and housing. All of these patterns constitute a form of cycle of disadvantage... (pp. 303–304)
However, the degree of intergenerational continuity in socioeconomic status is contingent on a number of circumstances. Intergenerational transmission of socioeconomic status is moderated by broad social changes such as the movement of the occupational opportunity structure from a more ascriptive oriented to a more achievement oriented structure, and changes in familial values from emphasis on obedience toward emphasis on autonomy (Biblarz, Bengtson, & Bucur, 1996). The change from ascription to achievement would decrease the extent to which social class in a later generation would be inherited from an earlier one as would the greater decision-making power of the child to choose his or her own destiny. Others (Harper et al., 2003) have noted that to achieve optimal child development and to forestall intergenerational transmission of poverty: ...two contexts need to exist. The first is that which enables individual participation in society through positive social relations and socio-political
168
Chapter 6 structures. The second is a wider enabling environment that presents opportunities for development, such as adequate labor markets and the state provision of public services... The wider enabling environment is vast. For example, in relationship to the generation of the retention of financial assets to allow for child survival, protection and development, a wide range of actions is needed. This enabling environment is well-documented in the development literature and includes (to name a few) the generation of nonexploitative adult labor markets, land distribution, asset retention, nondiscriminatory inheritance law, and social safety nets. Providing these and adequate education requires tax collection, debt relief, and reduced military spending to enable state financing, adequate provision of and access to formal education for children, and adult education, and where necessary motivational campaigns and legal action to prevent discrimination against particular children. (pp. 548–549)
The consideration of the instances of intergenerational continuity of parenting patterns and socioeconomic status serve to illustrate how observed patterns of intergenerational parallelism in deviance and its correlates might be explained in terms of intergenerational continuity of intragenerational (antecedents) causes of deviant behavior. In some cases, the explanation is that the observed intergenerational parallelism in deviance is the spurious outcome of continuity in causal contexts; that is, G1 deviant behavior has no direct or indirect causal impact on G2 deviant behavior. Rather, both G1 and G2 deviance seem to be continuous because each is the outcome of a stable deviance-inducing circumstance. The causal processes that are involved are the direct intragenerational effects of (for example) G1 and G2 antecedents on G1 and G2 deviance, respectively, the indirect effect of the G1 antecedents on G2 deviance via the G2 antecedent, and the indirect effect of G1 antecedents on G2 deviance via G1 deviance. However, these causal effects do not explain the intergenerational continuity of G1-G2 deviance. In other cases, the intergenerational parallelism in deviance is attenuated by specifying causal pathways implicating the intergenerational continuity of the G1-G2 antecedents. For example, the G1 antecedent affects G1 deviance; G1 deviance affects the G2 antecedent (thus attenuating the G1-G2 continuity of the antecedent construct), and the G2 antecedent affects G2 deviance (thus partly explaining the G1-G2 intergenerational parallelism in deviance). In any case, the continuity of G1-G2 intragenerational antecedents of deviance constitute part of the explanation for observed instances of
Intergenerational Continuity of Intragenerational Causes
169
intergenerational parallelism of deviance. Attention now turns to the estimation of a series of models informed by the theoretical framework outlined in chapter 1 that accounts for intergenerational parallelism in deviance, disposition to deviance, and negative self-feelings in terms of intergenerationally continuous common antecedents.
Explaining Intergenerational Parallelism of Deviance: Intergenerational Stability of Intragenerational Antecedents Aspects of the general theory of deviant behavior outlined in chapter 1 were tested by estimating models specifying intergenerational stability of common antecedents of deviant behavior (and its correlates) in the first and second generations. The intergenerational stability of theoretically informed constructs that were associated with the dependent variable (deviant behavior or selected correlates) at appreciable and statistically significant levels in each generation were expected to account for the original observation of intergenerational parallelism (in deviance or its correlates) in large measure. Although the intragenerational predictors were operationalized in terms of data that were collected at the same point in time as the dependent variables for the respective generation, earlier analyses (Kaplan & Johnson, 2001) established the temporal precedence of the putative causes of the dependent variables. In addition to estimating models that specify intergenerational stability of common antecedents of deviant behavior (including disposition to deviance), models will be estimated that specify intergenerational stability of common antecedents of disposition to deviance (including negative self-feelings) and of negative self-feelings. In all instances, it was expected that the specification of intergenerational continuity of intragenerational causes would (at least partially) decompose the observed intergenerational parallelism of deviant behavior (or disposition to deviance, or negative selffeelings, whatever the case).
Deviant Behavior Within both the first (G1) and second (G2) generations, it was hypothesized that each of four constructs would have a strong association with deviant behavior; and the specification of intergenerational stability of each construct would serve to attenuate the observed relationship
170
Chapter 6
between G1 and G2 deviance that was observed prior to specification of the G1 and G2 intragenerational association of the construct with deviant behavior and the intergenerational stability of the construct. The general theory specifies deviant behavior to be a consequence of the following mutually influential constructs that exert direct or indirect influences on deviant behavior: disposition to deviance, negative social sanctions, deviant friends, and experiences of social rejection. Each of these is considered in turn. Disposition to Deviance
A major influence on both G1 and G2 deviance is the disposition to engage in deviant behavior as this is reflected in the loss of motivation to conform to conventional behavioral prescriptions and the genesis of motivation to engage in deviant behaviors. The loss of motivation to conform to normative requirements in one’s membership groups is the outcome of experiences of rejection and failure in the person’s conventional membership groups. These experiences derive from any of a variety of sources, including congenitally given stigmatizing attributes, not the least salient of which are intrinsically disvalued group memberships. Failure and rejection also stem from the absence of resources that are prerequisite to the achievement of socially valued ends. The absence of resources might be traced to a variety of other circumstances, including congenitally given inadequacies and the interrelated experiences of sociocultural deprivation, faulty socialization experiences that fail to transmit adequate coping patterns but rather communicate maladaptive patterns of response to life stress, and life events that impose obligations on the person that cannot be met given resources that (where they were previously adequate) are no longer adequate to meet one’s obligations. The failure to meet one’s obligations and the correlated experiences of social rejection evoke negative self-evaluations and concomitant distressful self-feelings. In light of the failure of conventional responses to forestall experiences of failure and rejection by others and to assuage concomitant distressful self-feelings, the person is disposed to seek alternative (deviant) mechanisms to accomplish these ends. These deviant adaptive mechanisms have their own consequences (coming to the attention of the authorities, associating with deviant peers, exacerbation of social rejection) that further alienate the person from conventional socionormative systems and increase the disposition to adopt deviant
Intergenerational Continuity of Intragenerational Causes
171
adaptations to life stress. The disposition to engage in deviant adaptations to stressful circumstances would be expected to increase the likelihood of actually engaging in deviant behavior, partly as a consequence of the weakening of social controls that are secondary to the person’s alienation from conventional authorities. Hence, it was hypothesized that G1 disposition to engage in deviant behavior would increase the likelihood of G1 deviant behavior. It was further hypothesized that G1 disposition to deviant behavior would have consequences that would increase the likelihood that G2 youths would become disposed to engage in deviant behavior and that G2 youths who were disposed to adopt deviant coping patterns would be more likely to adopt deviant coping patterns. The hypothesized intergenerational continuity of disposition to deviance is based on the premise that G1 youths who acted out their deviant dispositions with contranormative acts would evoke negative social sanctions and, consequently, experience distressful self-feelings and intensified feelings of alienation from conventional socionormative systems throughout the life course. These feelings would be exacerbated by the perceived deprivation of needed resources and exclusion from conventional society. The latter would increase the likelihood of association with deviant peers, one manifestation of which might be inappropriate mate selection. This factor, together with chronic experience of distressful self-feelings and exclusion from normative influences, increases the likelihood that the G1 adult will engage in inappropriate parenting patterns characterized by neglect, abuse, and the modeling of deviant and maladaptive mechanisms. These outcomes, together with the paucity of resources available to the G2 youth and the stigmatizing circumstances associated with a conflictful and unconventional home life, increase the likelihood that the G2 youth would become disposed to adopt deviant mechanisms for dealing with life stress and such a disposition would increase the probability of in fact engaging in deviant behavior. These relations were tested by specifying a model in which G1 disposition to deviance was positively associated with G1 deviance, G2 disposition to deviance was positively associated with G2 deviance, G1 disposition to deviance influenced G2 disposition to deviance, and G1 deviance was associated with G2 deviance. The observation of an attenuation of the intergenerational effect of G1-G2 deviance noted prior to specifying the intragenerational effects and cross-generation stability of disposition to deviance would suggest that (1) the observed
172
Chapter 6
intergenerational parallelism in deviance was, in part, explainable in terms of intergenerationally continuous disposition to deviance, a construct that is associated with deviant behavior in each generation, and (2) some validation for the theoretical premises on which the model was based. The constructs deviant behavior and disposition to deviance for both G1 and G2 were measured as described in chapter 3. The latent construct deviant behavior was reflected in three measurement variables reflecting respectively theft, interpersonal violence, and substance use. The latent construct disposition to deviance was reflected in three measurement variables reflecting respectively aggressive coping patterns, disrespect of others, and contempt for conventional norms. The estimation of the structural equations model is summarized in Figure 6.1. In this and all other figures in the chapter, the unstandardized (b) and completely standardized (B) coefficients are presented for both the measurement and structural models. The completely standardized coefficients are in parentheses. As expected, within each generation, disposition to deviance has an appreciable and statistically significant effect on deviant behavior (B = 0.819 for G1 and B = 0.698 for G2) and some degree of stability exists between G1 and G2 disposition to deviance (B = 0.195). Modeling these relationships has the effect of attenuating the
Figure 6.1. Intergenerational Stability of Disposition to Deviance as a Common Antecedent of G1 and G2 Deviant Behavior
Intergenerational Continuity of Intragenerational Causes
173
intergenerational parallelism from the baseline effect of B = 0.239 to B = 0.097. The results support the conclusion that intergenerational continuity of this putative antecedent of deviant behavior (disposition to deviance) in each generation, in part, accounts for the previously observed intergenerational parallelism of deviant behavior. Negative Social Sanctions
Given the stability of deviant behavior over the life course, it is to be expected that in any generation, earlier deviant behavior will evoke negative social sanctions. The experience of negative social sanctions, in turn, will have consequences that will increase the probability of subsequent deviant behavior. The negative social sanctions function to stigmatize the person and to exclude the person from access to conventional resources that might permit resolution of salient needs. The combined circumstances of stigmatization and deprivation of resources alienate the person from the conventional sources of the sanctions and dispose the person to seek and employ alternative deviant adaptive mechanisms to assuage the distress that accompanies being the object of sanctions and to resolve life problems. As has been noted earlier, the disposition to seek alternative deviant response patterns increases the likelihood of adopting such patterns, particularly when opportunities to do so arise and perception of countervailing costs is minimal. The experience of negative social sanctions in any generation, then, has consequences that increase the probability of subsequent deviant behavior in that generation. The negative social sanctions in the earlier generation, through the intervening consequences of the G1 deviant behavior and its sequelae, are intergenerationally continuous into the second generation and have consequences for G2, similar to those described for G1, that eventuate in G2 deviant behavior. The intergenerational continuity of being the object of negative social sanctions is mediated by the G1 performance of deviant acts as adaptations to the consequences (in particular, negative social sanctions) of earlier deviance. The deviant adaptations continue to have consequences throughout the G1 life course that increase the likelihood of the G2 youth experiencing negative social sanctions. These intervening circumstances include G1 exclusion from conventional socialization experiences, conflictual and otherwise inappropriate performances of adult familial roles, socialization of G2 youths in the use of unconventional
174
Chapter 6
and maladaptive coping patterns, stigmatization of the G2 youth by association with deviant parents, G2 early performance of deviant acts as adaptive attempts to resolve life problems, and, finally, the G2 experience of negative social sanctions in response to early G2 deviant adaptations and other stigmatizing attributes. The G2 experience of negative social sanctions then has the same consequences that were described for G1 that increase the likelihood of future G2 deviant behavior. Consistent with these theoretical premises, it was hypothesized that G1 and G2 negative social sanctions would be significantly and appreciably associated with G1 and G2 deviant behavior, respectively, G1 negative social sanctions would manifest a significant intergenerational effect on G2 negative social sanctions, and specifying these effects would attenuate the originally observed G1-G2 intergenerational parallelism of deviant behavior. The latent construct, negative social sanctions, was reflected in two measurement variables. The first variable was a cumulative 3-item index indicating suspension or expulsion from school, being sent to the (school) office for punishment, and being sent to a psychiatrist, psychologist, or social worker within the last year. The second variable is a single item indicating coming to the attention of police, sheriff, or juvenile officers “for something you did or they thought you did.” The estimated structural equation model is summarized in Figure 6.2. As hypothesized, being the object of negative social sanctions was strongly associated with deviant behavior in both generations (B = 0.996 for G1 and B = 0.968 for G2); the cross-generation effect for negative social sanctions was statistically significant (B = 0.257), and the original intergenerational coefficient for deviant behavior (B = 0.239) was attenuated to nonsignificance (B = −0.029) as a result of specifying the intragenerational effects of negative social sanctions on deviant behavior and the intergenerational continuity of negative social sanctions. These results are compatible with theoretical assumptions regarding the intervening processes stemming from the consequences of being the object of negative social sanctions in the earlier generation, and they support the expectation that the observed intergenerational parallelism in deviance is explainable in terms of intergenerational continuity of negative social sanctions and the intragenerational effects of this construct on deviant behavior for both G1 and G2. The sequelae of G1 youths being the objects of negative social sanctions (including exacerbation of G1 deviance and its consequences) ultimately influences G2 youths being the
Intergenerational Continuity of Intragenerational Causes
175
Figure 6.2. Intergenerational Stability of Negative Social Sanctions as a Common Antecedent of G1 and G2 Deviant Behavior
object of negative social sanctions and the increase in G2 deviant behavior that results from this experience. Deviant Friends
On theoretical grounds, it is to be expected that membership in a deviant friendship network would increase the likelihood of engaging in deviant behavior. Participation in a deviant friendship network is the outcome of any of a number of circumstances. In some cases, youths are born into and raised in neighborhoods where what is regarded as deviant in the more inclusive (conventional) society is regarded as normative in this context. Thus, having deviant friends is the normal outcome of “conforming” to subcultural expectations. In other instances, having deviant friends is the end product of a sequence of events involving early deviance, being the object of negative social sanctions, feeling alienated and excluded from conventional society, and being attracted to and recruited into friendship groups by deviant peers. These circumstances are increasingly likely to occur as a result of sanctions that exclude the youth from conventional society.
176
Chapter 6
In any case, once a part of a deviant friendship group, the youth is increasingly likely to engage in deviant behavior. In some instances, the “deviant” behavior represents conformity to the norms of the friendship group that satisfies many of the youth’s quotidian needs. In other instances, the deviant friends provide social support for acting out motivations to engage in deviant behavior, resources that are prerequisite to engaging in such behavior (e.g., drugs), and opportunities or occasions for engaging in deviant behaviors. When youths in a given generation become part of a deviant friendship network, they set in motion a chain of events that eventuates in the youths of the next generation becoming part of a deviant peer network, with the increased likelihood that those youths would engage in deviant behavior. Participation in a deviant peer network evokes negative social sanctions from conventional authorities, alienates and excludes the person from conventional society, and increases the likelihood that the G1 adult would make unconventional choices in setting up a family, fail to be socialized to play the parenting role in a conventional manner, and fail to exercise effective social controls over the G2 youth’s activities (including choice of friends). At the same time, distressful experiences (familial and otherwise) associated with upbringing by parents with a deviant background would motivate the seeking for and adoption of deviant patterns that might forestall or assuage distressful self-feelings. These adaptations would include the selection of deviant peers, a circumstance that is facilitated by having been the object of negative social sanctions that exclude the youth from conventional contacts and facilitate contact with deviant peers. The entry into deviant friendship networks by the G2 youths would then have the same consequences for G2 deviant behavior as were described for the G1 youths. Based on these premises, a model was estimated that specified direct effects of deviant friends on deviant behavior within each generation, intergenerational continuity of deviant friends, and any residual intergenerational parallelism that might obtain for deviant behavior. The latent construct deviant friends was measured by two variables: a cumulative 2-item index indicating affirmative responses to “Do many of your good friends smoke marijuana?” and “Do many of your good friends take narcotic drugs to get high?”; and, agreement with the single item “Most of my close friends are the kinds of kids who get into trouble a lot?” The estimation of the structural equations model is summarized in Figure 6.3. As hypothesized, deviant friends was strongly associated with deviant
Intergenerational Continuity of Intragenerational Causes
177
Figure 6.3. Intergenerational Stability of Deviant Friends as a Common Antecedent of G1 and G2 Deviant Behavior
behavior in both generations (B = 0.963 for G1 and B = 0.787 for G2), the cross-generation stability effect (B = 0.255) was significant, and the intergenerational effect for deviant behavior observed prior to specifying the intrageneration and stability effects of deviant friends (B = 0.239) was attenuated to nonsignificance (B = −0.004). Thus, these findings are congruent with the theoretical framework that informed the hypothetical model and support the contention that the intragenerational effects of deviant friends on deviant behavior and the intergenerational stability effect of deviant friends are implicated in the explanation of the observed intergenerational parallelism of deviant behavior. Social Rejection
In each generation, on theoretical grounds, perceived rejection by conventional groups is expected to be associated with higher levels of deviant behavior. Social rejection, the outcome of the youth being characterized by disvalued attributes and behaviors, (further) alienates the youth from conventional normative structures, thus attenuating the social controls that might have impeded acting out deviant dispositions also resulting from the increased alienation from conventional membership groups.
178
Chapter 6
The experience of social rejection has consequences that increase the likelihood of social rejection by the next generation’s youths. Deviant behavior in G1 that is secondary to social rejection tends to be stable over the life course, causing stigmatization, social exclusion, a disposition to associate with deviant others (in part expressed in choice of marital partners), and unconventional performance of social roles (notably, parenting). Poor parenting, reflected in neglect and abuse of the G2 youth, is taken as parental rejection in the eyes of the G2 youth. The stigmatization attaching to membership in a deviant kinship network and any deviance that results from the paucity of resources and maladaptive coping patterns that are provided by the family as a result of being rejected and excluded from conventional society in turn cause the G2 youth to be the object of rejecting attitudes by other conventional membership groups such as peers and teachers. Social rejection of the G2 youth then increases the likelihood of deviant behavior through the same mechanism described for G2 youths. Informed by these theoretical premises, a structural equations model was estimated that specified: positive effects of G1 and G2 experience of social rejection on deviant behavior in G1 and G2, respectively, positive effects of G1 social rejection on G2 social rejection, and a residual effect of G1 deviant behavior on G2 deviant behavior. Experience of social rejection was modeled as a latent construct reflected in three measurement variables. The first measure, perceived rejection by parents, was a cumulative 4-item score based on the youth’s affirmation of My parents hardly ever trust me to do something on my own; As long as I can remember, my parents have put me down; My parents are usually no interested in what I say or do; My parents do not like me very much. The second measure, rejection by teachers, was a cumulative 4-item index based on the youth’s affirmation of My teachers are not very interested in what I say or do; By my teacher’s standards I am a failure; My teachers do not like me very much; My teachers usually put me down. The third measure, rejection by the “kids at school,” was a 4-item cumulative index based on the youth’s affirmation of More often than not I feel put down by the kids at school; I am not very good at the kinds of things the kids at school think are important; The kids at school are usually not very interested in what I say or do; Most of the kids at school do not like me very much. Support for the model would be taken as some validation of the theoretical premises that informed the model, and attenuation of the intergenerational stability coefficient for deviant behavior relative to that
Intergenerational Continuity of Intragenerational Causes
179
observed in the baseline model would support the proposition that intergenerational parallelism in deviant behavior is explainable partly in terms of intergenerational stability of a common intragenerational influence, namely perceived social rejection. The estimation of the model is summarized in Figure 6.4. As hypothesized, G1 and G2 perceived social rejection was strongly associated with G1 and G2 deviant behavior, respectively (B = 0.602 for G1 and B = 0.698 for G2) and a significant intergenerational stability coefficient for perceived social rejection was observed (B = 0.193). Also as expected, the baseline intergenerational stability coefficient for deviant behavior (B = 0.239) was attenuated (B = 0.145) as a result of specifying intragenerational effects of, and intergenerational stability of, perceived social rejection.
Disposition to Deviance Earlier, empirical support was observed for the theoretical premise that intergenerational parallelism in deviance is explainable, in part, by the intergenerational continuity of a common intragenerational influence on
Figure 6.4. Intergenerational Stability of Perceived Social Rejection as a Common Antecedent of G1 and G2 Deviant Behavior
180
Chapter 6
deviant behavior, namely disposition to deviance. The guiding theoretical framework that informed this expectation also leads to the expectations that intergenerational stability in disposition to deviance is explainable in terms of cross-generational stability of common intragenerational antecedents of disposition to deviance, namely negative self-feelings, negative social sanctions, membership in deviant friendship networks, and perceived social rejection. Negative Self-Feelings
The loss of motivation to conform to and the genesis of motivation to deviate from conventional socionormative systems that compose the disposition to deviate construct in each generation is the result of the experience of chronic negative self-feelings in the course of membership group experiences. These negative self-feelings stem from formal and informal negative social sanctions in response to disvalued attributes and behaviors and in response to the early deviant adaptations to being the object of such negative social sanctions (including being attracted to deviant friendship networks). The experience of distressful self-feelings in any given generation has consequences that increase the likelihood of intergenerational continuity of negative self-feelings. Within G1, distressful self-feelings associated with failure and rejection in conventional membership groups and the social rejection that accompanies the G1 youth’s disvalued attributes and behavior dispose the person to adopt deviant behaviors. Consequent deviant behaviors by G1 youths lead to stigmatization, social exclusion, deprivation of conventional social resources, and unconventional adaptations that increase the likelihood of conflictual and otherwise dysfunctional G1 family environments for the rearing of the G2 youth. The experience of neglectful and otherwise abusive childhood experiences along with the socialization into the adoption of ineffective and maladaptive coping patterns predispose the G2 youth to experience distressful negative self-feelings. These distressful self-feelings along with the recognition of their source alienate the G2 youth from the conventional environment and dispose the youth to adopt deviant response patterns. Based on these premises, it was hypothesized that within each generation, negative self-feelings would be associated with disposition to deviance, negative self-feelings at G1 would be associated with negative self-feelings at G2, and the stability coefficient for G1-G2 disposition to
Intergenerational Continuity of Intragenerational Causes
181
deviance would be attenuated relative to the observed coefficient for disposition to deviance in the baseline model. Disposition to deviance was measured as described in chapter 3 and earlier in this chapter as described with reference to the model reported in Figure 6.1. Negative self-feelings were measured as described in chapter 3, as reflected in three measurement variables. One measure was a cumulative 7-item scale reflecting self-derogatory attitudes. A second measure was a 9-item cumulative index reflecting affirmation of widely recognized symptoms of anxiety. The third measure was a 6-item cumulative index reflecting depressive affect. The estimation of the structural equation model is summarized in Figure 6.5. As hypothesized, within each generation, negative self-feelings was strongly associated with disposition to deviance (B = 0.602 for G1 and B = 0.765 for G2) and the intergenerational stability coefficient (B = 0.217) was statistically significant. Also as expected, the intergenerational stability coefficient for disposition to deviance was attenuated (B = 0.081) relative to the baseline coefficient (B = 0.216) observed prior to specifying the intragenerational effects of negative self-feelings on disposition to deviance and the intergenerational stability effect for negative self-feelings. Thus, the results are congruent with the theoretical premises
Figure 6.5. Intergenerational Stability of Negative Self-Feelings as a Common Antecedent of G1 and G2 Disposition to Deviance
182
Chapter 6
that informed the analyses, and the baseline intergenerational parallelism in disposition to deviance is explainable in part by the intergenerational continuity of a construct that has common intragenerational effects on disposition to deviance, namely negative self-feelings. Negative Social Sanctions
On theoretical grounds, being the object of negative social sanctions would be expected to dispose the youth to adopt deviant patterns as mechanisms for dealing with life stress. Being the object of negative social sanctions stigmatizes the youth, excludes the youth from conventional society and accessibility to resources that ordinarily accompany social acceptance, and so reduces motivation to conform to conventional norms and, rather, generates motivation to seek and adopt deviant alternatives to the conventional coping patterns that are no longer (if they ever were) available to the youth. Being the object of negative social sanctions in one generation has consequences that increase the likelihood of the next generation being the object of negative social sanctions, and in the next generation, negative social sanctions increase the likelihood of becoming disposed to deviant behavior. When the G1 youth becomes the object of negative social sanctions (perhaps due to deviant adaptations to earlier social rejection), the resulting acting out of deviant dispositions and their sequelae have consequences that eventuate in G1 parents modeling deviant behavior. In addition, deviant parenting practices (abuse and neglect) cause distressful self-feelings that motivate deviant adaptations by the G2 youth. The G2 deviance that is modeled after G1 patterns as ways of dealing with distressful self-feelings evokes negative social sanctions that, in turn, exacerbate disposition to deviance through the same intervening mechanisms that were described for the G1 youth. Based on these premises, a structural equations model was estimated that specified intragenerational effects of being the object of negative social sanctions on disposition to deviance for both G1 and G2 youths, a stability effect of G1 on G2 negative social sanctions, and a G1-G2 stability effect for disposition to deviance that was expected to be substantially weaker than the baseline intergenerational coefficient. Being the object of negative social sanctions was modeled as a latent construct as in the estimation summarized in Figure 6.2, reflecting sanctioning responses primarily by school and police-related authorities.
Intergenerational Continuity of Intragenerational Causes
183
Figure 6.6. Intergenerational Stability of Negative Social Sanctions as a Common Antecedent of G1 and G2 Disposition to Deviance
The estimated model is summarized in Figure 6.6. As expected, in both G1 and G2, being the object of negative social sanctions was strongly associated with disposition to deviance (B = 0.733 for G1 and B = 0.700 for G2) and a statistically significant relationship between G1 and G2 negative social sanctions (B = 0.267) was observed. Also as expected, the baseline intergenerational stability effect for disposition to deviance (B = 0.216) was attenuated to nonsignificance (B = 0.052) after specifying the intragenerational effects of negative social sanctions on disposition to deviance and the cross-generational stability of negative social sanctions. Thus, again, the results were compatible with the theoretical premises that informed the model, and the expectation that observed intergenerational parallelism in disposition to deviance was explainable, in part, by common intragenerational effects on disposition to deviance of a construct, namely negative social sanctions, that manifested intergenerational stability. Deviant Friends
Whether participation in deviant friendship networks was the outcome of (1) a deviant adaptation to social rejection by conventional groups, (2) being born in to an environment in which the individuals
184
Chapter 6
judged to be deviant by an external group were in fact conforming to normative expectations, or (3) being the object of negative social sanctions that demanded interaction among deviant peers, deviant friendships would be expected to dispose the youth to engage in deviant behavior. The expectations of deviant friends would motivate the G1 youth to conform in order to secure their approval, the deviant peers would provide resources and occasions for engaging in deviant behavior, and the deviant friendship nexus would provide rationalizations for otherwise conventional youths to adopt deviant responses. Informed by these theoretical premises, a structural equations model was estimated that specified intragenerational effects of deviant friends on disposition to deviance for both G1 and G2 youths and an intergenerational path between G1 and G2 deviant friends. The residual path between G1 and G2 disposition to deviance was expected to be weakened greatly relative to the baseline path. Deviant friends was measured as in the analysis summarized in Figure 6.3. The estimation of the structural equations model is summarized in Figure 6.7. As expected, the intragenerational effects of deviant friends were substantial (B = 0.762 for G1 and B = 0.811 for G2) and the intergenerational path for deviant friends (B = 0.246) was statistically
Figure 6.7. Intergenerational Stability of Deviant Friends as a Common Antecedent of G1 and G2 Disposition to Deviance
Intergenerational Continuity of Intragenerational Causes
185
significant. Also as hypothesized, the specification of the intragenerational effects of deviant friends on disposition to deviance and the crossgenerational stability of deviant friends served to attenuate the prespecification intergenerational stability for disposition to deviance (B = 0.216) to nonsignificance (B = 0.010). Thus, the theoretical premises that informed the hypothetical model received support, and the expectation that intergenerational parallelism in disposition to deviance would be explainable, in part, by specification of intergenerational continuity of a construct (deviant friends) that had intragenerational effects on disposition to deviance in both G1 and G2 was rewarded. Social Rejection
The perception of social rejection by conventional groups and the concomitant negative self-feelings increase alienation from those groups that are perceived as the source of the distressful self-feelings. Because conventional coping resources patently have been ineffective in forestalling or assuaging social rejection and negative self-feelings, it is to be expected that the youths in each generation who experience social rejection from conventional membership groups (family, school teachers, peers) would be disposed to seek and adopt deviant coping patterns that offer promise of reducing the distressful self-attitudes that accompany social rejection. The experience of social rejection by youths in one generation has consequences for the experience of social rejection in the next generation. As noted earlier, social rejection motivates the adoption of deviant mechanisms that offer promise of more effectively reducing distress than the manifestly ineffective coping mechanisms. Because the deviant mechanisms preclude full participation in conventional socialization activities as a result of evoking exclusionary responses, the G1 youth develops into an adult who frequently performs social roles (including spouse and parent) in deviant fashion and transmits to the G2 child deviant coping patterns while occasioning stressful family circumstances (abuse and neglect) that evoke these mechanisms. These outcomes, along with the stigma attaching to the family, evoke social rejection of the G2 youth and the same sequelae as described for the G1 youth that further dispose the G2 youth to adopt deviant patterns. Informed by these theoretical premises, a model was estimated that specified intragenerational effects of social rejection on disposition to
186
Chapter 6
deviance for both G1 and G2 youths and a G1-G2 intergenerational stability effect for social rejection. The residual intergenerational stability effect for disposition to deviance was specified but was expected to be much attenuated by the intragenerational and intergenerational effects of social rejection. Social rejection was again modeled as a latent construct reflected in three measurement variables (rejection by parents, teachers, peers, respectively) as described earlier for the model summarized in Figure 6.4. The structural equations model is specified in Figure 6.8. As hypothesized, significant and appreciable intragenerational effects of social rejection on disposition to deviance were observed (B = 0.858 for G1 and B = 0.835 for G2), as was a significant intergenerational stability effect for social rejection (B = 0.201). Also as expected, the residual intergenerational stability effect for disposition to deviance was attenuated to nonsignificance (B = 0.062) from the baseline coefficient observed prior to specification of the intragenerational and stability effects of social rejection (B = 0.216). Thus, the theoretical premises that informed the model received some support, as did the expectation that intergenerational parallelism in disposition to deviance was explainable, in part, in terms of the intergenerational stability of a construct (social rejection) that had intragenerationl influence on disposition to deviance within both G1 and G2.
Figure 6.8. Intergenerational Stability of Social Rejection as a Common Antecedent of G1 and G2 Deviant Behavior
Intergenerational Continuity of Intragenerational Causes
187
Negative Self-Feelings The construct negative self-feelings that, through its intergenerational stability and its intragenerational influence on disposition to deviance in G1 and G2, directly explains, in part, intergenerational parallelism in deviant disposition and thus indirectly explains intergenerational parallelism in deviant behavior is itself explainable in terms of the intergenerational stability and intragenerational effects of several constructs. Among the more salient of these are the interrelated constructs social rejection and negative social sanctions. Social Rejection
In the course of the normal socialization experience, individuals learn to need the approval of others, particularly those others who are in a position to satisfy or frustrate basic socioemotional or instrumental needs. Initially the most significant others are parents. Later, these others encompass school authorities and peers. The inability to evoke approving attitudes and, rather, the evocation of rejecting attitudes from these relevant others evoke distressful self-feelings. This pattern of linkages occurs in each generation. The experience of social rejection by youths in any generation has consequences that lead to the experience of social rejection by youths in the next generation. The experience of rejection by parents, teachers, and peers evokes distressful self-feelings that are associated in the youth’s mind with experiences in conventional groups. The result is to feel alienated from such groups and the readiness to adopt deviant responses, values, and groups that offer greater promise of forestalling rejection and concomitant distress than the patently ineffective conventional mechanisms. However, these mechanisms also evoke rejecting attitudes by virtue of their deviant nature. Further, the failure to learn conventional social role behavior as a parent (partly due to social exclusion as a deviant) frequently results in distressful neglect and abuse of the G2 youth. At the same time, the G1 parent fails to convey effective and conventional coping mechanisms to the G2 child that might have forestalled stressful life circumstance or assuaged the concomitant distressful self-feelings. Thus, the combination of learning maladaptive mechanisms, failing to learn effective conventional coping patterns, stigmatizing experiences as a member of a deviant familial groups, and being an object of neglectful or abusive parenting increases the
188
Chapter 6
likelihood that the G2 youth will parallel the G1 youth in both being the object of rejecting attitudes by conventional others and consequently experiencing negative self-feelings. Based on these theoretical premises, a structural equations model was estimated that specified intragenerational effects of social rejection on negative self-feelings in both G1 and G2 and G1-G2 stability in social rejection. Although a G1-G2 path for negative self-feelings was specified as well, it was expected that the coefficient would be attenuated greatly, relative to the baseline coefficient. Social rejection was measured as described in the analysis summarized in Figure 6.4. The estimation of the structural equations model is summarized in Figure 6.9. As hypothesized, social rejection (reflecting perceived rejection by parents, teachers, and peers) was appreciably and significantly related to psychological distress for both G1 and G2 youths (B = 0.897 for G1 and B = 0.754 for G2) and the intergenerational stability coefficient for social rejection (B = 0.245) was statistically significant. Also as expected, the baseline intergenerational stability coefficient for negative selffeelings (B = 0.228) was attenuated to nonsignificance (B = 0.053) as a result of specifying intergenerational stability of social rejection and the intrageneration effects of this construct on negative self-feelings in G1 and G2. The results, then, both support the theoretical premises that
Figure 6.9. Intergenerational Stability of Social Rejection as a Common Antecedent of G1 and G2 Negative Self-Feelings
Intergenerational Continuity of Intragenerational Causes
189
informed the analysis and were compatible with the expectation that intergenerational parallelism in negative self-feelings was explainable, in part, by the intergenerational continuity of a construct (social rejection) that had within-wave effects on negative self-feelings in both G1 and G2. Negative Social Sanctions
Among the causes of negative self-feelings are the negative social sanctions that are directed toward those youths who display objectionable characteristics or behaviors, including those deviant patterns that reflect deviant subcultural conformity or deviant adaptations to social rejection and other antecedents of alienation from conventional society. Being the object of negative social sanctions occasions negative self-feelings insofar as the socialization experience leads the person to value approving responses from these groups and being the object of negative social sanctions is a stigmatizing event, and these experiences deprive the person of access to valued social resources and have other outcomes that increase the likelihood of future distress-inducing experiences of rejection and failure. Being the object of negative social sanctions has consequences that increase the likelihood that the G2 youth also will be the object of negative social sanctions. Stigmatization and deprivation of resources that are consequences of being the object of negative social sanctions by G1 youths increases the likelihood of G1 youths being further alienated and adopting deviant patterns that continue into adulthood. These circumstances increase the likelihood of transmitting the disposition to adopt deviant coping patterns to G2 youths, as does the experience of stigmatizing association with the deviant familial context. The resultant deviant responses by G2 youths, in turn, increase the likelihood of G2 youths being the object of negative social sanctions, hence manifesting intergenerational continuity of negative social sanctions. In the second generation, being the object of negative social sanctions, in turn, increases the likelihood of negative self-feeling, operating through the same intervening processes as were described for G1 youths. Informed by these theoretical premises, a structural equations model was estimated that specified intergenerational continuity of being the object of negative social sanctions and within-wave effects of negative social sanctions on negative self-feelings for both G1 and G2. Negative social sanctions were measured as described in connection with the model summarized in Figure 6.6.
190
Chapter 6
Figure 6.10. Intergenerational Stability of Negative Social Sanctions as a Common Antecedent of G1 and G2 Negative Self-Feelings
The estimation of the structural equations model is summarized in Figure 6.10. As hypothesized, the intergenerational stability effect of negative social sanctions was statistically significant (B = 0.330) and withinwave effects of negative social sanctions on negative self-feelings were appreciable and significant (B = 0.515 for G1 and B = 0.426 for G2). Also as expected, the intergenerational stability coefficient for negative selffeelings was attenuated (B = 0.132) relative to the baseline coefficient observed prior to specifying the intragenerational effects of negative social sanctions on self-feelings and the intergenerational stability of social sanctions (B = 0.228). Thus, these specifications in part account for the intergenerational parallelism of negative self-feelings. In addition, the results suggest support for the theoretical premises that informed the structural equations model.
Summary In chapter 3, results were reported that demonstrated the conditional nature of intergenerational parallelism. The strength of the association between G1 and G2 phenomenon (deviant behavior or one or another of its theoretically informed correlates) is contingent on specified
Intergenerational Continuity of Intragenerational Causes
191
circumstances. However, whether such contingencies are specified the linear association between the G1 and G2 phenomena remains to be explained. Is the intergenerational continuity, explainable in terms of intervening processes such that the G1 phenomenon has consequences (mediating variables) that in turn influence the G2 phenomenon? The analyses conducted in chapter 4 suggest that part of the explanation of intergenerational continuity might well be found in intervening processes; or is the explanation of intergenerational parallelism to be found in the intergenerational continuity of common antecedents that are themselves intergenerationally continuous? Regarding the latter possibility, the analyses in the present chapter clearly suggest that part of the explanation of instances of intergenerational parallelism implicates intergenerational stability of common antecedents. Thus, observed intergenerational parallelism in deviant behavior in varying degrees was explained by the intergenerational stability of disposition to deviance, being the object of negative social sanctions and interpersonal rejection, participation in a network of deviant friends, and the common influence of these constructs on deviant behavior in both G1 and G2. Disposition to deviance, in turn, was explainable in large part by the intergenerational stability in the experience of negative self-feelings, being the object of negative social sanctions and interpersonal rejection, participating in a deviant friendship network, and the common influence of these constructs on negative self-feelings in both G1 and G2. Finally, negative self-feelings was explainable, in part, in terms of the stability of being the object of negative social sanctions and social rejection; that is, in each instance, the degree of interpersonal parallelism in the phenomenon of interest observed at baseline was substantially reduced when the intergenerational stability of the explanatory factor as well as the intragenerational influence of the explanatory factor in both G1 and G2 were specified. It now remains to consider the implications of the findings concerning mediating, moderating, and common antecedent constructs as explanations of instances of observed intergenerational parallelism within the context of the general theory of deviant behavior that informed the analyses. This will be the subject matter of Part IV.
Part IV RETROSPECT AND PROSPECT
In Part IV, the theoretical and methodological rationales are reviewed, as are the substantive findings with regard to their implications for a future research agenda. In chapter 7, Summary and Conclusions, the general logic of procedure that informed the organization of this volume is considered as a methodological template, not only for future research on intergenerational parallelism of deviance but also for in-depth analysis of any essentially bivariate relationship. Additionally, the findings relating to the moderating, mediating, and common explanatory factors are considered with regard to their complementarity and as offering validation of he general theory of deviant behavior that informed the analyses. In the course of the discussions, limitations of the analyses and lacunae in the theoretical formulations and substantive literature are highlighted and the implications of these for informing future research agendas are drawn.
193
7 Summary and Conclusions
Two tasks remain to be accomplished: to summarize the substantive findings on intergenerational parallelism within the context of the general theory of deviant behavior and, to highlight future elaborations that might obviate what might be perceived as limitations of the present study.
Substantive Findings Guided by a general theory of deviant behavior, a series of analyses was directed to the understanding of the intergenerational parallelism of deviant behavior. The methodology was informed by the conceptualization of intergenerational parallelism of deviance as a bivariate relationship between first-generation (G1) adolescent deviance and second-generation (G2) adolescent deviance that was explained/elaborated by specifying moderators, mediators, and intergenerationally continuous common antecedents of the G1 deviance–G2 deviance relationship. In addition to explaining the bivariate relationship between G1 deviant behavior and G2 deviant behavior, the analyses were directed toward explaining the bivariate relationships between G1 and G2 disposition to deviance and between G1 and G2 negative self-feelings. This was done (1) because on theoretical grounds, intergenerational parallelism in the two variables is implicated in the explanation of intergenerational parallelism in deviant behavior and (2) to demonstrate that explanations of the G1 deviant behavior–G2 deviant behavior relationship are themselves 195
196
Chapter 7
explainable in terms of mediating, moderating, and intergenerationally continuous antecedent constructs. According to the general theory, deviant behavior is motivated. Either (1) chronic experiences of failure and rejection in conventional membership groups lead to negative self-feelings that dispose the youth to lose motivation to conform to conventional expectations while seeking deviant alternatives that might assuage the negative self-feelings secondary to the experiences of failure and rejection or (2) the youth is motivated to perform deviant acts because the person perceives these acts as required by the (deviant) membership group, and so are performed in order to evoke positive responses from self and others. Given the central roles played by these constructs, the review of salient findings will emphasize the explanatory role of intergenerational parallelism in these constructs in accounting for the G1 deviance–G2 deviance relationship and the explanation of the intergenerational parallelism of negative selffeelings and disposition to deviance, although other salient findings will be discussed as well.
Deviant Behavior Consistent with other studies of intergenerational parallelism of deviance, G1 adolescent deviance was associated with G2 adolescent deviant behavior at a statistically significant but modest level (B = 0.239). The magnitude of the intergenerational parallelism in deviance is a function of a number of theoretically indicated conditions, which relate respectively to the G1 intragenerational stability of deviant behavior over the life course (particularly until the child-rearing years) and to the influence of the G1 parent on the G2 youth. The conditional G1 intragenerational stability of deviant behavior between early adolescence and the parenting years is explainable in terms of intervening processes and intergenerational continuities of common antecedents of G1 and G2 deviance. Moderating Variables
The magnitude of intergenerational parallelism in deviance is moderated by theoretically specified conditions that reflect alienation from the conventional world and is associated with the disposition to deviance (i.e., the loss of motivation to conform to and the acquisition of motivation to
Summary and Conclusions
197
deviate from conventional expectations). Alienation from the conventional socionormative system by G1 youths was reflected in latent constructs indicating rejection of conventional values or socionormative standards, perceptions of being rejected by conventional others, and emotional withdrawal from conventional others who otherwise might have served as agents of social control over the acting out of deviant dispositions by G1 youths. Under conditions such as these, the G1 youths were expected to maintain or increase levels of deviant behavior into the parenting years. Deviant behavior by the G1 parents, in turn, was expected to influence deviant behavior by G2 youths by providing an environment that was supportive of deviant behavior whether through modeling such behavior or by failing to effectively control association with deviant peers. As expected, under conditions reflecting G1 alienation from conventional socionormative systems (that presumably facilitated intragenerational continuity of deviant behavior) and G2 perceptions of a supportive deviant social environment (that was facilitated by G1 parental deviance), intergenerational parallelism in deviant behavior was substantially greater. In support of the general theory, for G1 subjects who scored higher on latent constructs reflecting alienation from the conventional socionormative environment (disposition to deviance) compared to those who scored lower on the constructs and for G2 subjects who scored higher on a construct that reflected a supportive deviant social environment compared to those who scored lower, the magnitude of intergenerational parallelism of deviant behavior was substantially greater (approximately twice as great). Intervening Processes
Consistent with the guiding theoretical framework, observed intergenerational parallelism of deviant behavior was partially explained in terms of intervening processes that relate to continuity of G1 deviant behavior between adolescence and the G1 parenting years and to the influence of the G1 parent on the G2 youth’s deviant behavior. The theoretically informed processes are said to be reflected in the G1 young adult socioeconomic status, G1 young adult distressful emotions, and G2 perceptions of the G1 parent’s religiosity. As expected, the specification of the intervening constructs partially explained (decomposed) the observed intergenerational parallelism of deviant behavior. The intervening constructs reflect the theoretical processes that informed the analyses.
198
Chapter 7
Deviant behavior by the G1 youth had the interrelated consequences of: evoking negative social sanctions that stigmatize and deprive the G1 youth of legitimate resources that are necessary for the youth’s need satisfaction, alienation from the conventional world that is secondary to the experience of stigmatization-induced distressful self-feelings, and the weakening of social controls that might have constrained the acting out of deviant motivations—all of which maintain the stability of G1 deviant behavior over the life course into the parenting years. These constructs reflect, also, the consequences of the parent’s deviance for G2 youth, including deprivation of resources, experiences of stigmatization and abusive parenting, resulting distressful self-feelings, alienation from the conventional socionormative system, and concomitant weakening of social controls that might have constrained the acting out of the G2 youth’s deviant impulses. Common Antecedents
In addition to explaining intergenerational parallelism in deviant behavior in terms of intervening processes, intergenerational parallelism in deviance is explainable in terms of the intergenerational parallelism of constructs that within each generation influence deviant behavior. The specification of intergenerational effects of each construct on deviant behavior along with the specification of intergenerational continuity of the construct in large measure explain the observed intergenerational parallelism of deviance; that is, a consequence of these specifications is the appreciable attenuation of the magnitude of the observed intergenerational parallelism of deviant behavior. Although on theoretical grounds, a number of common constructs were specified (negative social sanctions, deviant peer associations, experiences of social rejection), these are interpretable as antecedents of the most proximate of the common antecedents, namely disposition to deviance. A major influence on both G1 and G2 deviance is the disposition to engage in deviant behavior as is reflected in the loss of motivation to conform to conventional normative expectations and the genesis of motivation to engage in deviant adaptations as alternative responses to distressful experiences of failure and rejection associated with conventional membership experiences. The deviant adaptations have their own consequences (evoking negative social sanctions, associating with deviant peers, consequent social rejection) that exacerbate alienation from the
Summary and Conclusions
199
conventional normative system and increase the disposition to engage in deviant adaptations to life stress. The disposition to engage in deviant adaptations in each generation, in the presence of facilitating circumstances, is expected to increase the probability of actually engaging in the deviant behavior. This is, in part, a consequence of the weakening of conventional social controls that is a consequence of the youth’s alienation from the conventional normative system. The hypothesized intergenerational continuity of disposition to deviance is based on the theoretical premise that G1 youths, acting out their deviant dispositions, would evoke negative social sanctions and consequent distressful negative self-feelings and intensified feelings of alienation from conventional normative systems throughout the life course. Associated experiences of social rejection would increase interaction with deviant others, one manifestation of which might be inappropriate (deviant) mate selection. This circumstance along with the chronic experience of distressful self-feelings and the limited opportunity to be socialized in normative parenting patterns increased the likelihood of the G1 parent engaging in neglectful or abusive parenting and modeling deviant and maladaptive coping patterns for the G2 youth. These outcomes, together with the paucity of resources available to the G2 youth, and the stigmatizing and otherwise distressful circumstances associated with a conflictual and unconventional home life dispose the G2 youth to adopt deviant coping patterns for dealing with life stress, a disposition that increases the likelihood of the G2 youth actually engaging in deviant behavior. Consistent with the theoretical premises, the specification of intragenerational effects of disposition to deviate on deviant behavior and of intergenerational continuity of disposition to deviance in large measure explains (decomposes) the observed intergenerational parallelism of deviance. Given the strong explanatory value of disposition to deviance in accounting for intergenerational parallelism in deviant behavior, the theoretical premises underlying this expected effect are, to some degree, validated.
Disposition to Deviance Intergenerational parallelism in disposition to deviance, which on theoretical grounds is expected to play an important role in explaining
200
Chapter 7
intergenerational parallelism in deviant behavior, according to the same general theory would be expected to be contingent on specified moderating circumstances and to be decomposed (explained) in terms of specifying intervening processes and intergenerational continuity of common antecedents. Moderating Variables
Intergenerational stability in disposition to deviance is contingent on circumstances that facilitate (1) intragenerational stability of G1 disposition to deviance between youth and the parenting years and (2) intergenerational influences of the G1 parents on G2 youths. The results in general were supportive of the theoretical premises. The magnitude of intergenerational parallelism in disposition to deviance was several times greater under conditions that reflected facilitation of G1 intragenerational continuity of disposition to deviance and the influence of G1 on G2 disposition to deviance. Deviant dispositions were expected to be continuous throughout the life course under conditions where the youths characterized by such dispositions did not experience discomfort in their social relationships, whether this is reflected in rejection by others, new expectations incumbent upon one, thwarted aspirations, or distress related to failure to meet the expectations of others. Youths who were comfortable with who they were and their current situation in life would have no need to change their disposition to behave, nor would they be motivated to change if significant others were perceived as comfortable with the youths and deviant dispositions. However, discomfort with oneself (i.e, negative self-evaluations) would instigate changes in attitudes, including those related to deviant dispositions. Given intragenerational stability in G1 deviant dispositions between early adolescence and adulthood, intergenerational parallelism of deviant dispositions depends on circumstances that variously facilitate or impede communication of the antisocial disposition from the G1 adult to the G2 youth. Thus, the G2 youth would be more likely to identify with and adopt the parent’s antisocial attitudes if the youth held more, rather than less, favorable attitudes toward the parent. Congruent with the theoretical expectations, the results suggest that intergenerational continuity in deviant dispositions is conditional on three sets of moderators: social support for antisocial dispositions (suggested
Summary and Conclusions
201
by friendship variables reported by the G1 adults), stability in life circumstances (reflected in not being promoted at work, and not feeling unsure of oneself when thinking about one’s social class, reported by G1 adults), and close relations with parents (reflected in G2 reports of finding it easy to discuss problems with parents). Under conditions where peer support for deviant attitudes appears to exist (suggesting a deviant subculture), intragenerational stability in life circumstances is manifest, and a good relationship exists between G1 parents and G2 youths, the degree of intergenerational continuity in deviant dispositions tends to be relatively strong. Where peer support for deviant dispositions appears to be lacking, changing life circumstances and reference groups appears to be likely, and parent-child relations are strained, the degree of intergenerational parallelism is relatively weak. Intervening Processes
Congruent with theoretical expectations, indexes of structural (socioeconomic) disadvantage and familial conflict were observed to mediate (and, so, explain in part) the observed intergenerational parallelism of disposition to deviance. The expectations regarding the mediating role of socioeconomic status were based on the reasoning that the G1 youth’s deviant attitudes would evoke negative sanctions, including deprivation of needed resources. The experience of such deprivation and other negative sanctions would result in alienation from the conventional socionormative structure. Such alienative attitudes would be reflected in loss of motivation to conform to (among other norms) expectations regarding positive valuation of and striving for upward social mobility. The loss of motivation to pursue conventional values, along with the impediments to upward social mobility that accompany or reflect negative social sanctions increase the likelihood that the G1 youth will have a disadvantageous position in the conventional system of stratification. Undesirable placement in the system of stratification, in turn, has consequences, including the interrelated experiences of negative self-feelings associated with a paucity of material and interpersonal resources and consequent alienative attitudes that together hinder the G2 youth’s ability to forestall stress, gain social acceptance, and achieve success. The alienation of the G2 youth that is congruent to these outcomes, in the absence of learned effective conventional coping resources, is disposed to seek alternative deviant adaptations through which the G2 youth can satisfy his needs.
202
Chapter 7
The expectation that the familial conflict would mediate intergenerational parallelism of disposition to deviance, similarly, is predicated on a number of theoretical linkages. G1 disposition to deviance would be expected to eventuate in selecting deviant social networks, learning maladaptive coping mechanisms, and being deprived of instruction (and motivation in) conforming to conventional role definitions, including the roles of spouse and parent. The conflictful family relations that result from these outcomes throughout the G1 life course, in turn, affect the G2 youth’s own disposition to engage in conflictful relations. Family conflict serves as a model for the G2 youth’s maladaptive and antisocial attitudes. Further, the parent-child conflicts are intrinsically distressing situations that evoke attempts to forestall or assuage the distress. Because the G2 youth has learned deviant and maladaptive coping patterns rather than conventional and effective response patterns in the dysfunctional family, the G2 youth will be disposed to respond to family stress with similarly dysfunctional patterns. In general, the estimation of the models provided results that validated the general theory that informed the models. Common Antecedents
Intergenerational parallelism in disposition to deviance is expected to be explained, in part, by the intergenerational continuity of common antecedents of G1 and G2 disposition to deviate. On theoretical grounds, these common antecedents are expected to encompass negative self-feelings (and its antecedents, including negative social sanctions and perceived social rejection) and membership in deviant friendship networks. With regard to the latter common antecedent, whether participation in deviant friendship networks was the outcome of (1) a deviant adaptations to conventional social rejection, (2) being born into an environment in which the “deviant” patterns were in fact normative for that group, or (3) being the object of negative social sanctions that facilitated interaction among deviant peers, deviant friendships, on theoretical grounds, would be expected to dispose the youth to conform in order to secure deviant friends’ approval, the deviant peers would provide resources and occasions for engaging in deviant behavior, and the friendship network would provide rationalizations for otherwise conventional youths to adopt deviant responses. The participation in a deviant network in one generation should have consequences for the participation in a deviant network in the next
Summary and Conclusions
203
generation. These consequences include the evocation of negative social sanctions, selection of deviant mates, exclusion from conventional socialization into conventional spousal and parental roles, abusive and neglectful parenting, and the resultant experience of negative self-feelings and alienation from conventional groups on the part of the G2 youths, all of which facilitates participation in deviant memberships. Regarding intergenerational continuity of negative self-feelings as a common antecedent of disposition to deviance in each generation, negative self-feelings (stemming from formal and informal negative social sanctions and in response to the early deviant adaptations to being the object of negative sanctions), in the absence of conventional alternatives, dispose the youth to lose motivation to conform and to seek deviant responses as alternative modes of coping with life stress. The experience of distressful self-feelings in any generation has consequences that increase the likelihood of experiencing negative selffeelings in the next generation. Distressful self-feelings experienced by G1 youths deriving from failure and rejection in conventional membership groups dispose the G1 youths to adopt deviant patterns as the only viable alternatives to coping with life stress. Consequent social exclusion, deprivation, and unconventional adaptations increase the likelihood of forming conflictual and otherwise dysfunctional G1 family environments for the rearing of the G2 youth. The ensuing G2 experience of neglectful and otherwise abusive parenting, along with the socialization into ineffective and maladaptive coping patterns predispose the G2 youth to experience distressful negative self-feelings. Such distressful negative self-feelings, along with the recognition of their source alienate the G2 youth from the conventional environment and dispose the youth to adopt deviant patterns. Consistent with these theoretical premises, the model specifying intragenerational effects of negative self-feelings on disposition to deviance for both G1 and G2 youths and the intergenerational continuity of negative self-feelings between G1 and G2 in large measure explained the observed intergenerational parallelism of disposition to deviance.
Negative Self-Feelings The intergenerational parallelism of negative self-feelings, which, together with its intragenerational effects, in large measure explains
204
Chapter 7
intergenerational parallelism of disposition to deviance, has its own moderators and is explainable in terms of intervening processes and intergenerational continuity of common antecedents.
Moderating Variables
The results of estimating several structural equation models are congruent with the theoretical premises that informed the analyses. The strength of the intergenerational parallelism in negative self-feelings that is observed is contingent on circumstances that theoretically facilitate continuity of negative self-feelings between youth and parenthood for the G1 subjects and that facilitate the influence of the G1 parents on the G2 youths. The continuity of negative self-feelings between early adolescence and parenthood for the G1 subjects is facilitated by the early performance of stress-inducing behavior and its correlates (deviant behavior and consequent negative social sanctions), alienation from potential sources of social support that might forestall or assuage distress (not important what parents, teachers, kids at school think of me), consistent parental socialization in inadequate coping patterns (evidenced by the combination of negative self-feelings and parental agreement on patterns of child rearing), and the experience of life events that challenge inadequate coping patterns (evidenced by the combination of higher levels of negative self-feelings and life events reported in young adulthood such as marriage/cohabitation and the death of a significant other). The intergenerational transmission of negative self-feelings from G1 parents to G2 youths is conditional on the G2 youth’s experiences of rejection (by peers) and failure (no expectations of upward social mobility), absence of supportive parents who might have helped forestall stress and mitigate stressful negative self-feelings (G2 youth does not show affection or feel close to parents and does not receive positive reinforcement), and G2 youth perceptions of G1 parental use of ineffective and maladaptive coping mechanisms (substance use). Under each of these conditions, the magnitude of observed intergenerational parallelism in negative self-feelings is substantially greater than under mutually exclusive conditions, thus lending support to the general theory that informed the analyses.
Summary and Conclusions
205
Intervening Processes
The estimation of structural equation models produced results that are compatible with the theoretical premises that informed the analyses. The intergenerational parallelism of negative self-feelings is accounted for, in part, by the intervening processes related to young adult stressful emotional experiences and family conflict during the child-rearing years. These mediating variables are said to reflect, or incur at later stages of the life course, the experience of distressful life circumstance of failure and rejection, the lack of effective conventional coping mechanisms, and dysfunctional role performance in several social relational contexts. These outcomes, in turn, either constitute stressful life circumstances for the G2 youth or reflect the lack of effective conventional coping patterns that might otherwise have functioned to forestall or assuage the negative selffeelings associated with self-threatening life circumstances.
Common Antecedents
Intergenerational continuity in negative self-feelings, which influences disposition to deviance in each generation, in part accounts for intergenerational parallelism in disposition to deviance and so indirectly explains, in part, intergenerational continuity of deviant behavior. Intergenerational continuity in negative self-feelings is itself explainable partly in terms of intergenerational stability of constructs that have intragenerational effects on negative self-feelings in both G1 and G2. The constructs in question relate to social rejection and being the object of negative social sanctions. The results support the theoretical premises that informed the analyses. With regard to social rejection, in the course of normal socialization processes youths learn to need the approval of others, particularly those who are in a position to satisfy or frustrate basic socioemotional or instrumental needs. Within each generation, the inability to evoke approval and the evocation of rejecting attitudes from significant others evoke negative self-feelings. The experience of social rejection by youths in one generation has consequences that lead to the experience of social rejection by youths in the next generation. The youth’s experience of social rejection and consequent distressful self-feelings leads to alienation from the rejecting groups and readiness to adopt deviant responses that offer greater promise than
206
Chapter 7
conventional responses of forestalling or assuaging the distress that accompanies social rejection. Further, the deviant responses evoke social rejection and exclusion from normal opportunities to learn conventional role behavior as a parent. This factor along with the experience of negative self-feelings eventuate in abusive and neglectful parenting behavior and the modeling of ineffective coping mechanisms for forestalling or assuaging life stress. The combination of learning maladaptive coping mechanisms, stigmatizing experiences of membership in deviant family contexts, and being the object of neglectful or abusive parenting increase the likelihood that the G2 youth will parallel the G1 youth in being the object of rejecting responses by conventional others and, as a consequence, experiencing negative self-attitudes. With regard to negative social sanctions, objectionable traits and behavior evoke negative social sanctions and consequent negative selffeelings in each generation. Being the object of negative social sanctions evokes negative self-feelings because the youth values approving responses and disvalues rejecting responses; being the object of negative sanctions deprives the youth of access to valued social resources and has other outcomes that increase the likelihood of future experiences of rejection and failure that evoke negative self-feelings. The experience of negative social sanctions by G1 youths has consequences that increase the likelihood that G2 youths, as well, will be the object of negative social sanctions. As objects of negative social sanctions, G1 youths experience stigmatization and deprivation of valued social resources and consequent feelings of alienation from conventional groups and disposition to adopt alternative deviant coping patterns. These circumstances, continuing into the parenting years, increase the likelihood that the G1 parent will transmit the disposition to deviance to the G2 youth and will occasion distressful experiences that give G2 youth occasion to perform the deviant acts. Deviant coping response by the G2 youth, in turn, evokes negative social sanctions, thus paralleling the experience of the G1 youths. In the second generation, being the object of negative social sanctions, in turn, evokes negative self-feelings, operating through the same intervening mechanisms that were described for the G1 youths. In sum, although many of the intervening linkages remain to be tested, the models estimated are compatible with the general theory of deviant behavior that informed the analyses. Future specifications of moderating variables, intervening processes, and intergenerationally
Summary and Conclusions
207
continuous common antecedents will provide further opportunities for validation and elaboration of the general theory of deviant behavior.
Future Elaboration The theoretically informed empirical analyses reported in this volume derive from a research program that has a number of strengths, characterized by the use of a panel design, multigeneration measurement of successive biologically related panels at a comparable developmental stage, specification of mediating variables and contextual continuity, and consideration of the potential moderating influence on intergenerational parallelism (and on intervening causal linkages) of a range of variables. Many of these features that characterize our own studies approximate those that are thought to be ideal for the study of intergenerational transmission of, for example, parenting (Van Ijzendoorn, 1992): These studies should now begin to incorporate designs fitted to the goal of describing intergenerational transmission of parenting: longitudinal studies should be carried out, measuring parenting with comparable instruments at comparable times across the lifespan. Furthermore, contextual factors should be taken into account because the transmission may be stronger or weaker depending upon the influence of these contextual factors on two or three generations. (p. 97)
In addition, other design limitations (small sample sizes, use of clinical samples rather than in-community samples) are obviated by the use of a large general-population sample. Nevertheless, the conduct of the study also implicates a number of features that, arguably, are interpretable as limitations.
Sociodemographic Controls It is reasonable to anticipate, given the knee-jerk tendency in the social sciences to control certain sociodemographic variables such as gender, race, and socioeconomic status, that these analyses would be subject to criticism for failing to have done so. However, these variables (review of the literature would suggest) in themselves add little to our understanding of the mechanisms accounting for intergenerational parallelism whether they are treated as moderating or exogenous variables
208
Chapter 7
(although in the case of socioeconomic status, it seemed reasonable to model it as a mediating variables). Rather, it is through the more proximal influences exercised by correlates of these “background” variables as moderating, mediating, or common antecedents factors that insight into the processes underlying the observation of intergenerational parallelism is gained. The theoretically informed models, therefore, focus on the correlates of the sociodemographic variables that are more proximate to the mechanisms explaining instances of intergenerational parallelism.
Bidirectional Effects In attempting to assess the factors that intervene in intergenerational parallelism, such as parental distress or parenting patterns, the assumption is made that these variables are causally prior to the deviant patterns displayed by the G2 youths. However, these relationships do not necessarily reflect unidirectional causal effects. It is quite plausible to assert that characteristics of the child influence parental responses as it is to assert the interpersonal influence of parents on children. Thus, for example, the concurrent measurement of psychological dysfunction in the parent and antisocial character in the child does not permit easy separation of the causal relationships (Loeber & Stouthamer-Loeber, 1986) Parental rejection may drive children to rebel. However, it is also difficult to love children that make one’s life miserable. Thus, parental rejection can be both cause and consequence of children’s behavior. (p. 54).
This observation has not been obviated with the passage of time. A decade later, it might be observed that a weakness of many intergenerational studies is the failure to specify what is likely an existing mutual influence between putative intragenerational “causes” of the variable of interest and the tendency for the variable of interest to evoke or stimulate the so-called causal effect. It has been argued, for example, in the case of the relationship between antisocial behavior and discipline that the relation should be modeled as a mutual one in which parental discipline affects antisocial behavior, and antisocial behavior evokes parental influence (Ge et al., 1996). Ten years later, the present study has been unable to consistently attend to the modeling of bidirectional effects.
Summary and Conclusions
209
Meaning of Measurement Models By testing two panels of subjects a generation apart it is possible that the measures used for the G1 subjects are no longer valid indicators of the latent constructs under consideration. This is a special case of the situation that exists even in studies in which data are collected at a single point in time; that is, it is possible that subgroups of the population reflect a psychosocial phenomena in different ways. Males, for example, might reflect alienation from conventional society with different behaviors than females. It is possible that certain behaviors that are considered deviant for females are considered acceptable for males. Hence, these indicators might not be appropriate measures of alienation for males, but they remain valid indicators of alienation for females. Similarly, subcultural variation in ethnic groups might preclude one ethnic group from expressing selfderogation with what are face-valid indicators of self-derogation for other groups. One ethnic group might by normative consent be precluded from admitting to self-derogatory feelings, whereas another group might be permitted to express such feelings when self-derogation is truly present; or one ethnic group migh be required, as a matter of good form, to depreciate their personal qualities. In such groups, the expression of selfdeprecatory feelings would not be valid indicators of self-derogation. One must always be wary of the possibility that measurement variables do not reflect the same underlying meanings across time and in different gender, ethnic, racial, or socioeconomic groupings. Even where measurement items cohere, the internally reliable scale might reflect different constructs. Therefore, all analyses should examine the internal reliability and construct validity of measures for groups differentiated in terms of gender, race/ethnicity, and socioeconomic status at different times. Where measurement models over time and space vary across these sociodemographic groupings, the decision will have to be made as to whether the functionally equivalent but metrically dissonant measurement models reflect the same underlying theoretical construct. If they do, the models will be estimated for different groups using functionally equivalent but metrically different indicators. This issue has been all but ignored in the literature. It is necessary to begin thinking of how one deals with the problem of temporal and sociocultural differences in reflections of underlying constructs. Among the relatively rare instances in which this issue has been considered, Baumrind (1994) observed with reference to the meaning of specific forms of punishment:
210
Chapter 7 Strict discipline and the use of corporal punishment do not per se constitute child abuse. The high value placed on obedience and respect for authority by African-American families that results at times in the use of corporal punishment can be traced to African tradition (Kohn, 1977; Peters, 1976; Young, 1970). African-American parents claim to use hitting as a teaching method with young children to improve behavior, to teach respect, obedience, and right from wrong, and to deal with children whose language is limited. Some African-American parents use coercive tactics strategically to force the aggressiveness and guardedness needed for African-American children to survive in hostile environments (McLoyd, 1990; Ogbu, 1981). (p. 362)
Genetic Influences Our study shares with many other studies of intergenerational transmission of deviant outcomes a limitation that relates to the difficulty in assessing genetic influences on intergenerational continuity. Referring to a group of studies on intergenerational continuity and the transfer of risk, Serbin and Stack (1998) made observations that are appropriate for our study as well: Conceptually, these projects were not designed for examining genetic transmission of risk or for studying nature–nurture issues. All of these studies deal with prediction of behaviors that are likely to have a complex and interactive set of causes, including genetic, social–experimental, cultural, and contextual factors. However, the sample sizes in these projects and information about family histories are typically too limited to extract genetically relevant information, at least by using conventional genetic research designs. With improvements in technology for studying genetic profiles and markers, however, examination of parent–child similarities in genetic patterns is likely to be added to ongoing studies in the near future. In fact, the availability of information about parents’ behavior at earlier points in time may make it possible to use these data sets to discover valuable information about the genetic basis of continuity in human behavior, both within and between generations. (p. 1160)
In the present study, residual intergenerational effects, after specifying mediating variables and intragenerational effects of intergenerationally continuous causal variables, are often interpretable in terms of social learning influences. However, these residual effects are also interpretable in terms of genetic inheritance. Because the intergenerational dyads are biologically related, we cannot choose between these interpretations.
Summary and Conclusions
211
Parent Reports Another limitation of our study, one that is shared by numerous other studies of intergenerational continuity (Serbin & Stack, 1998) concerns the assessment of intergenerational continuity using data from only one of the G2 child’s parent’s childhood experiences. As is typical of studies of this sort, only one of the parents provided data at the time of their youthful experiences. As Serbin and Stack (1998) observed: This may lead to an underestimation of intergenerational continuity, because half of the equation is missing, so to speak. Solutions to this problem may be found in future inclusion of spousal and extended-family data, as well as the inclusion of entire sibships, rather than a single child from each family. (p. 1161)
Sociohistorical Trends To what extent do sociohistorical trends influence the degree of intergenerational comparability? Are 12- and 13-year-olds in 1970 equally likely to engage in deviant acts as their counterparts in 1995? Are the causes of deviant behavior in 1970 different from the causes of deviant behavior in 1995? Are changes in the correlates of deviant behavior or in the degree of deviant behavior over a 25-year period due to changes in social conditions over that period in time? Period effects are said to “reflect the impact of sociohistorical events as they produce change and continuity in different social groups (i.e., social generations) at different points in time” (Bengtson, 1987, p. 452). Substantial social changes have occurred over the years between the times that the first- and second-generation youths were interviewed during their early adolescent years. To the extent that these social changes have some impact on the circumstances in which the individuals were reared, we would expect lesser degrees of parallelism to be observed than in circumstances where the social context in which the youths were reared were similar. A case in point is the changes in behaviors and attitudes relating to working mothers. Regarding behavior change Rindfuss, Brewster, and Kavee (1996) stated: A substantial change has occurred in the childrearing practices of American families with young children. They have moved from a childrearing regimen
212
Chapter 7 under which most preschool-age children were cared for in their own homes by their mothers to a more diverse system in which the majority of mothers of preschoolers are in the labor force and their children are cared for in a variety of settings by assorted caregivers. While part of this shift is the result of more children being in female-headed families, the overwhelming portion is explained by the changing labor force behavior of married women. (pp. 476–477)
Parallel attitudinal changes have been observed as well (Rindfuss et al., 1996) particularly between 1970 and 1985: Not only has there been a substantial decline in the percent of respondents thinking that preschool children will suffer if their mother work, but this shift occurred across all social subgroups and all ages within American society. (pp. 478–479)
We have yet to examine secular trends over the generations that might account for the presence or absence of intergenerational parallelism. Intergenerational trends in deviant behavior, societal responses to deviant behavior, and related phenomena might attenuate or exacerbate intergenerational parallelism in antisocial character. In the absence of a cohort-sequential design, it is not possible to evaluate the degree to which the findings relating to intergenerational parallelism (or their absence) are due to patterns of sociocultural change. During periods of social stability, intergenerational transmission becomes a major mechanism for maintaining the stability of attitudes. The mass media and contemporary peers will tend to reinforce intergenerationally transmitted attitudes and behaviors. In times of rapid social change, however, the inevitable intergenerational influence will be attenuated by the transmission of the nature and acceptability of novel ideas in the mass media and by heterogeneous peers who model such novel ideas and behaviors. The interpretation of our findings must take such circumstances into account.
Range of Variables The models that were specified and estimated in chapters 3–6 implicated a large number of theoretical constructs. However, each relationship was predicated on a number of theoretical suppositions that remain to be tested by modeling the constructs that reflect the theoretical assumptions. It still remains to specify what other common antecedents or intervening processes explain the intergenerational relationship in deviant behavior
Summary and Conclusions
213
between early adolescents who ultimately will become parents of the adolescents assessed in the second generation at the same developmental stage and the G2 adolescents themselves. With regard to common antecedents, intergenerational parallelism in one form of deviant behavior might reflect a common circumstance to which the members of the respective generations adapt in like fashion. A variety of common circumstances might be summarized, for example, in the general experience of psychological dysfunction. Continuity in psychological dysfunction across the generations, whether due to common experiences or not, might lead to common deviant adaptations. Regarding intervening processes, understanding intergenerational transmission requires the ever-more precise specification of mediating mechanisms. As Velleman (1992) observed: ...in seeking to account for the intergenerational transmission of alcohol problems, researchers have suggested an explanation couched in terms of marital and family problems. Yet explaining the transmission of problems by simply citing such factors as violence, parental conflict, parental loss, and parental inconsistency is in itself no explanation, for the same question can be re-asked about these questions: why should these factors lead to a greater incidence of problems. (p. 382)
Further, before we approximate a fuller explanatory model accounting for intergenerational parallelism in deviant behavior and the mechanisms that account for such continuity, we must initiate further investigation of moderating constructs that condition the relationships between the transgenerational characteristics and the variables that represent common antecedents or mediating influences in the crossgenerational relationships. Indeed, we have yet to examine fully the intergenerational parallelisms in deviant behaviors. Both on empirical and theoretical grounds, we might expect that the degree of intergenerational continuity and the factors that account for such continuity will vary according to any of a number of factors. When we have satisfied ourselves that such variables indeed moderate the processes accounting for intergenerational parallelism, we must further seek out the true meaning of these constructs in terms of their correlates that are more proximal moderators of causal linkages. These remarks apply not only to the investigation of intergenerational parallelism in deviant behavior (and disposition to deviance and negative self-feelings) but also to the mediators, moderators, and common antecedents of these instances of intergenerational parallelism. As
214
Chapter 7
reference to Figure 1.1 will remind us, each moderator, mediator, and intergenerationally continuous common antecedent has its own mediators, moderators, and intergenerationally continuous common antecedents.
Inclusive Models In examining the explanatory power of theoretically indicated mediating, moderating, and intergenerationally continuous common antecedent variables in accounting for intergenerational parallelism in deviant behavior, disposition to deviance, and negative self-feelings, the models were estimated seriatim. Thus, it is possible that the several moderators (or mediators or intergenerationally continuous common antecedents) were redundant indicators of a common latent construct or were linearly related to each other. The simultaneous estimation of all of these effects challenges the capabilities of this dataset. The specification of inclusive models that simultaneously specify the several theoretically informed mediating, moderating, and intergenerationally continuous common antecedents of deviant behavior and its correlates remains to be accomplished.
Conclusion This volume is viewed as accomplishing the following objectives. First, it has presented an overview of the methodologies and substantive findings of empirical studies that investigate instances of intergenerational parallelism. Second, it has offered a template for a systematic logic of procedure for analyses of instances of intergenerational parallelism (and, more generally, for the analysis of any bivariate relationships). The suggested methodology for analyses and elaboration of the bivariate relationships between G1 deviant behavior and G2 deviant behavior implicates the systematic specification of the moderators, mediators, and intergenerationally continuous common antecedents of the intergenerational parallelism of deviant behavior and its correlates. Third, the volume offers a guiding theoretical framework that accommodates the general literature and informs the empirical findings reported herein. Finally, the results of a series of analyses are reported that are informed by the general theory of deviant behavior. The models that are
Summary and Conclusions
215
estimated specify the moderators, mediators, and intergenerationally continuous common antecedents of the relationship between G1 and G2 deviance (and of the intergenerational parallelism of two correlates of the deviance). The findings collectively confirm and complement those reported in the general literature; and by being congruent with theoretically informed expectations, the findings lend support to the general theory. The analyses reported in this volume are presented as flawed but provocative contributions to a research agenda that aims at describing and understanding intergenerational parallelism in deviant adaptations. In the process of producing these findings, we have followed a logic of procedure that systematically specifies (1) the conditions under which intergenerational parallelism in deviant behavior (and its correlates) is observed and (2) the intervening and intergenerationally stable common antecedents that explain the instances of observed intergenerational parallelism. The systematic adherence to this methodology and the resulting substantive findings move us closer to understanding the extent and explanation of intergenerational parallelism in deviant adaptations. However, ultimately, greater approximation to the goal of understanding intergenerational parallelism in deviant behavior requires that future studies attend to the above-specified elaborations.
References
Acock, A. C., & Bengtson, V. L. (1980). Socialization and attribution: Actual versus perceived similarity among parents and youth. Journal of Marriage and the Family, 42, 501–515. Agnew, R. (1992). Foundation for a general strain theory of crime and delinquency. Criminology, 30, 47–88. Akers, R. L. (1973). Deviant behavior: A social learning approach. Belmont, CA: Wadsworth. Akers, R. L. (2000). Criminological theories: Introduction, evaluation, and application. Los Angeles: Roxbury Publishing. Amato, P. R. (1996). Explaining the intergenerational transmission of divorce. Journal of Marriage and the Family, 58, 628–640. Amato, P. R., & Booth, A. (1991). Consequences of parental divorce and marital unhappiness for adult well-being. Social Forces, 69, 905–914. Amato, P. R., & Booth, A. (1997). A generation at risk: Growing up in an era of family upheaval. Cambridge, MA: University of Harvard Press. Amato, P. R., & DeBoer, D. D. (2001). The transmission of marital instability across generations: Relationship skills or commitment to marriage? Journal of Marriage and the Family, 63, 1038–1051. Amato, P. R., & Keith, B. (1991). Parental divorce and the well-being of children: A meta-analysis. Psychological Bulletin, 110, 26–46. Amato, P. R., & Rogers, S. J. (1997). A longitudinal study of marital problems and subsequent divorce. Journal of Marriage and the Family, 59, 612–624. Anderton, D., Tsuya, N. O., Bean, L. L., & Mineau, G. P. (1987). Intergenerational transmission of relative fertility and life course patterns. Demography, 24, 467–480. Aquilino, W. S. (1999). Two views of one relationship: Comparing parents’ and young adult children’s reports of the quality of intergenerational relations. Journal of Marriage and the Family, 61, 858–870. Atkinson, L. P., & Dodder, R. A. (1990). Differences over time and generation in sexual attitudes. International Review of Modern Sociology, 20, 193–210. Avison, W. R. (1999). The impact of mental illness on the family. In C. S. Aneshensel & J. C. Phelan (Eds.), Handbooks of sociology and social research (pp. 495–515). New York: Kluwer Academic/Plenum. Bandura, A. (1973). Aggression: A social learning analysis. Englewood Cliffs, NJ: Prentice-Hall. Bao, W.-N., Whitbeck, L. B., Hoyt, D. R., & Conger, R. D. (1999). Perceived parental acceptance as a moderator of religious transmission among adolescent boys and girls. Journal of Marriage and the Family, 61, 362–74.
217
218
References
Barber, B. K., Olsen, J. E., & Shagle, S. C. (1994). Associations between parental psychological and behavioral control and youth internalized and externalized behaviors. Child Development, 65, 1120–1136. Barber, J. S. (2000). Intergenerational influences on the entry into parenthood: Mothers’ preferences for family and nonfamily behavior. Social Forces, 79, 319–48. Barber, J. S. (2001a). Ideational influences on the transition to parenthood: Attitudes toward childbearing and competing alternatives. Social Psychology Quarterly, 64(2), 101–127. Barber, J. S. (2001b). The intergenerational transmission of age at first birth among married and unmarried men and women. Social Science Research, 30, 219–247. Barnes, G. M., & Farrel, M. (1992). Parental support and control as predictors of adolescent drinking, delinquency, and related problem behaviors. Journal of Marriage and the Family, 54, 763–776. Bartle, S. E., & Anderson, S. A. (1991). Similarity between parents’ and adolescents’ levels of individuation. Adolescence, 26, 913–924. Baumrind, D. (1966). Effects of authoritative parental control on child behavior. Child Development, 37, 887–907. Baumrind, D. (1978). Parental disciplinary practices and social competence in children. Youth and Society, 9, 239–296. Baumrind, D. (1991). The influence of parenting style on adolescent competence and substance use. Journal of Early Adolescence, 11, 56–95. Baumrind, D. (1994). The social context of child maltreatment. Family Relations, 43, 360–368. Beck, P. A., & Jennings, M. K. (1975). Parents as “middlepersons” in political socialization. Journal of Politics, 37, 83–107. Becker, H. S. (1963). Outsiders: Studies in the sociology of deviance. New York: The Free Press. Behrman, J., & Taubmann, P. (1985). Intergenerational earnings mobility in the United States: Some estimates and a test of Becker’s intergenerational endowments model. Review of Economics and Statistics, 67, 144–151. Belsky, J., & Pensky, E. (1988). Marital change across the transition to parenthood. Marriage and Family Review, 12, 133–156. Belsky, J., Youngblade, L. M., & Pensky, E. M. (1989). Childrearing history, marital quality, and maternal affect: Intergenerational transmission in a low-risk sample. Development and Psychopathology, 1, 291–304. Bengtson, V. L. (1975). Generation and family effects in value socialization. American Sociological Review, 40, 358–371. Bengtson, V. L. (1987). Parenting, grandparenting, and intergenerational continuity. In J. B. Lancaster, J. Altmann, A. S. Rossi, & L. R. Sherrod (Eds.), Parenting across the life span: Biosocial dimensions (pp. 435–457). Hawthorne, NY: Aldine Publishing. Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107, 238–246. Berkowitz, L. (1962). Aggression: A social psychological analysis. New York: McGraw-Hill. Biblarz, T. J., & Raftery, A. E. (1993). The effects of family disruption on social mobility. American Sociological Review, 58, 97–109. Biblarz, T. J., Bengtson, V. L., & Bucur, A. (1996). Social mobility across three generations. Journal of Marriage and the Family, 58, 188–200. Biblarz, T. J., Raftery, A. E., & Bucur, A. (1997). Family structure and social mobility. Social Forces, 75, 1319–1341. Billings, A. G., Kessler, M., Gomberg, C. A., & Weiner, S. (1979). Marital conflict resolution of alcoholic and nonalcoholic couples during drinking and nondrinking sessions. Journal of Studies on Alcohol, 40, 183–195. Blee, F. M., & Tickameyer, A. R. (1986). Black-white differences in mother-to-daughter transmission of sex-role attitudes. Sociological Quarterly, 28(2), 205–222.
References
219
Bollen, K. A. (1989). Structural equations with latent variables. New York: Wiley. Bond, J. B., & Harvey, C. (1991). Ethnicity and intergenerational perceptions of family solidarity. International Journal of Aging and Human Development, 33, 33–44. Borjas, G. J. (1992). Ethnic capital and intergenerational mobility. Quarterly Journal of Economics, 107, 123–150. Borjas, G. J. (1995). Ethnicity, neighborhoods, and human-capital externalities. The American Economist, 85, 365–390. Boyd, R., & Richerson, P. J. (1985). Culture and the evolutionary process. Chicago, IL: University of Chicago Press. Briar, S., & Piliavin, I. (1965). Delinquency, situational inducements, and commitment to conformity. Social Problems, 13, 35–45. Brook, J. S., Brook, D. W., Gordon, A. S., Whiteman, M., & Cohen, P. (1990). The psychosocial etiology of adolescent drug use: A family interactional approach. Genetic, Social, and General Psychology Monographs, 116, 111–126. Brook, J. S., Whiteman, M., & Zheng, L. (2002). Intergenerational transmission of risks for problem behavior. Journal of Abnormal Child Psychology, 30(1), 65–76. Brooks-Gunn, J., & Chase-Lansdale, P. L. (1995). Adolescent parenthood. In M. Bornstein (Ed.), Handbook of parenting: Vol. 3. Status and social conditions of parenting (pp. 113–149). Mahwah, NJ: Erlbaum. Brown, G. R., & Anderson, B. (1991). Psychiatric morbidity in adult inpatients with childhood histories of sexual and physical abuse. American Journal of Psychiatry, 148, 55–61. Brown, R. L. (1994). Efficacy of the indirect approach for estimating structural equation models with missing data: A comparison of five methods. Structural Equation Modeling, 1, 287–316. Bumpass, L. L., Martin, T. C., & Sweet, J. A. (1991). The impact of family background and early marital factors on marital disruption. Journal of Family Issues, 12, 22 –42. Burgess, R. L., & Youngblade, L. M. (1988). Social incompetence and the intergenerational transmission of abusive parental practices. In G. T. Hotaling, D. Finkelhor, J. Kirkpatrick, & M. A. Straus (Eds.), Family abuse and its consequence: New directions in research (pp. 38–60). Newbury Park, CA: Sage. Bussey, K., & Bandura, A. (1984). Influence of gender constancy and social power on sex-linked modeling. Journal of Personality and Social Psychology, 47, 1292–1302. Cairns, R. B., Cairns, B. D., Xie, H., Leung, M.-C., & Hearne, S. (1998). Paths across generations: Academic competence and aggressive behaviors in young mothers and their children. Developmental Psychology, 34(6), 1162–1174. Cairns, R. B., Elder, G. H., Jr., & Costello, E. J. (1996). Developmental science. New York: Cambridge University Press. Cairns, R. B., Gariepy, J. L., & Hood, K. E. (1990). Development, microevolution, and social behavior. Psychological Review, 97, 49–65. Caliso, J. A., & Milner, J. S. (1994). Childhood physical abuse, childhood social support, and adult child abuse potential. Journal of Interpersonal Violence, 9(1), 27–44. Capaldi, D. M., & Clark, S. (1998). Prospective family predictors of aggression toward female partners for atrisk young men. Developmental Psychology, 34, 1175–1188. Capaldi, D. M., Crosby, L., & Stoolmiller, M. (1996). Predicting the timing of first sexual intercourse for at-risk adolescent males. Child Development, 67, 344–359. Card, J. J. (1981). Long-term consequences for children of teenage parents. Demography, 18, 137–156. Carlin, A. S., Kemper, K., Ward, N. G., Sowell, H., Gustafson, B., & Stevens, N. (1994). The effects of differences in objective and subjective definitions of childhood abuse on estimates of its prevalence and relationship to psychopathology. Child Abuse and Neglect, 18, 393–399. Cashmore, K., & de Haas, N. (1995). Legal and social aspects of the physical punishment of children. Canberra, Australia: AGPS.
220
References
Caspi, A., Bem, S., & Elder, G. H., Jr. (1989). Continuities and consequences of interactional styles across the life course. Journal of Personality, 57, 375–406. Caspi, A., & Elder, G. H., Jr. (1988). Emergent family patterns: The intergenerationa construction of problem behavior and relationships. In R. A. Hinde & J. Stevenson-Hinde (Eds.), Relationships within families: Mutual influences (pp. 218–240). Oxford: Clarendon Press. Cavalli-Sforza, L. L., & Feldman, M. W. (1981). Cultural transmission and evolution: A quantitative approach. Princeton, NJ: Princeton University Press. Chase-Lansdale, P. L., & Brooks-Gunn, J. (1995). Escape from poverty: What makes a difference for children? New York: Cambridge University Press. Chassin, L., Pillow, D. R., Curran, P. J., Molina, B. S. G., & Barrera, J. M. (1993). Relation of parental alcoholism to early adolescent substance use: A test of three mediating mechanisms. Journal of Abnormal Psychology, 102, 3–19. Chassin, L., Presson, C. C., Todd, M., Rose, J. S., & Sherman, S. J. (1998). Maternal socialization of adolescent smoking: The intergenerational transmission of parenting and smoking. Developmental Psychology, 34(6), 1189–1201. Chen, Z.-Y., & Kaplan, H. B. (2001). Intergenerational transmission of constructive parenting. Journal of Marriage and the Family, 63, 17–31. Cherlin, A. J. (1992). Marriage, divorce, remarriage. Cambridge, MA: Harvard University Press. Cherlin, A. J., Chase-Lansdale, P. L., & McRae, C. (1998). Effects of parental divorce on mental health throughout the life course. American Sociological Review, 63, 239–249. Chodorow, N. (1978). The reproduction of mothering. Berkeley: University of California Press. Christiansen, K. O. (1977). A review of studies of criminality among twins. In S. A. Mednick & K. O. Christiansen (Eds.), Biosocial basis of criminal behavior. New York: Gardner. Clayton, S. (1991). Gender differences in psychosocial determinants of adolescent smoking. Journal of School Health, 61(3), 115–120. Cloward, R. A., & Ohlin, L. E. (1960). Delinquency and opportunity. New York: The Free Press. Cochran, M., & Brassard, J. (1979). Child development and personal social networks. Child Development, 50, 601–616. Cohen, A. K. (1955). Delinquent boys. Glencoe, IL: The Free Press. Cohen, A. K. (1965). The sociology of the deviant act: Anomie theory and beyond. American Sociological Review, 30, 5–14. Cohen, A. K. (1966). Deviance and control. Englewood Cliffs, NJ: Prentice-Hall. Cohen, A. K., & Short, J. (1966). Juvenile delinquency. In R. K. Merton & R. A. Nisbet (Eds.), Contemporary social problems (2nd ed., pp. 47–100). New York: Harcourt, Brace, and World. Cohen, Y., & Tyree, A. (1986). Escape from poverty: Determinants of intergenerational mobility of sons and daughters of the poor. Social Science Quarterly, 67, 803–813. Cohler, D. J., Gallant, D. H., & Grunebaum, H. U. (1977). Disturbance of attention among schizophrenic, depressed and well mothers and their five-year old children. Journal of Child Psychology and Psychiatry, 18, 115–136. Coleman, L. M. (1986). Stigma: An enigma demystified. New York: Plenum Press. Colletta, N. (1981). Social support and the risk of maternal rejection by adolescent mothers. Journal of Psychology, 109, 191–197. Conger, R. D., Ge, X., Elder, G. H., Jr., Lorenz, F. O., & Simons, R. L. (1994). Economic stress, coercive family process, and developmental problems of adolescents. Child Development, 65, 541–561. Conger, R. D., Patterson, G. R., & Ge, X. J. (1995). It takes 2 to replicate—A mediational model for the impact of parents’ stress on adolescent adjustment. Child Development, 66(1), 80–97. Corcoran, M., Gordon, R., Laren, D., & Solon, G. (1990). Effects of family and community background on economic status. The American Economic Review, 80, 362–366. Corcoran, M., Gordon, R., Laren, D., & Solon, G. (1992). The association between men’s economic status and their family and community origins. The Journal of Human Resources, 27, 575–601.
References
221
Curtis, A. (1991). Perceived similarity of mothers and their early adolescent daughters and relationship to behavior. Journal of Youth and Adolescence, 20(3), 381–397. Dalhouse, M., & Frideres, J. S. (1996). Intergenerational congruency: The role of the family in political attitudes of youth. Journal of Family Issues, 17, 227–248. de Beer, G. R. (1958). Embryos and ancestors (3rd ed). London: Oxford University Press. DeGarmo, D. S., Forgatch, M. S., & Martinez Jr., C. R. (1999). Parenting of divorced mothers as a link between social status and boys’ academic outcomes: Unpacking the effects of SES. Child Development, 70, 1231–1245. Dickei, J. R., Eshleman, A. K., Merasco, D. M., Shepard, A. M., Wilt, V., & Johnson, M. (1997). Parent-child relationships and children’s images of God. Journal for the Scientific Study of Religion, 36, 25–43. Doumas, D., Margolin, G., & John, R. S. (1994). The intergenerational transmission of aggression across three generations. Journal of Family Violence, 9(2), 157–175. Downey, G., & Coyne, J. C. (1990). Children of depressed parents: An integrative review. Psychological Bulletin, 108, 50–76. Downs, W. R., Miller, B. A., Testa, M., & Panek, D. (1992). Long-term effects of parent-to-child violence for women. Journal of Interpersonal Violence, 7, 365–382. Dunlap, E., Golub, A., Johnson, B. D., & Wesley, D. (2002). Intergenerational transmission of conduct norms for drugs, sexual exploitation and violence: A case study. British Journal of Criminology, 42, 1–20. Egeland, B. (1988). Breaking the cycle of abuse: Implications for prediction and intervention. In K. Browne, C. Davies, & P. Stratton (Eds.), Early prediction and prevention of child abuse (pp. 87–99). New York: Wiley. Egeland, B. (1993). A history of abuse is a major risk factor for abusing the next generation. In R. J. Gelles & D. R. Loseke (Eds.), Current controversies on family violence (pp. 197–208). Newbury Park, CA: Sage. Egeland, B., Jacobvitz, D., & Sroufe, L. A. (1988). Breaking the cycle of abuse. Child Development, 59, 1080–1088. Eichenbaum, L., & Orbach, S. (1983). Understanding women. New York: Basic Books. Elder, G. H., Jr. (1978). Approaches to social change and the family. In J. Demos & S. Boocock (Eds.), Turning points (pp. 1–38). Chicago, IL: University of Chicago Press. Elder, G. H., Jr., & Caspi, A. (1988). Human development and social change: An emerging perspective on the life course. In N. Bolger, A. Caspi, G. Downey, & M. Moorehouse (Eds.), Persons in context: Developmental processes (pp. 77–113). New York: Cambridge University Press. Elder, G. H., Jr., Caspi, A., & Downey, G. (1986). Problem behavior and family relationships: Life course and intergenerational themes. In A. Sorensen, F. Weinert, & L. Sherrod (Eds.), Human development and the life course: Multi-disciplinary perspectives (pp. 293–340). Hillsdale, NJ: Erlbaum. Emery, R., Weintraub, S., & Neale, J. M. (1982). Effects of marital discord on the children of schizophrenic, affectively disordered, and normal parents. Journal of Abnormal Child Psychology, 10, 215–228. Enders, C. K. (2001). A primer on maximum likelihood algorithms available for use with missing data. Structural Equation Modeling, 8, 128–141. Enders, C. K., & Bandalos, D. L. (2001). The relative performance of full information maximum likelihood estimation for missing data in structural equation models. Structural EquationModeling, 8, 430–457. Eron, L., & Huesmann, L. R. (1987). The control of aggressive behavior by changes in attitudes, values, and the conditions of learning. Advances in the Study of Aggression, 1, 139–171. Eron, L. D., Walder, L. O., & Lefkowitz, M. M. (1971). Learning of aggression in children. Boston: Little, Brown.
222
References
Esping-Anderson, G. (1990). The Three Worlds of Welfare Capitalism. Princeton, NJ: Princeton University Press. Farber, N. B., & Iversen, R. R. (1998). Family values about education and their transmission among black inner-city young women. In A. Colby, J. James, & D. Hart (Eds.), Competence and character through life (pp. 141–167). Chicago, IL: University of Chicago Press. Farrington, D. P. (1993). Understanding and preventing bullying (Vol. 17). Chicago: University of Chicago Press. Farrington, D. P. (1995). The development of offending and antisocial behavior from childhood: Key findings from the Cambridge Study of Delinquent Development. Journal of Child Psychology and Psychiatry, 36, 929–964. Farrington, D. P. (2002). Families and crime. In J. Q. Wilson & J. Petersilia (Eds.), Crime: Public policies for crime control (pp. 129–148). Oakland, CA: Institute for Contemporary Studies Press. Farrington, D. P., & West, D. J. (1995). Effects of marriage, separation and children on offending by adult males. In J. Hagan (Ed.), Current perspectives on aging and the life cycle, Vol. 4. Delinquency and disrepute in the life course (pp. 249–281). Greenwich, CT. JAI Press. Fawzy, F. I., Coombs, R. H., & Gerber, B. (1983). Generational continuity in the use of substances: The impact of parental substance use on adolescent substance use. Addictive Behaviors, 8, 109–114. Feng, D., Giarrusso, R., Bengtson, V. L., & Frye, N. (1999). Intergenerational transmission of marital quality and marital instability. Journal of Marriage and the Family, 61, 451–463. Field, T., Healy, B., Goldstein, S., & Guthertz, M. (1990). Behavior-state matching and synchrony in mother-infant interactions of nondepressed versus depressed dyads. Developmental Psychology, 26, 7–14. Filsinger, E. E., & Lamke, L. K. (1983). The lineage transmission of interpersonal competence. Journal of Marriage and the Family, 45, 75–80. Fisher, L., & Jones, J. E. (1980). Child competence and psychiatric risk: II. Areas of relationship between child and family functioning. Journal of Nervous and Mental Disease, 168, 332–342. Flax, J. (1981). The conflict between nurturance and autonomy in mother-daughter relationships and within feminism. In E. Howell & M. Bayes (Eds.), Women and mental health (pp. 51–68). New York: Basic Books. Frias-Armenta, M. (2002). Long-term effects of child punishment on Mexican women: A structural model. Child Abuse and Neglect, 26, 371–386. Fry, D. P. (1993). The intergenerational transmission of disciplinary practices and approaches to conflict. Human Organization, 52(2), 176–185. Furstenberg, F., Brooks-Gunn, J., & Morgan, S. (1987). Adolescent mothers in later life. New York: Cambridge University Press. F. F., Jr. (1976). Unplanned parenthood: The social consequences of teenage childbearing. New York: The Free Press. Furstenberg, F. F., Jr., Levine, J. A., & Brooks-Gunn, J. (1990). The children of teenage mothers: Patterns of early childbearing in two generations. Family Planning Perspectives, 22(2), 54–61. Garfinkel, I., & McLanahan, S. S. (1986). Single mothers and their children: A new American dilemma. Washington, DC: Urban Institute Press. Ge, X, Conger, R. D., Cadoret, R. J., Neiderhiser, J. M., Yates, W., Troughton, E., et al. (1996). The developmental interface between nature and nurture: A mutual influence model of child antisocial behavior and parent behaviors. Developmental Psychology, 32, 574–589. Gecas, V., & Seff, M. A. (1990). Families and adolescents: A review of the 1980’s. Journal of Marriage and the Family, 52, 941–958. Gest, S. D., Neeman, J., Hubbard, J. J., Masten, A. S., & Tellegen, A. (1993). Parenting quality, adversity, and conduct problems in adolescence: Testing process-oriented models of resilience. Development and Psychopathology, 5, 663–682.
References
223
Giever, D. (1995, November). An empirical assessment of the core elements of Gottfredson and Hirschi’s general theory of crime. Paper presented at the 47th annual meeting of the American Society of Criminology, Boston, MA. Gilligan, C. (1982). In a different voice. Cambridge, MA: Harvard University Press. Glass, J., Bengtson, V. L., & Dunham, C. C. (1986). Attitude similarity in three-generation families: Socialization, status inheritance, or reciprocal influence? American Sociological Review, 51, 685–698. Glenn, N. D., & Kramer, K. B. (1987). The marriages and divorces of the children of divorce. Journal of Marriage and the Family, 49, 811–825. Goffman, E. (1961). Asylums: Essays on the social situation of mental patients and other inmates. New York: Doubleday. Goodman, S. H., & Brumley, H. E. (1990). Schizophrenic and depressed mothers: Relational deficits in parenting. Developmental Psychology, 26, 31–39. Gottfredson, M., & Hirschi, T. (1990). A general theory of crime. Stanford, CA: Stanford University Press. Greenstein, T. N. (1995). Gender ideology, marital disruption, and the employment of married women. Journal of Marriage and the Family, 57, 31–42. Greenwald, R. L., Bank, L., Reid, J. B., & Knutson, J. F. (1997). A discipline-mediated model of excessively punitive parenting. Aggressive Behavior, 23, 259–280. Grych, J. H., & Fincham, F. D. (1990). Marital conflict and children’s adjustment: A cognitive-contextual framework. Psychological Bulletin, 108, 267–290. Hagan, J., & Palloni, A. (1990). The social reproduction of a criminal class in working-class London, circa 1950–1980. American Journal of Sociology, 96(2), 265–299. Hanson, R. A., & Mullis, R. L. (1986). Intergenerational transfer of normative parental attitudes. Psychological Reports, 59, 711–14. Harburg, E., Davis, D. R., & Caplan, R. (1982). Parent and offspring alcohol use. Journal of Studies in Alcohol, 43, 497–516. Harburg, E., Gleiberman, L., DiFranceisco, W., Schork, A., & Weissfeld, L. (1990). Familial transmission of alcohol use, III. Impact of imitation/nonimitation of parent alcohol use (1960) on the sensible/problem drinking of their offspring (1977). British Journal of Addiction, 85, 1141–1155. Hardy, J. B., Astone, N. M., Brooks-Gunn, J., Shapiro, S., & Miller, T. L. (1998). Like mother, like child: Intergenerational patterns of age at first birth and associations with childhood and adolescent characteristics and adult outcomes in the second generation. Developmental Psychology, 34(6), 1220–1232. Hardy, J. B., Shapiro, S., Mellits, D., Skinner, E. A., Astone, N. M., Ensminger, M., et al. (1997). Selfsufficiency at ages 27–33 years: Factors present between birth and 18 years that predict educational attainment among children born to inner-city families. Pediatrics, 99, 80–87. Harper, C., Marcus, R., & Moore, K. (2003). Enduring poverty and the conditions of childhood: Life course and intergenerational poverty transmissions. World Development, 31(3), 535–554. Harvey, D. M., Curry, C. J., & Bray, J. H. (1991). Individuation and intimacy in intergenerational relationships and health: Patterns across two generations. Journal of Family Psychology, 5(2), 204–236. Haveman, R., Wolfe, B., & Peterson, E. (1997). How do the children of early childbearers fare as young adults? In R. Maynard (Ed.), Kids having kids: Economic costs and social consequences of teen pregnancy (pp. 257–284). Washington, DC: Urban Institute Press. Hayes, H. R., & Emshoff, J. G. (1993). Substance abuse and family violence. In R. L. Hampton, T. P. Gullotta, G. R. Adams, E. H. Potter III, & R. P. Weissberg (Eds.), Family violence: Prevention and treatment, issues in children’s and families’ lives, vol. 1. Newbury Park, CA: Sage. Heinicke, C., Diskin, S., Ramsey-Klee, D., & Given, K. (1983). Prebirth parent characteristics and family development in the first year of life. Child Development, 54, 194–208.
224
References
Hetherington, E. M., & Frankie, G. (1965). Effects of parental dominance, warmth, and conflict on imitation in children. Journal of Personality and Social Psychology, 20, 1004–1016. Hewitt, J. P. (1970). Social stratifications and deviant behavior. New York: Random House. Hirschi, T. (1969). Causes of delinquency. Berkeley: University of California Press. Hofferth, S. L., Kahn, J. R., & Baldwin, W. (1987). Premarital sexual activity among U.S. teenage women over the past three decades. Family Planning Perspectives, 19, 46–53. Hofferth, S. L., & Moore, K. A. (1979). Early childbearing and later economic well-being. American Sociological Review, 44(5), 784–815. Hogan, D. P., & Kitagawa, E. M. (1985). The impact of social status, family structure and neighborhood on the fertility of black adolescents. American Journal of Sociology, 90, 825–855. Holloway, S. D., & Machinda, S. (1991). Childrearing effectiveness of divorced mothers: Relationship to coping strategies and social support. Journal of Divorce and Remarriage, 14, 179–201. Hood, K. E., & Cairns, R. B. (1988). A developmental-genetic analysis of aggressive behavior in mice: II. Cross-sex inheritance. Behavior Genetics, 18, 605–619. Hooley, J. M., Richters, J. E., Weintraub, S., & Neale, J. M. (1987). Psychopathology and marital distress: The positive side of positive symptoms. Journal of Abnormal Psychology, 96, 27–33. Hout, M. (1984). Status, autonomy, and training in occupational mobility. American Journal of Sociology, 89, 1379–1409. Hout, M, (1988). More universalism, less structural mobility: The American occupational structure in the 1980s. American Journal of Sociology, 93, 1358–1400. Huesmann, L. R., Lagerspetz, K., & Eron, L. D. (1984). Intervening variables in the television violence-aggression relation: Evidence from two countries. Developmental Psychology, 20, 746–775. Huesmann, R. L., Eron, L. D., Lefkowitz, M. M., & Walder, L. O. (1984). Stability of aggression over time and generations. Developmental Psychology, 20(6), 1120–1134. Hunter, R. S., & Kilstrom, N. (1979). Breaking the cycle in abusive families. American Journal of Psychiatry, 136, 1320–1322. Israel, M., & Seeborg, M. (1998). The impact of youth characteristics and experiences on transitions out of poverty. Journal of Socio-Economics, 27(6), 753–776. Jacob, T., & Krahn, G. L. (1988). Marital interactions of alcoholic couples: Comparison with depressed and nondistressed couples. Journal of Consulting and Clinical Psychology, 56, 73–79. Jacob, T., Ritchey, D., Cvitkovic, J. F., & Blane, H. T. (1981). Communication styles of alcoholic and nonalcholic families when drinking and not drinking. Journal of Studies on Alcohol, 42, 466–482. Jankowski, M. K., Leitenberg, H., Henning, K., & Coffey, P. (1999). Intergenerational transmission of dating aggression as a function of witnessing only same sex parents vs. opposite sex parents vs. both parents as perpetrators of domestic violence. Journal of Family Violence, 14(3), 267–279. Jennings, M. K., & Niemi, R. G. (1981). Generations and politics: A panel study of young adults and their parents. Princeton, NJ: Princeton University Press. Johnson, V., & Bennett, M. E. (1995). Assessing and tracking family histories of alcoholism. Journal of Studies on Alcohol, 56, 654–660. Joreskog, K. G., & Sorbom, D. (1994a). New features in PRELIS 2. In K.G. Joreskog & D. Sorbom, PRELIS 2: User’s reference guide, 1996 (3rd ed.). Chicago, IL: Scientific Software International. Joreskog, K. G., & Sorbom, D. (1994b). New features in LISREL 8. In K.G. Joreskog & D. Sorbom, LISREL 8: User’s reference guide, 1996 (2nd ed.). Chicago, IL: Scientific Software International. Kahn, J. R., & Anderson, K. E. (1992). Intergenerational patterns of teenage fertility. Demography, 29(1), 39–57.
References
225
Kahn, J. R., Rindfuss, R. R., & Guilkey, D. K. (1990). Adolescent contraceptive method choices. Demography, 27(5), 323–335. Kalmuss, D. (1984). The intergenerational transmission of marital aggression. Journal of Marriage and the Family, 46(1), 11–19. Kandel, D. B. (1990). Parenting styles, drug use, and children’s adjustment in families of young adults. Journal of Marriage and the Family, 52, 183–196. Kandel, D. B., & Wu, P. (1995). The contribution of mothers and fathers to the intergenerational transmission of cigarette smoking in adolescence. Journal of Research Adolescence, 5, 225–252. Kaplan, D. S., Kaplan, H. B., & Liu, X. (2000). Family structure and parental involvement in the intergenerational parallelism of school adversity. The Journal of Educational Research, 93(4), 235–244. Kaplan, H. B. (1972). Toward a general theory of psychosocial deviance: The case of aggressive behavior. Social Science & Medicine, 6(5), 593. Kaplan, H. B. (1975). Self-attitudes and deviant behavior. Pacific Palisades, CA Goodyear. Kaplan, H. B. (1980). Deviant behavior in defense of self. New York: Academic Press. Kaplan, H. B. (1982). Self-attitudes and deviant behavior: New directions for theory and research. Youth and Society, 14(2), 185–211. Kaplan, H. B. (1983). Psychological distress in sociological context: Toward a general theory of psychosocial stress. In H. B. Kaplan (Ed.), Psychosocial stress: Trends in theory and research (pp. 195–264). New York: Academic Press. Kaplan, H. B. (1984). Patterns of juvenile delinquency. Beverly Hills, CA: Sage. Kaplan, H. B. (1986). Social psychology of self-referent behavior. New York: Plenum Press. Kaplan, H. B. (1995). Drugs, crime, and other deviant adaptations. In H. B. Kaplan (Ed.), Drugs, crime, and other deviant adaptations: Longitudinal studies (pp. 3–46). New York: Plenum Press. Kaplan, H. B. (2003) “Testing an Integrative Theory of Deviant Behavior: Theory-Syntonic Findings from a Long-Term Multi-Generational Study.” pp. 185–204 in Taking Stock of Delinquency: An Overview of Findings from Contemporary Longitudinal Studies, edited by T. Thornberry and M. Krohn. New York, NY: Kluwer Academic/Plenum. Kaplan, H. B., & Johnson, R. J. (2001). Social deviance: Testing a general theory. New York: Kluwer Academic/Plenum Publishers. Kaplan, H. B., & Lin, C. (2000). Deviant identity as a moderator of the relation between negative selffeelings and deviant behavior. Journal of Early Adolescence, 20(2), 150–177. Kaplan, H. B., & Lin, C. (2005). Deviant identity, negative self-feelings, and decreases in deviant behavior: The moderating influence of conventional social bonding. Psychology, Crime & Law, 11(3), 289. Kaplan, H. B., & Liu, X. (1999). Explaining transgenerational continuity in antisocial behavior during early adolescence. In P. Cohen, C. Slomkowski, & L. Robins (Eds.), Historical and geographical aspects of psychopathology (pp. 163–191). Mahwah, NJ: Erlbaum and Associates. Kaplan, H. B., & Liu, X. (2000). Social protest and self-enhancement: A conditional relationship. Sociological Forum, 14(4), 595–616. Kaufman, J., & Zigler, E. (1987). Do abused children become abusive parents? American Journal of Orthopsychiatry, 57, 186–192. Kaufman, J., & Zigler, E. (1993). The intergenerational transmission of abuse is overstated. In R. J. Gelles & D. R. Loseke (Eds.), Current controversies on family violence (pp. 209–221). Newbury Park, CA: Sage. Keith, V. M., & Finlay, B. (1988). The impact of parental divorce on children’s educational attainment, marital timing, and the likelihood of divorce. Journal of Marriage and the Family, 50, 797–809. Kellam, S., Ensminger, M. E., & Turner, R. (1977). Family structure and the mental health of children. Archives of General Psychiatry, 34, 1012–1022.
226
References
Kiernan, K. E., & Diamond, I. (1983). The age at which childbearing starts—A longitudinal study. Population Studies, 37, 363–380. Kirkpatrick, L. A., & Shaver, P. R. (1990). Attachment theory and religion: Childhood attachments, religious beliefs, and conversion. Journal for the Scientific Study of Religion, 29, 315–334. Kitsuse, J. I. (1962). Societal reaction to deviant behavior: Problems of theory and method. Social Problems, 9, 247–257. Knutson, J. F., & Bower, M. E. (1994). Physically abusive parenting as an escalated aggressive response. In M. Potegal & J. F. Knutson (Eds.), The dynamics of aggression: Biological and social processes in dyads and groups (pp. 195–225). Hillsdale, NJ: Erlbaum. Kohn, M. (1977). Class and conformity (2nd ed.). Chicago, IL: University of Chicago Press. Kramer, L., & Baron, L. A. (1995). Intergenerational linkages: How experiences with siblings relate to the parenting of siblings. Journal of Social and Personal Relationships, 12(1), 67–87. Krein, S. F., & Beller, A. H. (1988). Educational attainment of children from single-parent families: Differences by exposure, gender, and race. Demography, 25, 221–234. Kreuz, L. E., & Rose, R. M. (1972). Assessment of aggressive behaviors and plasma testosterone in a young criminal population. Psychosomatic Medicine, 34, 321–332. Kumpfer, K. L. (1995, November). Impact of maternal characteristics and parenting processes on children of drug abusers. Paper presented at the 47th annual meeting of the American Society of Criminology, Boston, MA. La Sorsa, V. A., & Fodor, I. G. (1990). Adolescent daughter/midlife mother dyad. Psychology of Women Quarterly, 14, 593–606. Lagerspetz, K., & Lagerspetz, K. Y. H. (1974). Genetic determination of aggressive behavior. In I. H. F. von Abeelen (Ed.), Behavioral genetics (pp. 321–346). Amsterdam: North-Holland. Lee, C. M., & Gotlib, I. H. (1991). Adjustment of children of depressed mothers: A ten-month follow-up. Journal of Abnormal Psychology, 100, 473–477. Lefkowitz, M. M., Huesmann, L. R., & Eron, L. D. (1978). Parental punishment: A longitudinal analysis of effects. Archives of General Psychiatry, 35, 186–191. Lemert, E. (1951). Social pathology. New York: McGraw-Hill. Little, R. J. A., & Rubin, D. B. (1987). Statistical analysis with missing data. New York: Wiley. Loeber, R., & Stouthamer-Loeber, M. (1986). Family factors as correlates and predictors of juvenile conduct problems and delinquency. In M. Tonry & N. Morris (Eds.), Crime and justice: An annual review of research, Vol. 7 (pp. 29–149). Chicago, IL: University of Chicago Press. Lundberg, M., Perris, C., Schlette, P., & Adolfsson, R. (2000). Intergenerational transmission of perceived parenting. Personality and Individual Differences, 28, 865–877. Luntz, B. K., & Widom, C. S. (1994). Antisocial personality disorder in abused and neglected children grown up. American Journal of Psychiatry, 151, 670–674. MacEwen, K. E. (1994). Refining the intergenerational transmission hypothesis. Journal of Interpersonal Violence, 9(3), 350–365. Magnusson, D., & Cairns, R. B. (1996). Developmental science: Principles and illustrations. In R. B. Cairns, Elder, G. H., Jr., & J. Costello (Eds.), Developmental science (pp. 7–30). New York: Cambridge University Press. Main, M., & Goldwyn, R. (1984). Predicting rejection of her infant from mother’s representation of her own experience: Implications for the abused-abusing intergenerational cycle. Child Abuse and Neglect, 23, 305–319. Main, M., Kaplan, N., & Cassidy, J. (1985). Security in infancy, childhood, and adulthood: A move to the level of representation. In I. Bretherton, & E. Water (Eds.), Monographs of the society for research in child development [Special issue]. Growing Points of Attachment Theory and Research, 50(1–2), 66–104. Malinosky-Rummell, R., & Hansen, D. J. (1993). Long term consequences of childhood physical abuse. Psychological Bulletin, 114, 68–79.
References
227
Manlove, J. (1997). Early motherhood in an intergenerational perspective: The experiences of a British cohort. Journal of Marriage and the Family, 59, 263–279. Mann, B. J., & MacKenzie, E. P. (1996). Pathways among marital functioning, parental behaviors, and child behavior problems in school-age boys. Journal of Clinical Child Psychology, 25, 183–191. Mark, V., & Ervin, F. (1970). Violence and the brain. New York: Harper and Row. Markowitz, F. E. (2001). Attitudes and family violence: Linking intergenerational and cultural theories. Journal of Family Violence, 16(2), 205–218. McCloskey, L. A., & Bailey, J. A. (2000). The intergenerational transmission of risk for child sexual abuse. Journal of Interpersonal Violence, 15(10), 1019–1035. McCord, J. (1988). Parental behavior in the cycle of aggression. Psychiatry, 51, 14–23. McCord, J. (1991). The cycle of crime and socialization practices. The Journal of Criminal Law and Criminality, 82, 211–228. McLanahan, S. S., & Bumpass, L. L. (1988). Intergenerational consequences of family disruption. American Journal of Sociology, 94(1), 130–152. McLeod, J. D. (1991). Childhood parental loss and adult depression. Journal of Helath and Social Behavior, 32, 205–220. 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. Mead, G. H. (1918). The psychology of punitive justice. American Journal of Sociology, 23, 577–602. Merton, R. K. (1938). Social structure and anomie. American Sociological Review, 3, 672–682. Michael, R. T., & Tuma, N. B. (1985). Entry into marriage and parenthood by young men and women: The influence of family background. Demography, 22(4), 515–544. Miller, J. B. (1986). Toward a new psychology of women (2nd ed.). Boston: Beacon Press. Miller, L., Kramer, R., Warner, V., Wickramaratne, P., & Weissman, M. (1997). Intergenerational transmission of parental bonding among women. Journal of the American Academy of Child and Adolescent Psychiatry, 36(8), 1134–1139. Miller, R. B., & Glass, J. (1989). Parent-child attitude similarity across the life course. Journal of Marriage and the Family, 51, 991–997. Miller, W. B. (1958). Lower class culture as a generating milieu of gang delinquency. Journal of Social Issues, 14, 5–19. Moen, P., Erickson, M. A., & Demptser-McClain, D. (1997). Their mothers’ daughters? The intergenerational transmission of gender attitudes in a world of changing roles. Journal of Marriage and the Family, 59, 281–293. Moore, K. (2001). Frameworks for understanding the intergenerational transmission of poverty and well-being in developing countries (CPRC working paper No. 8). Manchester, UK: IDPM, University of Manchester. Moore, K. A., & Burt, M. R. (1982). Private crisis, public cost: Policy perspectives on teenage childbearing. Washington, DC: Urban Institute. Moore, K. A., Hofferth, S. L., & Wertheimer, R. F. (1981). Teenage childbearing: Consequences for women, families, and government welfare expenditures. In K. G. Scott, T. Field, & E. G. Robertson (Eds.), Teenage parents and their offspring (pp. 35–54). New York: Grune and Stratton. Muller, R. T., Fitzgerald, H. E., Sullivan, L. A., & Zucker, R. A. (1994). Social support and stress factors in child maltreatment among alcoholic families. Canadian Journal of Behavioral Science, 26, 438–461. Muller, R. T., Hunter, J. E., & Stollak, G. (1995). The intergenerational transmission of corporal punishment: A comparison of social learning and temperament models. Child Abuse and Neglect, 19(11), 1323–1335. Muthen, L. K., & Muthen, B. O. (2004). Mplus user’s guide. Los Angles: Muthen & Muthen.
228
References
Muthen, B. O., & Satorra, A. (1995). Technical aspects of Muthen’s LISCOMP approach to estimation of latent variable relations with a comprehensive measurement model. Psychometrika, 60, 489–503. Narang, D. S., & Contreras, J. M. (2000). Dissociation as a mediator between child abuse history and adult abuse potential. Child Abuse and Neglect, 24(5), 653–665. Newcomer, S. F., & Udry, J. R. (1984). Mothers’ influence on the sexual behavior of their teenage children. Journal of Marriage and the Family, 46, 477–485. Newson, J., Newson, E., & Adams, M. (1993). The social origins of delinquency. Criminal Behavior and Mental Health, 3, 19–29. Noddings, N. (1984). Caring: A feminine approach to ethics and moral education. Berkeley: University of California Press. Noller, P., Seth-Smith, M., Bouma, R., & Schweitzer, R. (1992). Parent and adolescent perceptions of family functioning: A comparison of clinic and non-clinic families. Journal of Adolescence, 15, 101–114. Offer, D., Kaiz, M., Ostrov, E., & Albert, D. B. (2003). Continuity in family constellation. Adolescent and Family Health, 3(1), 3–8. Ogbu, J. (1981). Origins of human competence: A cultural-ecological perspective. Child Development, 52, 413–249. Oliver, J. E. (1993). Intergenerational transmission of child abuse: Rates, research, and clinical implications. American Journal of Psychiatry, 150(9), 1315–1324. Offord, D. R., Allen, N., & Abrams, N. (1978). Parental psychiatric illness, broken homes and delinquency. Journal of American Academy of Child Psychiarty, 17, 224–238. Orford, J., & Velleman, R. (1991). The environmental intergenerational transmission of alcohol problems: A comparison of two hypotheses. British Journal of Medical Psychology, 64, 189–200. Parke, R. D., & Buriel, R. (1997). Socialization in the family: Ethnic and ecological perspectives. In W. Damon & N. Eisenberg (Eds.), Handbook of child psychology: Social, emotional, and personality development (5th ed., pp. 463–552). New York: Wiley. Patterson, G. (1998). Continuities: A search for causal mechanisms. Developmental Psychology, 34, 1263–1268. Patterson, G., De Baryshe, B., & Ramsey, E. (1989). A developmental perspective on antisocial behavior. American Psychologist, 44, 329–335. Patterson, G. R., & Dishion, T. J. (1988). Multilevel process models: Traits, interactions, and relationships. In R. A. Hinde & J. Stevenson-Hinde (Eds.), Relationships within families (pp. 283–310). Oxford: Clarendon Press. Patterson, G. R., & Stouthamer-Loeber, M. (1984). The correlation of family management practices and delinquency. Child Development, 55, 1299–1307. Patterson, G. R., Reid, J. B., & Dishion, T. J. (1992). Antisocial boys. Eugene, OR: Castalia Publishing. Pears, K. C., & Capaldi, D. M. (2001). Intergenerational transmission of abuse: A two-generational prospective study of an at-risk sample. Child Abuse and Neglect, 25, 1439–1461. Perry, B. D. (1997). Incubated in terror: Neurodevelopmental factors in the “cycle of violence.” In J. D. Osofsky (Ed.), Children in a violent society (pp. 124–149). New York: Guilford Press. Perry, D. G., & Bussey, K. (1979). The social learning theory of sex differences: Imitation is alive and well. Journal of Personality and Social Psychology, 37, 1699–1712. Peters, E. H. (1992). Patterns of intergenerational mobility in income and earnings. The Review of Economics and Statistics, 74, 456–466. Peters, M. F. (1976). Nine black families: A study of household management and childrearing in black families with working mothers. Ann Arbor, MI: University Microfilms. Peterson, P. L., Hawkins, J. D., Abbott, R. D., & Catalano, R. F. (1994). Disentangling the effects of parental drinking, family management, and parental alcohol norms on current drinking by black and white adolescents. Journal of Research on Adolescence, 4(2), 203–227.
References
229
Petit, G. S., Clawson, M. A., Dodge, K. A., & Bates, J. E. (1996). Stability and change in peer-rejected status: The role of child behavior, parenting and family ecology. Merill Palmer Quarterly, 42, 267–294. Phalet, K., & Schönpflug, U. (2000). Intergenerational transmission in Turkish immigrant families: Parental collectivism, achievement values, and gender differences. Journal of Comparative Family Studies, 32(4), 489–504. Philips, R. (1991). Untying the knot: A short history of divorce. Cambridge, MA: Harvard University Press. Plass, P. S., & Hotaling, G. T. (1995). The intergenerational transmission of running away: Childhood experiences of the parents of runaways. Journal of Youth and Adolescence, 24(3), 335–348. Polk, F., & Halferty, D. (1966). Adolescence, commitment, and delinquency. Journal of Research in Crime and Delinquency, 4, 82–96. Putallaz, M., Costanzo, P. R., Grimes, C. L., & Lipton, D. (1998). Intergenerational continuities and their influences on children’s social development. Social Development, 7, 389–427. Quinton, D., & Rutter, M. (1984). Parents with children in care: I. Intergenerational continuities. Journal of Child Psychology and Psychiatry, 25, 231–250. Raajimakers, Q. (1999). Beliefs on politics and society. In W. Meeus & H. ‘t Hart (Eds.), Youth in the Netherlands (pp. 106–132). Amersfoort, The Netherlands: Academische Uitgeverij Amersfoort. Rada, R. T., Kellner, R., & Winslow, W. W. (1976). Plasma testosterone and aggressive behavior. Psychosomatics, 16, 138–142. Radke-Yarrow, M., Campbell, J. D., & Burton R. V. (1968). Child rearing: An inquiry in research and methods. San Francisco: Jossey-Bass. Rand, D. (1992). Intergenerational continuity of parental rejection and depressed affect. Journal of Personality and Social Psychology, 63, 1036–1045. Reckless, W. (1967). The crime problem (4th ed). New York: Appleton-Century-Crofts. Reckless, W., Dinitz, S., & Murray, E. (1956). Self concept as an insulator against delinquency. American Sociological Review, 21, 744–746. Rickard, K. M., Forehand, R., Wells, K. C., Griest, D. L., & McMahon, R. J. (1981). Factors in the referral of children for behavioral treatment: A comparison of mothers of clinic-referred deviant, clinic-referred non-deviant, and non-clinic children. Behavioral Research and Therapy, 19, 201–205. Ricks, M. (1985). The social transmission of parental behavior: Attachment across generations. In I. Bretherton, & E. Water (Eds.), Monographs of the society for research in child development [Special issue]. Growing Points of Attachment Theory and Research, 50(1–2), 211–227. Rindfuss, R. R., Brewster, K. L., & Kavee, A. L. (1996). Women, work, and children: Behavioral and attitudinal change in the United States. Population and Development Review, 22(3), 457–482. Rohan, M. J., & Zanna, M. P. (1996). Values transmission in families. In C. Seligman, J. M. Olson, & M. P. Zanna (Eds.), The psychology of values: The Ontario symposium, Volume 8 (pp. 253–276). Mahwah, NJ: Erlbaum. Roosa, M. W., Tien, J., Groppenbacher, N., Michaels, M., & Dumka, L. (1993). Mothers’ parenting behavior and child mental health in families with a problem drinking parent. Journal of Marriage and the Family, 55, 107–118. Ross, C. E., & Mirowsky, J. (1999). Parental divorce, life-course disruption, and adult depression. Journal of Marriage and the Family, 28, 331–357. Rossi, A. (1980). Aging and parenthood in the middle years. In P. Baltes & O. Brim (Eds.), Life-span development and behavior (Vol. 2). New York: Academic Press. Rubin, D. B. (1976). Inference and missing data. Biometrika, 63, 581–592 Ruscher, S. M., & Gotlib, I. H. (1988). Marital interaction patterns of couples with and without a depressed partner. Behavior Therapy, 19, 455–470. Rutter, M. (1987). Psychosocial resilience and protective mechanisms. American Journal of Orthopsychiatry, 57, 316–331.
230
References
Rutter, M. (1998). Some research considerations on intergenerational continuities and discontinuities: Comment on the special section. Developmental Psychology, 34, 1269–1273. Rutter, M., Giller, H., & Hagell, A. (1998). Antisocial behavior by young people. Cambridge, MA: Cambridge University Press. Rutter, M., & Madge, N. (1976). Cycles of disadvantage: A review of research. London: Heinemann. Rutter, M., Quinton, D., & Liddle, C. (1983). Parenting in two generations: Looking backwards and looking forwards. In N. Madge (Ed.), Families at risk (pp. 60–98). London: Heinemann. Rutter, M., Quinton, D., & Yule, B. (1976). Family pathology and disorder in the children. Chichester: Wiley. Sameroff, A. J., Seifer, R., & Zax, M. (1982). Early development of children at risk for emotional disorder. Monographs of the Society for Research in Child Development, 47(7). Sampson, R. J., & Laub, J. H. (1993). Crime in the making: Pathways and turning points through life. Cambridge, MA: Harvard University Press. Sandefur, G. D., & Wells, T. (1999). Does family structure really influence educational attainment. Social Science Research, 28, 331–371. Sansbury, L. E., & Wahler, R. G. (1992). Pathways to maladaptive parenting with mothers and their conduct disordered children. Behavior Modification, 16(4), 574–592. Scheff, T. (1966). Being mentally ill: A sociological theory. Chicago, IL: Aldine. Schwartz, M., & Tangri, S. S. (1965). A note on self-concept as an insulator against delinquency. American Sociological Review, 30, 922–926. Serbin, L. A. (1996, November). Girls’ childhood aggression as a predictor of adult family functioning and behavior problems in offspring: A prospective study over 20 years and two generations. Paper presented at the Séminaire L’Aggression Chez Les Filles, Montreal, Canada. Serbin, L. A., Cooperman, J. M., Peters, P. L., Lehoux, P. M., Stack, D. M., & Schwartzman, A. E. (1998). Intergenerational transfer of psychosocial risk in women with childhood histories of aggression, withdrawal, or aggression and withdrawal. Developmental Psychology, 34, 1246–1262. Serbin, L. A., Moskowitz, D. S., Schwartzman, A. E., & Ledingham, J. E. (1991). Aggressive, withdrawn, and aggressive/withdrawn children in adolescence: Into the next generation. In D. J. Pepler & K. H. Rubin (Eds.), The development and treatment of childhood aggression (pp. 55–70). Hillsdale, NJ: Erlbaum. Serbin, L. A., & Stack, D. M. (1998). Introduction to the special section: Studying intergenerational continuity and the transfer of risk. Developmental Psychology, 34, 1159–1161. Sher, K. J. (1991). Children of alcoholics: A critical appraisal of theory and research. Chicago, IL: University of Chicago Press. Sheridan, M. J. (1995). A proposed intergenerational model of substance abuse, family functioning, and abuse/neglect. Child Abuse and Neglect, 19, 519–530. Shlonsky, H. R. (1984). Continuity in poverty along family lines: A reexamination of the intergenerational cycle of poverty. Human Relations, 37(6), 455–472. Silverman, A. B., Reinherz, H. Z., & Giaconia, R. M. (1996). The long-term sequelae of child and adolescent abuse: A longitudinal community study. Child Abuse and Neglect, 20, 709–723. Simons, R. L., Beaman, J., Conger, R. D., & Chao, W. (1992). Gender differences in the intergenerational transmission of parenting beliefs. Journal of Marriage and the Family, 54, 823–836. Simons, R. L., Beaman, J., Conger, R. D., & Chao, W. (1993). Childhood experience, conceptions of parenting, and attitudes of spouse as determinants of parenting behavior. Journal of Marriage and the Family, 55, 91–106. Simons, R. L., Wu, C. I., Johnson, C., & Conger, R. D. (1995). A test of various perspectives on the intergenerational transmission of domestic violence. Criminology, 33(1), 141–171. Simons, R. L., Johnson, C., & Conger, R. D. (1994). Harsh corporal punishment versus quality of parental involvement as an explanation of adolescent maladjustment. Journal of Marriage and the Family, 56, 591–607.
References
231
Simons, R. L., Whitbeck, L. B., Conger, R. D., & Wu, C. I. (1991). Intergenerational transmission of harsh parenting. Developmental Psychology, 27, 159–171. Small, S. A. (1988). Parental self-esteem and its relationship to child-rearing practices, parent-adolescent interaction and adolescent behavior. Journal of Marriage and the Family, 50, 1063–1072. Smith, C. A., & Farrington, D. P. (2004), Continuities in antisocial behavior and parenting across three generations. Journal of child psychology and psychiatry. 45(2), 230–247. Smith, M. D., & Self, G. (1980). The congruence between mothers’ and daughters’ sex-role attitudes: A research note. Journal of Marriage and the Family, 42, 105–109. Smyth, N. J., Miller, B. A., Janicki, P. I., & Mudar, P. (1995, November). Mothers’ protectiveness and child abuse: The impact of her history of childhood sexual abuse and alcohol diagnosis. Paper presented at the 47th annual meeting of the American Society of Criminology, Boston, MA. Solon, G. (1992). Intergenerational income mobility in the United States. The American Economic Review, 82, 393–408. Solon, G., Corcoran, M., Gordon, R., & Laren, D. (1991). A longitudinal analysis of sibling correlations in economic status. The Journal of Human Resources, 26, 509–534. South, S. J. (1995). Do you need to shop around? Age at marriage, spousal alternatives, and marital dissolution. Journal of Family Issues, 16, 432–449. Stafleu, A., Van Staveren, W. A., De Graaf, C., Burema, J., Joseph, G. A., & Hautvast, J. (1995). Family resemblance in beliefs, attitudes, and intentions towards consumption of 20 foods: A study among three generations of women. Appetite, 25(3), 201–216. Steiger, J. H., & Lind, J. C. (1980). Statistically-based tests for the number of common factors, Spring Meeting of the Psychometric Society. Iowa City. Steiger, H., Stotland, S., Trottier, J., & Ghadirian, A. M. (1996). Familial eating concerns and psychopathological traits: Causal implications of transgenerational effects. International Journal of Eating Disorders, 19(2), 147–157. Steinberg, L. (2000). We know some things: Parent-adolescent relations in retrospect and prospect. Journal of Reserch on Adolescence, 10, 83–110. Stern, S. B., & Smith, C. A. (1995). Family processes and delinquency in an ecological context. Social Service Review, 69, 705–731. Stern, S. B., & Smith, C. A. (1999). Urban families and adolescent mental health. Social Work Research, 23, 15–17. Stith, S. M., Rosen, K. H., Middleton, K. A., Busch, A. L., Lundeberg, K. C., & Russell P. (2000). The intergenerational transmission of spouse abuse: A meta-analysis. Journal of Marriage and the Family, 62, 640–654. Straus, M. A., & Gelles, R. J. (1986). Societal change and change in family violence from 1975 to 1985 as revealed by two national surveys. Journal of Marriage and the Family, 48, 465–479. Straus, M. A., & Gelles, R. J. (1988). How violent are American families? Estimates from the National Family Violence Resurvey and other studies. In G. T. Hotaling, J. T. Finkelhor, D. Kirkpatrick, & M. A. Straus (Eds.), Family abuse and its consequences: New directions in research (pp. 14–37). Newbury Park, CA: Sage. Straus, M. A., Gelles, R. J., & Steinmetz, S. K. (1980). Behind closed doors: Violence in the American family. Garden City, NY: Anchor Books. Straus, M. A., & Kantor, G. K. (1994). Corporal punishment of adolescents by parents: A risk factor in the epidemiology of depression, suicide, alcohol abuse, child abuse, and wife beating. Adolescence, 29(115), 543–561. Styron, T., & Janoff-Bulman, R. (1997). Childhood attachment and abuse: Long-term effects on adult attachment, depression and conflict resolution. Child Abuse and Neglect, 21, 1015–1023.
232
References
Sutherland, E. H., & Cressey, D. (1974). Criminology (9th ed.). Philadelphia: J. B. Kippincott. Tallman, I., Gray, L. N., Kullberg, V., & Henderson, D. (1999). The intergenerational transmission of marital conflict: Testing a process model. Social Psychology Quarterly, 62(3), 219–239. Tapscott, M., Frick, P. J., Wooton, J., & Kruh, I. (1996). The intergenerational link to antisocial behavior: Effects of paternal contact. Journal of Child and Family Studies, 5(2), 229–240. Taris, T. W. (2000). Quality of mother-child interaction and the intergenerational transmission of sexual values: A panel study. The Journal of Genetic Psychology, 161(2), 169–181. Taris, T. W., Semin, G. R., & Bok, I. A. (1998). The effect of quality of family interaction and intergenerational transmission of values on sexual permissiveness. The Journal of Genetic Psychology, 159(2), 237–250. Tarter, R. E., Blackson, T., Martin, C., Loeber, R., & Moss, H. B. (1993). Characteristics and correlates of child discipline practices in substance abuse and normal families. American Journal on Addictions, 2(1), 18–25. Tedeschi, J. T., & Felson, R. B. (1994). Violence, aggression, and coercive actions. Washington, DC: American Psychological Association. Thompson, K. M., & Wilsnack, R. W. (1984). Drinking and drinking problems among female adolescents: Patterns and influences. In S. C. Wilsnack & L. J. Beckman (Eds.), Alcohol problems in women (pp. 37–65). New York: Guilford. Thompson, L., Clark, K., & Gunn, W., Jr. (1985). Developmental stage and perceptions of intergenerational continuity. Journal of Marriage and the Family, 47(4), 913–920. Thornton, A., (1991). Influence of the marital history of parents on the marital and cohabitational experiences of children. American Journal of Sociology, 96(4), 868. Thornton, A., & Camburn, D. (1987). The influence of the family on premarital sexual attitudes and behavior. Demography, 24, 323–240. Van Ijzendoorn, M. H. (1992). Intergenerational transmission of parenting: A review of studies in nonclinical populations. Developmental Review, 12, 76–99. Van Ijzendoorn, M. H. (1995). Adult attachment representations, parental responsiveness, and infant attachment: A meta-analysis on the predictive validity of the adult attachment interview. Psychological Bulletin, 117, 387–403. Velleman, R. (1992). Intergenerational effects—A review of environmentally oriented studies concerning the relationship between parental alcohol problems and family disharmony in the genesis of alcohol nad other problems. II: The intergenerational effects of family disharmony. International Journal of the Addictions, 27(4), 367–389. Velleman, R., & Orford, J. (1990). Adult offspring of parents with drinking problems: Recollection of parent’s drinking and its immediate effects. British Journal of Clinical Psychology, 29, 297–317. Vollebergh, W. A., Iedema, J., & Raaijmakers, Q. (1999). The intergenerational transmission of cultural and economic conservatism. In H. De Witte & P. Scheepers (Eds.), Ideology in the low countries: Trends, models and lacunae (pp. 51–69). Assen, The Netherlands: Van Gorcum. Vollebergh, W. A., Iedema, J., & Raaijmakers, Q. (2001). Intergenerational transmission and the formation of cultural orientations in adolescence and young adulthood. Journal of Marriage and the Family, 63, 1185–1198. Voss, H. (1969). Differential association and containment theory: A theoretical convergence. Social Forces, 47, 381–391. Wadsworth, M. E. J., & Freeman, S. R. (1983). Generation differences in beliefs: A cohort study of stability and change in religious beliefs. The British Journal of Sociology, 34(3), 416–437. Wells, L. E. (1978). Theories of deviance and the self-concept. Social Psychology, 41, 189–204. Werner, E. E., & Smith, R. S. (1982). Vulnerable but invincible: A study of resilient children. New York: McGraw-Hill.
References
233
Werner, E. E., & Smith, R. S. (1992). Overcoming the odds: High risk children from birth to adulthood. Ithaca, NY: Cornell University Press. West, D., & Farrington, D. (1977). The delinquent way of life. London: Heinemann Educational. West, O. M., & Prinz, R. J. (1987). Parental alcoholism and childhood psychopathology. Psychological Bulletin, 102, 204–218. Whitbeck, L. B., Hoyt, D. R., Simons, R. L., Conger, R. D., Elder, G. H., Jr., Lorenz, F. O., et al. (1992). Intergenerational continuity of parental rejection and depressed affect. Journal of Personality and Social Psychology, 63(6), 1036–1045. Wickrama, K. A. S., Conger, R. D., Wallace, L. E., & Elder, G. H., Jr. (1999). The intergenerational transmission of health-risk behaviors: Adolescent lifestyles and gender moderating effects. Journal of Health and Social Behavior, 40, 258–272. Willerman, L., & Cohen, D. B. (1990). Psychopathology. New York: McGraw-Hill. Wilson, M. (1989). Child development in the context of the black extended family. American Psychologist, 44, 380–383. Wilson, W. J. (1987). The truly disadvantaged: The inner city, the underclass and public policy. Chicago, IL: University of Chicago Press. Wolfinger, N. H. (1999). Trends in the intergenerational transmission of divorce. Demography, 36(3), 415–420. Wootton, J. M., Frick, P. J., Shelton, K. K., & Silverthorn, P. (1997). Ineffective parenting and childhood conduct problems: The moderating role of callous-unemotional traits. Journal of Consulting and Clinical Psychology, 65, 301–308. Wu, L. (1996). Effects of family instability, income, and income instability on the risk of premarital birth. American Sociological Review, 61, 386–406. Wu, L., & Martinson, B. C. (1993). Family structure and the risk of a premarital birth. American Sociological Review, 58, 210–232. Young, V. (1970). Family and childhood in a southern Negro community. American Anthropologist, 72, 269–288. Youngblade, L. M., & Belsky, J. (1990). Social and emotional consequences of child maltreatment. In R. T. Ammerman & M. Hersen (Eds.), Children at risk: An evaluation of factors contributing to child abuse and neglect (pp. 109–140). New York: Plenum Press. Zaidi, L. Y., Knutson, J. F., & Mehm, J. (1989). Transgenerational patterns of abusive parenting: Analog and clinical tests. Aggressive Behavior, 15, 137–152. Zuravin, S., McMillen, C., DePanfilis, D., & Risley-Curtiss C. (1996). The intergenerational cycle of child maltreatment: Continuity versus discontinuity. Journal of Interpersonal Violence, 11(3), 315–334.
Index
deviant friends, 175 (dis)continuity of, 24 disposition to deviance, 170 distressful emotions, 144 family conflict, 134 integrative and other theories, 29–35 integrative theory of, 18 intervening processes, 197 intragenerational causes of, 157 moderating variables, 196 motivation to engage in acts of, 18 negative social sanctions, 173 parental religiosity, 146 parenting patterns, 160, 162–163, 165 psychological distress, 135 social rejection, 177 socioeconomic status, 139–140, 165 Deviant behavior, correlates of, 78 abusive parenting or experiences of abuse, 131 divorce, 130 early childbearing, 129 Deviant behavior, integrative and other theories labeling theory, 35 social control/bonding theory, 33 strain-related theories, 31 subcultural/social learning theories, 32
A Abusive parenting, 65 B Barbarians or subhumans, 8 Bidirectional effects, 208 Bivariate relationship of interest, 2 D Deviance, 2 defining of, 6 disposition to, 9 first-generation, 2 motivated, 8 nature of, 6–10 second-generation, 2 unmotivated, 9 Deviant behavior, 5, 68, 86, 90, 125, 133, 140, 169, 196; see also Deviance acting out deviant dispositions, 22 adult deviant behavior, 134 alienation from conventional others, 91 child-rearing patterns, 137 common antecedents, 198 continuation or escalation of, 25 correlates of, 78, see Deviant behavior, correlates of
235
236 Deviant behavior, parenting patterns intervening processes, 163 intragenerational effects, 162 moderating variables, 162 social modeling, 165 Deviant patterns deviance in social role performance, 63 psychiatric disorders, 62 socially devalued behaviors, 62 Disposition to deviance, 9, 70, 98, 148, 179, 199 common antecedents, 202 deviant friends, 183 familial relations, 150 intergenerational influences, 106 intervening processes, 201 intragenerational stability, 98 moderating variables, 200 negative self-feelings, 180 negative social sanctions, 182 social rejection, 185 structural disadvantage, 149 E Etiology of deviance, 1 F Future elaborations, of intergenerational parallelism in deviance bidirectional effects, 208 genetic influences, 210 inclusive models, 214 measurement models, meaning of, 209 parent reports, 211 range of variables, 212 sociodemographic controls, 207 sociohistorical trends, 211 G Genetic influences, 210 I Inclusive models, 214 Integrative theory, 29 application, 37 applying an, 37–40
Index Integrative theory and other deviance theories, 29–31 labeling theory, 35–37 social control/bonding theory, 33–35 strain-related theories, 31–32 subcultural/social learning theories, 32–33 Integrative theory of deviant behavior, 18–29 acting out deviant dispositions, 22–24 continuation or escalation of deviant behavior, 25–29 (dis)continuity of deviant behavior, 24 motivation to engage in acts defined as deviant, 18–19 motivation to engage in acts that are deviant by the standards of one’s membership group(s), 19–22 Intergenerational continuities, 11–13 in poverty, 81 Intergenerational correlation, 1 Intergenerational (dis)continuities, 73 decomposition of, 125 in religious beliefs, 80 Intergenerational influence, 1 Intergenerational parallelism, 11–14 conditional nature of, 57 decomposing, 125 of deviance, see Intergenerational parallelism, of deviance deviance of, 1 in deviant behavior, 57 in disposition to deviance, 57 in divorce, 79 estimates of, 66 moderators of, 73–124 in negative self-feeling, 57 theoretical framework, 15–40 Intergenerational parallelism, literature on correlates of deviant behavior, 59 deviant patterns, 61 Intergenerational parallelism, moderators of, 89 deviant behavior, 90 disposition to deviance, 98 methodological, 75
237
Index negative self-feelings, 107 research literature on, 74, 77 substantive, 75 Intergenerational parallelism, of deviance, 169 retrospect and prospect, 193 substantive findings, 195 summary and conclusion, 195 Intergenerational parallelism in deviance, 59–72 Intergenerational parallelism in deviant behavior, 10 parental diagnosis and child functioning, 10 Intergenerational parallelism of deviance comparable developmental stages in studies of, 42–44 independent sources of data on, 47–50 mediators, common antecedents, moderators, 45–47 methodological limitations in research on, 41–42 a multigeneration prospective study on, 50–55 prospective longitudinal studies of, 44–45 understanding of, 5–40 Intergenerational replication, 1 Intergenerational transmission, 12–13 of abuse, 85 crude rates of, 73 of divorce, 79 Intervening processes, 127 literature on, 128 Intragenerational causes, intergenerational continuities of literature on, 158 M Measurement models, 209 Methodological limitations to intergenerational parallelism of deviance, 41 comparable developmental stages, 42 independent sources of data, 47 mediators, common antecedents, moderators, 45 prospective longitudinal studies, 44
Motivated deviance, 8 Multigeneration prospective study, 51–56 data collection, 51 missing data, 52 multigroup analysis, 56 statistical methods and latent variable model estimation, 55 variable construction, 53 variable nonnormality, 54 N Negative self-feelings, 70, 107, 152, 203 common antecedents, 205 family conflict, 154 G1 young adult stressful emotions, 152 intervening processes, 205 moderating variables, 204 moderators of intragenerational continuity, 108 moderators of transgenerational continuity, 115 negative social sanctions, 189 social rejection, 187 P Parent reports, 211 Psychosocial phenomena, 1 R Range of variables, 212 S Shared normative expectations, 7 Sociodemographic controls, 207 Sociohistorical trends, 211 Socionormative systems, 1 Structural equation modeling, 3 Structural equation models, 140 Supportive G2 social context, 95 T Transgenerational parallelism, 65 U Unmotivated deviance, 9