jmp cover (i).qxd
17/04/2008
10:21
Page 1
ISSN 0268-3946
Volume 23 Number 3 2008
Journal of
Managerial Psychology Complexities and challenges in the work-family interface Guest Editors: Noreen Heraty, Michael J. Morley and Jeanette N. Cleveland
www.emeraldinsight.com
Journal of Managerial Psychology
ISSN 0268-3946 Volume 23 Number 3 2008
Complexities and challenges in the work-family interface Guest Editors Noreen Heraty, Michael J. Morley and Jeanette N. Cleveland
Access this journal online ______________________________
207
Editorial boards __________________________________________
208
INTRODUCTION Complexities and challenges in the work-family interface Noreen Heraty, Michael J. Morley and Jeanette N. Cleveland ____________
209
The impact of work/family demand on work-family conflict Scott L. Boyar, Carl P. Maertz Jr, Donald C. Mosley Jr and Jon C. Carr __
215
Effort-reward imbalance, over-commitment and work-life conflict: testing an expanded model Gail Kinman and Fiona Jones ____________________________________
236
Examining couple agreement about work-family conflict Michelle Streich, Wendy J. Casper and Amy Nicole Salvaggio ___________
252
Positive effects of nonwork-to-work facilitation on well-being in work, family and personal domains Pam Allis and Michael O’Driscoll__________________________________
Access this journal electronically The current and past volumes of this journal are available at:
www.emeraldinsight.com/0268-3946.htm You can also search more than 175 additional Emerald journals in Emerald Management Xtra (www.emeraldinsight.com) See page following contents for full details of what your access includes.
273
CONTENTS
CONTENTS continued
The influence of family responsibilities, career fields and gender on career success: an empirical study Wolfgang Mayrhofer, Michael Meyer, Michael Schiffinger and Angelika Schmidt___________________________________________
292
The expatriate family: an international perspective Arno Haslberger and Chris Brewster ______________________________
324
www.emeraldinsight.com/jmp.htm As a subscriber to this journal, you can benefit from instant, electronic access to this title via Emerald Management Xtra. Your access includes a variety of features that increase the value of your journal subscription.
Structured abstracts Emerald structured abstracts provide consistent, clear and informative summaries of the content of the articles, allowing faster evaluation of papers.
How to access this journal electronically
Additional complimentary services available
To benefit from electronic access to this journal, please contact
[email protected] A set of login details will then be provided to you. Should you wish to access via IP, please provide these details in your e-mail. Once registration is completed, your institution will have instant access to all articles through the journal’s Table of Contents page at www.emeraldinsight.com/0268-3946.htm More information about the journal is also available at www.emeraldinsight.com/ jmp.htm
Your access includes a variety of features that add to the functionality and value of your journal subscription:
Our liberal institution-wide licence allows everyone within your institution to access your journal electronically, making your subscription more cost-effective. Our web site has been designed to provide you with a comprehensive, simple system that needs only minimum administration. Access is available via IP authentication or username and password. Emerald online training services Visit www.emeraldinsight.com/training and take an Emerald online tour to help you get the most from your subscription.
Key features of Emerald electronic journals Automatic permission to make up to 25 copies of individual articles This facility can be used for training purposes, course notes, seminars etc. This only applies to articles of which Emerald owns copyright. For further details visit www.emeraldinsight.com/ copyright Online publishing and archiving As well as current volumes of the journal, you can also gain access to past volumes on the internet via Emerald Management Xtra. You can browse or search these databases for relevant articles. Key readings This feature provides abstracts of related articles chosen by the journal editor, selected to provide readers with current awareness of interesting articles from other publications in the field. Non-article content Material in our journals such as product information, industry trends, company news, conferences, etc. is available online and can be accessed by users. Reference linking Direct links from the journal article references to abstracts of the most influential articles cited. Where possible, this link is to the full text of the article. E-mail an article Allows users to e-mail links to relevant and interesting articles to another computer for later use, reference or printing purposes.
Xtra resources and collections When you register your journal subscription online you will gain access to additional resources for Authors and Librarians, offering key information and support to subscribers. In addition, our dedicated Research, Teaching and Learning Zones provide specialist ‘‘How to guides’’, case studies and interviews and you can also access Emerald Collections, including book reviews, management interviews and key readings. E-mail alert services These services allow you to be kept up to date with the latest additions to the journal via e-mail, as soon as new material enters the database. Further information about the services available can be found at www.emeraldinsight.com/alerts Emerald Research Connections An online meeting place for the world-wide research community, offering an opportunity for researchers to present their own work and find others to participate in future projects, or simply share ideas. Register yourself or search our database of researchers at www.emeraldinsight.com/connections
Choice of access Electronic access to this journal is available via a number of channels. Our web site www.emeraldinsight.com is the recommended means of electronic access, as it provides fully searchable and value added access to the complete content of the journal. However, you can also access and search the article content of this journal through the following journal delivery services: EBSCOHost Electronic Journals Service ejournals.ebsco.com Informatics J-Gate www.j-gate.informindia.co.in Ingenta www.ingenta.com Minerva Electronic Online Services www.minerva.at OCLC FirstSearch www.oclc.org/firstsearch SilverLinker www.ovid.com SwetsWise www.swetswise.com
Emerald Customer Support For customer support and technical help contact: E-mail
[email protected] Web www.emeraldinsight.com/customercharter Tel +44 (0) 1274 785278 Fax +44 (0) 1274 785201
JMP 23,3
208
Journal of Managerial Psychology Vol. 23 No. 3, 2008 p. 208 # Emerald Group Publishing Limited 0268-3946
EDITORIAL ADVISORY BOARD
EDITORIAL REVIEW BOARD
Professor Neil Anderson University of Amsterdam, The Netherlands Professor Chris Argyris Harvard Graduate School of Business Administration, USA Professor Yehuda Baruch Norwich Businesss School, University of East Anglia, UK Professor Frank Bournois Universite´ Panthe´on-Assas, Paris II, France Professor Cary L. Cooper Lancaster University Management School, Lancaster University, UK Professor Dov Eden Tel Aviv University, Israel Martin Euwema University of Utrecht, The Netherlands Professor Adrian Furnham University College London, UK Professor Hugh P. Gunz Joseph L. Rotman School of Management, University of Toronto, Ontario Dr Frank Heller Tavistock Institute, UK Professor Geert Hofstede CentER for Economic Research, University of Tilburg, The Netherlands Professor Paul Iles Leeds Metropolitan University, UK Professor Jim Jawahar Illinois State University, USA Professor Andrew Kakabadse Cranfield School of Management, UK, Founding Editor of Journal of Managerial Psychology Dr Bruce Kirkcaldy International Centre for the Study of Occupational and Mental Health, Germany Professor Harold J. Leavitt Stanford University, USA Professor Manuel London State University of New York, NY, USA Professor Dr Wolfgang Mayrhofer Vienna University of Economics and Business Administration, Austria Professor Greg Northcraft College of Business, University of Illinois, USA Dr Francisco Gil Rodriguez Universidad Complutense de Madrid, Spain Wilmar B. Schaufeli, PhD Research Institute for Psychology & Health, Utrecht University, Utrecht, The Netherlands Dr Chay Yue Wah SIM University, Singapore
Professor Yochanan Altman London Metropolitan University, UK Prof. dr. Arnold Bakker Erasmus University Rotterdam, The Netherlands Dr Dean Bartlett London Metropolitan University, UK Dr Gayle Baugh University of West Florida, USA Professor Ce´leste Brotheridge Universite´ Du Que´bec a` Montre´al, Canada Dr Adrian Carr University of Western Sydney, Australia Professor Kerry Carson University of Louisiana at Lafayette, USA Dr Alf Crossman The University of Surrey, UK Dr Patricia Hind Ashridge Business School, UK Professor Henry S.R. Kao University of Hong Kong, Hong Kong Dr Christian Kiewitz University of Dayton, USA Dr Ute-Christine Klehe Programmagroep A&O Psychologie, Amsterdam, The Netherlands Professor Steven D. Maurer Old Dominion University, USA Dr Ioannis Nikolaou Athens University of Economics and Business, Greece Dr Chris Rees University of Manchester, UK Dr Ramon Rico Universidad Autonoma de Madrid, Spain Professor Alain M. Roger IAE de Lyon, Universite´ Jean Moulin, Lyon Dr Raymond Saner Centre for Socio-Economic Development, Geneva, Switzerland Dr Rene´ Schalk Tilburg University, The Netherlands Dr Ruth Simpson Brunel University, UK Lynda J. Song Renmin University of China, China Dr Sherry E. Sullivan Bowling Green State University, USA Professor Dr Danie¨l Vloeberghs University of Antwerp and K.U. Leuven, Belgium Dr Lichia Yiu Centre for Socio-Economic Development, Geneva, Switzerland
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0268-3946.htm
INTRODUCTION
Introduction
Complexities and challenges in the work-family interface Noreen Heraty and Michael J. Morley
209
University of Limerick, Limerick, Ireland, and
Jeanette N. Cleveland The Pennsylvania State University, University Park, Pennsylvania, USA Abstract Purpose – The purpose of this brief paper is to introduce the papers in this special issue of Journal of Managerial Psychology, focused on “Complexities and challenges in the work-family interface”. Design/methodology/approach – The paper first introduces the theme of the special issue, and a brief outline of each paper contained in it is given. Findings – There is concern that progress in the work-family research area has been somewhat restricted and may have failed to take sufficient account of the complexity of work-family issues. Originality/value – The literature on the work-family interface is complex, and theory in the field is uncertain and under-developed. The papers in this special issue should further understanding of the challenges and complexities underscoring the work-family interface. Keywords Family, Sociology of work Paper type Viewpoint
Work-family relationships are complex and multidimensional and remain an important ongoing academic and social policy area that requires multidisciplinary and multi-level investigation and collaboration. Growing out of research on inter-role conflict (Kahn et al., 1964), historically work-family research has tended to focus on relationships between specific work and family variables and usually from either a family-focused or a work-focused perspective. This has resulted in an expanding body of research coalescing around the perceived ability of individuals to control stressors stemming from one or other of the work and family domains. In their review of 190 work-family studies published in IO/OB, Eby et al. (2005, p. 180) note that while there is a growing body of research to suggest that work and family can positively influence one another, there is far more that points to a negative spillover in terms of work-family conflict. The literature on the work-family interface is complex on several accounts. Theory in the field is uncertain and underdeveloped. Concepts, often loosely defined, abound and many of the relationships between them are not well understood. Empirical efforts are variable and the implications, especially for organisations, are often unclear. Despite these complexities, however, the work-family interface is an area of immense importance, personally, professionally and socially, as increasing numbers of families attempt to juggle work and family commitments and experience underlying difficulties in so doing. At the individual level, tensions in the work-family interface have been adjudged to affect, inter alia, stress, well-being, reduced spousal and parental effectiveness, decreased life satisfaction and increased psychologically threatening
Journal of Managerial Psychology Vol. 23 No. 3, 2008 pp. 209-214 q Emerald Group Publishing Limited 0268-3946 DOI 10.1108/02683940810861347
JMP 23,3
210
activities. In the work and organisational sphere, critical issues include the potential for a negative impact on organisational commitment (affective and continuance), job performance, job satisfaction, absenteeism and turnover. At the societal level, concerns relate to family disruption and community disconnect, reduced social citizenship and community engagement. Against this backdrop, there is an ongoing concern that progress in the work-family research area has been somewhat restricted and may have failed to take sufficient account of the complexity of work-family issues. Voydanoff (1988, 2005) calls for a better reconceptualisation of the work-family field (to include non-paid work and non-traditional family structures) and better measures of work-family fit and balance; Kossek and Ozeki (1998) similarly call for more consistency and robustness in measurement, and better sampling techniques; Barnett and Hyde (2001) call for new ways of thinking about the work-family interface, which Rotondo et al. (2003) describe as a permeable boundary; while Zedeck and Mosier (1990) and Frone (2003) highlight developments in organisational strategies and policies for promoting work-family balance at the individual and organisational level. Moreover, there appears to be a dearth of research that focuses on the larger macro societal level within which work and family domains exist and which can play a highly influential role in the work-family interface. This special issue brings together six papers that further our understanding of the challenges and complexities underscoring the work-family interface, an interface which has been characterised as one of the primary social challenges of our era because of the perceived imbalance which people experience in these domain areas of their lives (Halpern, 2005). The range of issues tackled here is both timely and insightful and opens up new lines of enquiry in addition to deepening previous research efforts. Thematically, the papers address the following: . the development of direct measures of perceived work demand and family demand in order to overcome construct difficulties apparent in the literature heretofore; . the testing of an expanded model of effort-reward imbalance (ERI) in predicting work-life conflict; . the examination of self and partner perceptions of work family interference among dual earner couples; . the exploration of whether spillover from non-work to work contributes to individuals’ wellbeing; . the analysis of the consequences of family responsibilities for career success; and . the development of a new model of the adjustment of expatriate families to living abroad. Our first paper, by Scott Boyar, Carl Maertz, Donald Mosley and Jon Carr, explores the domain areas of work demand and family demand that are seen to impact work-family conflict (both work interfering with family, or WIF, and family interfering with work, or FIW). The authors note that while there is general agreement that increased demand leads to increased conflict, to date no direct measures of demand have been identified. Using a sample of 698 university employees, this paper seeks to develop new knowledge in this area. Demand is defined as a global perception of the level and intensity of responsibility within the work or family domain. Here, the authors argue
that this demand must be subjectively experienced for it to influence WIF or FIW. Moreover, they propose that work and family domain variables act to influence levels of perceived work and family demand. In contrast to earlier research, the authors propose that perceived work demand and perceived family demand fully mediate the relationship between work and family domain variables, and WIF and FIW. Finally, the authors consider the relationship between work/family centrality and various demand effects. Their results confirm that work domain and family domain variables do indeed influence perceived work and family demand. Although not fully supported, their results suggest that these two forms of demand partially mediate the effects of domain variables on conflict. Furthermore, the relative strength of the demand-conflict relationship was partly determined by values of work/family centrality. In our second contribution to this Special Issue, Gail Kinman and Fiona Jones propose an expanded model of effort-reward imbalance, over-commitment and work-life conflict based on a study of 1,108 university employees in the UK. A key focus of this paper is to enhance knowledge of the well-established effort-reward imbalance (ERI) model of job stress proposed by Siegrist (1996) by examining its performance as a predictor of perceived conflict between work and home. This ERI model suggests that employees’ perceived imbalance/inequity between the effort they put into their jobs and the rewards that they receive is experienced as stressful and is likely to compromise heath and well-being over time. It also predicts that this imbalance will be experienced more frequently by employees who are seen as over-committed to their work. Beyond this, the authors here test whether factors associated with support for work-life balance such as degree of integration/segmentation between work and home, and individual working practices (scheduling flexibility, organisational support for balance), account for additional variance in work-life conflict over that explained by the ERI components. Their results suggest that the components of the ERI model are powerful predictors of work-life conflict. They further found evidence that lack of schedule flexibility and higher levels of work-life integration (fewer boundaries) are independent risk factors that are considered likely to compromise work life balance suggesting the importance of monitoring specific working conditions and occupational cultures, and providing promising avenues for further research in this area. Our third contribution comes from Michelle Streich, Wendy Casper and Amy Salvaggio, and it examines the interesting angle of couple agreement about work-family conflict. The authors note that much research to date has focused on how workers experience and manage conflict between their work and family roles. The point of departure in this paper is that, while WIF is becoming increasingly better understood, it has not really been examined from the couple perspective. To this end the authors provide an empirical examination of self and partner perceptions of WIF among 224 dual earner couples. Using family systems theory to explain how individuals’ attitudes are affected by other family members’ attitudes and behaviours, the authors explore to what extent couples agree about their WIF, and whether this agreement has a moderating effect on individual level organisational commitment. Building on the established relationship between WFC and both affective and continuance commitment, the authors here are interested in exploring whether influences from the family system are seen to act as moderators on the relationship between organisational commitment and WFC. The issue of gender differences with
Introduction
211
JMP 23,3
212
respect to agreement on WIF is also taken up, in recognition that evidence to date suggests that women generally report more WIF than men. The results highlight substantial agreement among couples when rating the WIF experienced by both male and female partners, though agreement was found to be higher when rating female WIF. Interestingly, couple agreement about female WIF was found to moderate the relationship between her self-rated WIF and her continuance commitment (the relationship was weaker when agreement was high). This suggests that spouse support can act to reduce WFC and is in broadly in line with the general view that social support acts as a moderator of the stressor-strain relationship. Moreover, the authors posit that WIF contributes to decreased affective commitment and increased continuance commitment suggesting that those who experience higher WIF may well have poorer work performance overall. Studies investigating the relationship between work and specific non-work roles (excluding family roles) are relatively few and this sets the context for our next contribution by Pam Allis and Michael O’Driscoll. While there have been many investigations of conflict between the two domain areas of work and non-work, their paper examine the positive effects of non work-to-work facilitation on well-being in work family and personal domains and seeks to examine whether spillover from non-work to work can be seen to contribute to individual well-being. Focusing specifically on family and personal benefit activities, they test the relationship between time demands in these domains and conflict between these roles and work. Personal benefit activities are defined as those that an individual undertakes that are associated with feelings of freedom, intrinsic satisfaction, positive mood and with consequent benefit for well-being. The authors propose that there is a positive knock-on effect (facilitation) from these activities but there may also be time conflicts associated with them. Utilising the enhancement hypothesis associated with multiple roles, and with a sample of 938 New Zealand local government employees, they test whether participation in one domain area (work) is made easier because of the experiences, skills and resources gained or developed in different domains (family/personal benefits). They further investigate whether the time demands associated with these activities has a consequent effect on work conflict. Their results point to moderate levels of family to work facilitation and personal activities to work. They found that high psychological involvement in a non-work domain correlated positively with facilitation to the work domain. However, involvement in personal benefits activities does not significantly relate with positive well-being. They propose that individuals need to develop strategies to enhance facilitation levels across different domains since facilitation is positively associated with well-being. In our fifth paper, Wolfgang Mayrhofer, Michael Meyer, Michael Schiffinger and Angelika Schmidt take us through the influence of family responsibilities, career fields and gender on career success. The authors contend that family responsibilities are an important factor influencing the amount of time and energy individuals are able and willing to devote to work, i.e. their work centrality. In turn, work centrality is seen to be positively related to both objective and subjective dimensions of career success. They further contend that the work and career context of the individual constitute important factors when analysing the effects of family situations on career success. Here they identify job alternatives and changeability of work content and professional relationships as key factors in the work context. Using a sample of 305 business
school graduates, they test the consequences of family responsibilities for their career success, and the influence of career context variables and gender on this relationship. Overall the impact of family responsibility on work centrality is negative, though not significantly so for men. Furthermore, work centrality was positively linked with both subjective and objective career success, which points to a negative relationship between family responsibilities and objective and subjective career success via work centrality. The authors found support for the effect of contextual factors on the relationship between family situations and career success. Changeability, for example, was found to be positively related to work centrality, for both men and women. Finally, evidence of gender effects was found to exist throughout, and as expected in most instances. The authors conclude that the results underscore the important relationship between external factors such as family responsibilities and work-related factors such as work centrality and career success, and point to the role of HR systems in managing the work-family interface. Our final contribution to this special issue comes from Arno Haslberger and Chris Brewster, who provide a theoretical exploration of work-family conflict from the perspective of the expatriate family. As the number of people who are sent abroad by their organisation continues to grow, and with recent figures suggesting that about 60 per cent of these are accompanied by a spouse/partner and about 50 per cent by their children, there remains a critical need to understand the work-family interface in expatriate assignments. Indeed, the authors suggest that evidence from both expatriate and family research shows systematic omissions about work-family interface among this population. Citing the FAAR model of Family Adjustment and Adaptation Response by Patterson (1988), which examines the balancing process between demands on a family (stressors, strains and daily hassles) and its capabilities to cope with its demands (resource capabilities, coping behaviours, and meaning ascribed) the authors explore the process of adjustment among expatriate families. They note that the literature on expatriates and expatriate families cover only some of the demands and capabilities involved, and argue that although expatriate families face more demands than single expatriates, they also have a broader range of capabilities available to them, but the range of resources and coping behaviours are less well researched. The authors contend that the family is under-represented in studies of expatriation and that there is considerable scope for crossover and spillover effects that are likely to affect the adjustment process. Furthermore, the integration of ideas from family systems theory may bring additional insights into the dynamics of expatriate adjustment. They call for greater theoretical and empirical research on expatriates from the level of the individual to the level of the family unit to gain additional insights into the process of cross-cultural adjustment. This should allow for the better management of, and support for, expatriate assignments that involve families and draws attention to the work-family interface, and the potential for work-family conflict arising from international assignments and global careers. References Barnett, R.C. and Hyde, J.S. (2001), “Women, men, work, and family: an expansionist theory”, American Psychologist, Vol. 56 No. 10, pp. 781-96. Eby, L., Casper, W., Lockwood, A., Bordeaux, C. and Brinley, A. (2005), “Work and family research in IO/OB: content analysis and review of the literature (1980-2002)”, Journal of Vocational Behavior, Vol. 66, pp. 124-97.
Introduction
213
JMP 23,3
214
Frone, M.R. (2003), “Work-family balance”, in Quick, J.C. and Tetrick, L.E. (Eds), Handbook of Occupational Health Psychology, American Psychological Association, Washington, DC. Halpern, D.F. (2005), “Psychology at the intersection of work and family: recommendations for employers, working families and policy makers”, American Psychologist, Vol. 60, pp. 397-409. Kahn, R.L., Wolfe, D.M., Quinn, R., Snoek, J.D. and Rosenthal, R.A. (1964), Organizational Stress: Studies in Role Conflict and Ambiguity, Wiley, New York, NY. Kossek, E.E. and Ozeki, C. (1998), “Work-family conflict, policies, and the job-life satisfaction relationship: a review and directions for organizational behavior-human resources research”, Journal of Applied Psychology, Vol. 83, pp. 39-149. Rotondo, D.M., Carlson, D.S. and Kincaid, J.F. (2003), “Coping with multiple dimensions of work-family conflict”, Personnel Review, Vol. 32 No. 3, pp. 275-96. Voyandoff, P. (1988), “Work and family: a review and expanded conceptualisation”, Journal of Science Behavior and Personality, Vol. 3, pp. 1-22. Voyandoff, P. (2005), “Towards a conceptualisation of perceived work-family fit and balance: a demands and resources approach”, Journal of Marriage and Family, Vol. 67, pp. 822-36. Zedeck, S. and Mosier, K.L. (1990), “Work in the family and employing organization”, American Psychologist, Vol. 45 No. 2, pp. 240-51. Corresponding author Noreen Heraty can be contacted at: noreen.heraty.ul.ie
To purchase reprints of this article please e-mail:
[email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0268-3946.htm
The impact of work/family demand on work-family conflict
Impact of work/ family demand
Scott L. Boyar Department of Management and Entrepreneurship, Williams College of Business, Xavier University, Cincinnati, Ohio, USA
215
Carl P. Maertz Jr Department of Management, John Cook School of Business, Saint Louis University, St Louis, Missouri, USA
Donald C. Mosley Jr Department of Management, Mitchell College of Business, University of South Alabama, Mobile, Alabama, USA, and
Jon C. Carr Department of Management & Marketing, University of Southern Mississippi, Hattiesburg, Mississippi, USA Abstract Purpose – The current study seeks to argue that the constructs of work demand and family demand have been neglected in the work-family conflict (WFC) literature. The authors aim to help clarify the definition and utilize direct measures of perceived work and family demand to test main effect, mediated, and interactive hypotheses. Design/methodology/approach – A sample of 698 university employees participated in a comprehensive computer survey that considered various manifest indicators and multiple scales across work and family domains. Moderator hierarchical regression and LISREL 8.0 were used in analyzing the data. Findings – The results indicate that both forms of demand have significant direct effects on work interfering with family (WIF) and family interfering with work (FIW). Both demand constructs partially mediate the effects of three categories of domain variables on the two forms of conflict. Finally, the work demand-WIF relationship is found to be stronger for those with relatively high family centrality. Research limitations/implications – A cross-sectional design was used and may be problematic when examining relationships that occur over time. Further, capturing all scales with a single survey could result in common method bias, which may have inflated the predictive relationships. Practical implications – Organizations can work to reduce WFC by adopting family-friendly programs that help employees balance work and family demands. Specifically, this study implies that organizations should find ways to hold constant or reduce perceptions of work and family demand, along with other direct antecedents of WIF and FIW. Originality/value – This study provides a relatively comprehensive model of antecedents that can be useful in future research. The authors also examine interactive effects of demand and work-family centrality on conflict using direct measures of perceived demand. Methodologically, the research improves on some past studies by measuring perceived demand directly and by not narrowing our sample to employees who are married or those with children. Hopefully, these contributions will help stimulate continued growth in the work-family literature. Keywords Family, Sociology of work, Role conflict Paper type Research paper
Journal of Managerial Psychology Vol. 23 No. 3, 2008 pp. 215-235 q Emerald Group Publishing Limited 0268-3946 DOI 10.1108/02683940810861356
JMP 23,3
216
Work and family are the key domains of life to many people (Whitely and England, 1977), and, not surprisingly, work-family conflict (WFC) research has become a major area in organizational research (Parasuraman and Greenhaus, 2002). In addition, this is an important area for practitioners because work-family conflicts have been empirically related to negative work attitudes (Frone et al., 1992a; Parasuraman et al., 1996; Yang et al., 2000), absenteeism (Goff et al., 1990), tardiness (Hepburn and Barling, 1996), leaving work early (Boyar et al., 2005), turnover intentions (Burke, 1988), and other negative work behaviors (Frone et al., 1996). Moreover, in the USA, data suggest trends toward reduced percentage of traditional family structures (US Bureau of Labor Statistics, 2000), an increasing number of dual-worker families (US Bureau of Labor Statistics, 1999), a growing number of working single mothers (Paulin and Lee, 2002), and an increased need to provide elder care for aging family members (Erwin, 2000). These trends and an individual’s limited time and energy combine to ensure that work and family roles are likely to conflict for some time into the future, and these conflicts need to be fully understood. Two separate forms of work-family conflict – work interfering with family (WIF) and family interfering with work (FIW) – have been shown to negatively impact organizational, family, and personal outcomes. In order to better understand and control both forms of conflict, researchers have deliberately focused on identifying their antecedents. Researchers have successfully demonstrated that some antecedents are related to both WIF and FIW (e.g. family support, age, number of children) (Frone et al., 1992b, 1996, 1997b; Kinnunen and Mauno, 1998; Netemeyer et al., 1996; Parasuraman et al., 1996). However, to confirm the construct validity of WIF and FIW and to better target management interventions for each, the unique antecedents of each must be clearly delineated. Research suggests that omitted mediators and moderators may be a major reason for the inability to consistently identify causal antecedents of related constructs (e.g. Frone et al., 1992a). We contend that work demand and family demand are foremost among the most important yet problematic factors surrounding WIF and FIW. Specifically, there has been inadequate conceptual work and measurement on these demand constructs, which has led to a lack of explicit consideration in work-family conflict models. The purpose of this study is to address these limitations by clarifying the constructs of work and family demand and their relationship to WIF and FIW. We test direct, mediated, and moderated effects involving these constructs. In doing so, we provide a clear picture of the antecedents for WIF and FIW. First, though, we review some background literature and describe the contributions of this paper in more detail.
Background and contributions Work-family conflict is defined as “a form of interrole conflict in which the role pressures from the work and family domains are mutually incompatible in some respect” (Greenhaus and Beutell, 1985, p. 77). Current research has demonstrated that two types of work-family conflict exist: (1) “work interfering with family” conflict (WIF); and (2) “family interfering with work” conflict (FIW) (e.g. Carlson et al., 2000; Frone et al., 1992a, 1996; Gutek et al., 1991; Netemeyer et al., 1996).
Researchers have come to implicitly agree that increases in demand are a primary cause of WIF and FIW (e.g. Carlson and Kacmar, 2000; Parasuraman et al., 1996). Despite its purported importance, work and family demand has been poorly conceptualized and rarely directly measured (cf. Boyar et al., 2007; Voydanoff, 1988; Yang et al., 2000). Researchers have claimed to measure work or family demand when they actually measure hours worked (Frone, 2000), marital status and family work hours (Voydanoff, 1988), or number of children (Kinnunen and Mauno, 1998). Many of these are more accurately conceived as demographic or situational causes of demand, or possibly surrogate measures of demand, rather than direct measures of demand itself. First, objective measures of demand, such as number of children, do not account for other ameliorating variables that affect the subjective experience of demand. For instance, more children may generally increase one’s family demand, but older children may offer significant help and support that reduces demand (e.g. Adams et al., 1996). The same argument can be made for marital status. Second, objective background variables may be perceived differently and may not lead to conflict at all (Martins et al., 2002; Parasuraman et al., 1996), unlike work or family demand. This lack of precision in measurement may have obscured the true effects of demand constructs on both WIF and FIW. Therefore, directly capturing individual perception of demand in each domain is necessary to ensure that the construct is adequately measured. Moreover, definitions of work demand or family demand have been ambiguous or too narrow; definitions often simply reflect role overload, which is based on having a negative response to work pressures. For example, Yang et al. (2000) define work demand as “pressures arising from excessive workloads and typical workplace time pressures such as rush jobs and deadlines” (p. 114). Role overload only reflects demand at extremely high levels of the continuum and is associated with a negative affect (Karasek, 1979). More specifically, role overload assesses the negative component of an individual’s workload; thus, low levels of role overload do not necessarily mean low levels of demand, whereas perceived demand covers a wider range of levels and possible affective responses. Because work demand and family demand are such critical constructs in the area, clear definitions distinct from overload are absolutely necessary so that they can be adequately measured and causal relationships tested. The current study To remedy these shortcomings, we incorporate a clear and broad definition and measure of work and family demand and examine the linkages among these measures to WIF and FIW. Predictors frequently used in the WFC literature were chosen for the analyses along with the two forms of conflict. Additionally, we assess the mediating effect of perceived demand for work and family domain variables (e.g. hours worked, work and family conflict, and social support) on WIF and FIW. We use regression to examine this possibility for both domains when most use correlation analysis. We also examine interactive effects of demand and work-family centrality on conflict using direct measures of perceived demand. Methodologically, we improve on some past studies by measuring perceived demand directly and by not narrowing our sample to employees who are married or those with children as other studies have, which skews the results and limits generalizability (Geurts et al., 2003; Parasuraman and Greenhaus, 2002). Finally, we help to illustrate the importance and establish the place of work and family demand in the nomological net surrounding work-family conflicts.
Impact of work/ family demand
217
JMP 23,3
218
Definition of demand Here, we define demand as “a global perception of the level and intensity of responsibility within the work (or family) domain” (Boyar et al., 2007). It is important to note that this demand is not an objective characteristic of the work or family domain – that is, demand must be subjectively experienced by the employee for it to influence WIF or FIW. Thus, it is actually perceived work (PWD) and family demand (PFD) that are the constructs of interest. Measures should capture an employee’s overall reaction to the level and intensity of responsibilities that may be affectively positive, negative, or neutral. Hypotheses Generally, we propose that work and family domain variables influence levels of perceived work and family demand. Increases in demand mediate the effects of these domain variables on WIF and FIW. Finally, these demand-conflict relationships are subject to moderators like values and roles related to the work and family domains (see Figure 1). Following Judge and Colquitt’s (2004) framework, we organize the domain variables into three categories for each domain: (1) responsibilities and expectations (e.g. role characteristics such as hours worked); (2) psychological demands (e.g. role stressors such as work role conflict); and (3) organizational policies and activities (e.g. social support). While many variables fit into each category, for parsimony, we use the examples provided by Judge and Colquitt (2004) in our analyses and refer to them as role characteristics, role stressors, and social support. Work domain variables and perceived work demand For work characteristics impacting work demand, we examine several important variables. Working more hours means that the employee is at work for more hours and may have more work duties and has less time for other activities (Frone et al., 1997b). As hours increase, so should one’s level of perceived work demand (Greenhaus et al., 1987). Supervisors are often held accountable for the work of their employees and are given more responsibility to manage employees; thus, one would expect supervisors to experience more work demand (Frone, 2000). As income increases, one would expect the level of responsibility to increase as well (Beutell and Wittig-Berman, 1999) and thereby having more work demand. Furthermore, having autonomy and the ability to manage one’s job activities should reduce demand levels by giving the employee the freedom to handle work demand without as many restrictive schedules or demands (Greenhaus et al., 1987; Karasek, 1979; Parasuraman et al., 1996). Being able to adjust ones schedule should allow the employee to manage his/her time in completing work duties. Thus: H1a. Hours worked will be positively related to perceived work demand. H1b. Supervisory status will be positively related to perceived work demand. H1c. Income will be positively related to perceived work demand. H1d. Work autonomy will be negatively related to perceived work demand.
Impact of work/ family demand
219
Figure 1. Proposed WFC model
In addition to work characteristics, role stressors should also directly affect an employee’s level of demand (Voydanoff, 1988). Receiving conflicting direction from multiple sources or working in ambiguous work environment may affect employee work activities and be perceived as work demand. Role conflict is defined as the “[S]imultaneous occurrence of two (or more) sets of pressures such that compliance with one would make more difficult compliance with the other” (Kahn et al., 1964, p. 19). Role ambiguity occurs when individuals are unsure of what it expected of them for a given role (House et al., 1983). Thus, as role conflict and role ambiguity increase, perceived level of demand should also increase. For social support, researchers have considered the relationship between work social support and WIF, but no study has directly considered support as an independent variable along with objective and perceptual work demand variables in determining an individual’s level of work and family demand. Frone et al. (1997b), however, demonstrated that work-role overload mediated the relationship between
JMP 23,3
220
instrumental supervisor support and WIF. Social support is generally considered a buffer to strain (Greenhaus and Beutell, 1985; House et al., 1983) and it may reduce demand levels while stressors and other variables discussed will increase demand levels. Supervisors can be supportive through encouragement, recognition, and resource allocation as they relate to work activities. Generally, as work support increases for the employee, level of work demand should decrease. In addition to work support, organizations can be family-friendly and provide work-to-family social support. Being supportive of employee family obligations and providing, for example, flexible work schedules should reduce perceived work demand. Therefore: H1e. Work role conflict will be positively related to perceived work demand. H1f. Work role ambiguity will be positively related to perceived work demand. H1g. Work social support will be negatively related to perceived work demand. H1h. Work-to-family social support will be negatively related to perceived work demand. Family domain variables and perceived family demand For family role characteristics impacting family demand, several variables are considered. Hours spent providing care in the family (Carlson and Perrewe, 1999; Frone et al., 1997b; Gutek et al., 1991; Parasuraman et al., 1996) have been shown to significantly affect FIW. Many researchers would argue that number of children should directly affect demand (see Parasuraman et al., 1996). Intuitively, being responsible for children in the home requires allocating time and energy in providing for their day-to-day needs. However, results have not been conclusive (Frone et al., 1996). Nevertheless, Netemeyer et al. (1996) and Kossek et al. (2001) found significant relationships between the number of children living at home and FIW, lending indirect support. Also, family should include dependent relationships (e.g. grandparents) that might directly affect family demand levels (Rothausen, 1999). While the results for marital status (Frone et al., 1996) have not yielded strong findings with WIF and FIW (Frone, 2000; Martins et al., 2002), it should be a strong predictor of perceived family demand. Married individuals should have more family demand in the form of obligations to a spouse that individuals who are not married do not have: H2a. Hours in caregiving will be positively related to family demand. H2b. Number of children living at home will be positively related to family demand. H2c. Number of dependents (other than children) living at home will be positively related to family demand. H2d. Being married will be positively related to family demand. Like work role stressors, family role conflict and role ambiguity should increase the resources required to obtain needed clarity that results in more time spent on family (Carlson and Kacmar, 2000). Specifically, having more stress in the family domain should result in higher levels of family demand. Further, family can provide support for family responsibilities or work obligations potentially reducing family demand. In fact, Frone et al. (1997b) found that family-role
overload (extreme demand) mediated the relationship between instrumental spouse support and FIW. In addition to reducing family demand, family members may work to reduce family obligations that specifically free energy and time for work duties, lowering family demand. Thus:
Impact of work/ family demand
H2e. Family role conflict will be positively related to family demand. H2f. Family role ambiguity will be positively related to family demand. H2g. Family social support will be negatively related to family demand. H2h. Family-to-work social support will be negatively related to family demand. Direct effects of demand on conflict It has been argued previously that an increase in work demand may lead work to interfere with family (Carlson and Kacmar, 2000). The employee may see that increases in work demand make completing family responsibilities more difficult, potentially causing a conflict which is attributed to work (i.e. WIF). In the same way, as family demand increases, putting more pressure on the individual to attend to family-related activities, work activities may suffer, causing perceived FIW. Therefore: H3. Work demand will be positively related to WIF. H4. Family demand will be positively related to FIW. Demand as a mediator Because researchers have not measured demand directly, the question of whether role characteristics and stressors lead to WIF and FIW directly or indirectly through work and family demand remains unanswered. Researchers have so far suggested that work and family role characteristics are directly related to WIF and FIW, but their arguments have typically suggested intermediate increases in demand levels in the causal chain (e.g. Voydanoff, 1988). Further, domain-specific support should reduce an individual’s demand level, and thus should be directly related to demand rather than to conflict directly. Thus, in contrast to earlier research, we propose that perceived work demand and perceived family demand will fully mediate the relationship between work and family domain variables and WIF and FIW, respectively. Carlson and Perrewe (1999) showed that the support-conflict relationship was mediated by role stressors, providing some indirect evidence for our hypothesized mediated relationship. H5a. The relationship between work domain variables (role characteristics, stressors, and support) and WIF will be fully mediated by perceived work demand. H5b. The relationship between family domain variables (role characteristics, stressors, and support) and FIW will be fully mediated by perceived family demand. Work/family centrality £ demand effects In addition to work role centrality, researchers have considered central life interest (Dubin, 1956; Dubin et al., 1975), values in life roles (Brown and Crace, 1996), and life role values (Carlson and Kacmar, 2000) in attempting to capture those aspects of life which
221
JMP 23,3
222
are valued and deemed important. Recently, Carlson and Kacmar (2000) used Whitely and England’s (1977) measure to capture employee’s life role values focusing on these main categories. However, independently capturing work centrality and family centrality poses difficulties. For these constructs, no relative comparisons across individuals can be made because there are no meaningful absolute standards. A relative scale capturing the importance of work versus family and family versus work provides a basis for comparison and hypothesizing. While others have suggested it (e.g. Burke, 1988), Carlson and Kacmar (2000) were the first to integrate a relative work-family values scale into the area. They did not, however, propose specific hypotheses. Nevertheless, they found that when work was more central in the respondent’s life, family domain antecedents had a significant impact on FIW. When family was valued as more central, work domain antecedents impacted WIF. It seems plausible that individuals who value work over family or family over work might react differently to work and family demand levels. Specifically, we propose that they may attribute the blame for experienced conflict to the domain which they value relatively less (Feldman, 1981; Weiner, 1985). Such self-serving bias works to maintain or enhance one’s self-image or identity (Arkin et al., 1980). In other words, individuals will tend to see the less valued domain as the cause rather than the more valued one. Therefore: H6a. The relationship between work demand and WIF will be stronger when individuals are relatively low on work/family centrality (i.e. high on family). H6b. The relationship between family demand and FIW will be stronger when individuals are relatively high on work/family centrality (i.e. high on work). Method Sample The survey was provided to all 3,498 university employees; 698 completed the questionnaire, providing a 20 percent response rate. Most were women (66 percent) and married (78.5 percent). Most (85 percent) were Caucasian, 10 percent were African-American, and the rest were either Latino, Native Americans, or mixed race. The average age was 42.7 years. Many of the participants had a graduate degree (48.3 percent), while many others had some college or a college degree (39.5 percent). Procedure Data for this study was collected over a two-week period. Potential participants were sent a request through e-mail asking for their fully voluntary participation in the study and inviting them to visit a specified website. At that site, measures were collected using a computer survey. Measures Most items, other than demographics, were captured using a five-point Likert-type scale with responses ranging from strongly disagree (1) to strongly agree (5). Dependent variables. Work-interfering with family (WIF) and family interfering with work (FIW) items were culled from established scales (e.g. Carlson et al., 2000). A sample item for WIF is “My work often interferes with my family responsibilities”. A sample item for FIW is “My family responsibilities prevent me from effectively performing my job”. Confirmatory factor analysis (CFA) employing LISREL 8
(Joreskog and Sorbom, 1996) was used to examine the latent constructs. The model with the two scales produced a x 2 with eight degrees of freedom of 35.35 (p ¼ 0:00) and a GFI of 0.98, an AGFI of 0.96, and an RMSEA of 0.07. The alphas for WIF and FIW were 0.94 and 0.86, respectively. Antecedent variables. Hours worked per week was kept as a continuous variable. Supervisory status was coded 1 (supervisory responsibilities) or 2 (no supervisory responsibilities). Annual salary was self-reported on the survey. Work autonomy was assessed using three items from Sims et al.’s (1976) autonomy scale. “To what extent are you able to do your job independently of others?”. Researchers have shown adequate internal reliability coefficients of 0.84 and high convergent validity correlations (r ¼ 0:72) (Aiken and Hage, 1966). The alpha in this sample was 0.72. Participants were asked to report how many children and other dependents are presently living with them. Each was measured as a continuous variable. Marital status was coded 1 or 2 into two groups, one for those single or living alone and another for those married or living with a significant other. Work role conflict (WRC), work role ambiguity (WRA), family role conflict (FRC), and family role ambiguity (FRA) were measured using items adapted from Rizzo et al. (1970) and House et al. (1983). The alpha for WRC and WRA was 0.75 and 0.84, respectively. The alpha for FRC and FRA was 0.68 and 0.68, respectively. Items from Thomas and Ganster (1995) and Cutrona and Russell (1987) were used to measure the perceived level of support provided to employees from supervisors. An example item is “My supervisor is understanding or sympathetic”. Nine items adapted from King et al. (1995) were used to measure the level of support from family members and both work-to-family and family-to-work support. An example item for family support is “Members of my family are willing to straighten up the house when it needs it”. An example item for work-to-family is “My supervisor holds my family responsibilities against me” (reverse coded) and for family-to-work support is “If my job gets very demanding, someone in my family will take on extra household duties”. The alphas for work support and family support were 0.89 and 0.86, respectively. The alphas for work-to-family support and family-to-work support were 0.75 and 0.86, respectively. Work and family demand. Perceived work demand (PWD) and perceived family demand (PFD) scales were measured with scales developed and validated by Boyar et al. (2007). The alpha for the five-item work demand scale was 0.89. A sample item is “My job requires all of my attention”. It produced a x 2 with five degrees of freedom of 13.85 (p ¼ 0:17), a GFI of 0.99, an AGFI of 0.98, and an RMSEA of 0.05. The four-item family demand scale had an alpha of 0.77. A sample item is “I have a lot of responsibility in my family”. It produced a x 2 with two degrees of freedom of 0.37 (p ¼ 0:83), a GFI of 1.00, an AGFI of 0.99, and an RMSEA of 0.01. Work/family centrality. A work centrality scale developed by Paullay et al. (1994) was adapted to capture work/family centrality in a relative sense rather than for the work domain only. The scale consists of five items and had an alpha of 0.82. An example item is “The major satisfaction in my life comes from my work rather than my family”. Analysis Structural equation modeling (SEM) using LISREL 8 (Joreskog and Sorbom, 1996) was used to test direct and mediated effects, comparing nested models. Using procedures
Impact of work/ family demand
223
JMP 23,3
suggested by Baron and Kenny (1986) and implemented by Sapienza and Korsgaard (1996), we compared three models – a direct, an indirect, and a saturated model – with each other to determine whether PWD and PFD fully or partially mediated the relationship between their respective domain variables and conflict. Multiple regression was used to assess the interaction effects. Pairwise deletion method was employed.
224 Results The means, standard deviations, and correlations are presented in Tables I and II. As with previous research, WIF and FIW were highly and positively correlated (r ¼ 0:42; p ¼ 0:00). Figure 2 provides a summary of standardized path coefficients and significance levels for both work and family domain models. The main effects hypotheses between the work domain variables and perceived work demand received some support (see Table III). Hours worked (H1a), supervisory status (H1b), income (H1c), work role conflict (H1e), and work-to-family social support (H1h) were all supported at the p , 0:05 level of significance. However, autonomy (H1d ), work role ambiguity (H1f ), and work social support (H1g) were not significant in the hypothesized direction. See Table III for results of indirect models. While not supported, work role ambiguity (r ¼ 0:08; p ¼ 0:05) had significant and positive correlations with work demand, suggesting that a relationship may yet exist. Overall, most domain variables predicted work demand as expected. The main effects hypotheses between family domain variables and perceived family demand received mixed support (see Table IV). Hours spent providing care (H2a), number of children living at home (H2b), marital status (H2d ), and family role conflict (H2e) were supported at p , 0:05 level of significance. However, number of dependents living at home (H2c), family role ambiguity, family role social support, and family-to-work social support were not significantly related to perceived family demand, but family social support (r ¼ 20:17; p ¼ 0:00), and family-to-work social support (r ¼ 20:23; p ¼ 0:00) had a significant correlation with family demand. The main effects hypotheses between work demand and WIF (H3) and between family demand and FIW (H4) were both supported at the p ¼ 0:05 level of significance. The mediating hypotheses were tested using an SEM nested model comparison. First, the direct model was compared to the saturated model to determine if mediation exists. For the work domain, the x 2 difference of 26.42 (634:95 2 608:53) with one (258 2 257) degree of freedom was significant at the p ¼ 0:00 level. For the family domain, the x 2 difference of 47.84 (634:41 2 586:57) with one (190 2 189) degree of freedom was significant at the p ¼ 0:00 level. Therefore, for each model comparison, there is a significant difference, suggesting mediation exists. Second, the indirect model was compared to the saturated model to determine if demand fully mediated (i.e. indirect model) or partially mediated (saturated model) the relationship between exogenous variables and conflict. For the work domain, the x 2 difference of 158.29 (766:53 2 608:24) with eight (265 2 257) degrees of freedom was significant at the p ¼ 0:00 level. For the family domain, the x 2 difference of 46.80 (633:37 2 586:57) with eight (197 2 189) degrees of freedom was significant at the p ¼ 0:00 level. Therefore, the partially mediated model fit the data better than a fully mediated model for the both the work and family domains. While mediation exists, it is partial; thus, H5 was not supported.
Variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Mean
SD
WIF 2.57 1.06 FIW 2.13 0.85 PWD 3.79 0.84 PFD 3.07 0.84 Hours worked 45.26 9.21 Supervisory status 1.51 0.50 Income 40,282 24,403 Autonomy 4.12 0.56 Work role conflict 2.87 0.95 Work role ambiguity 2.23 0.91 Work social support 3.76 0.99 Work-to-family support 3.98 0.76 Hours in care 33.46 31.39 Children at home 1.18 1.20 Dependents at home 0.39 0.62 Marital status 0.85 0.36 Family role conflict 2.78 0.84 Family role ambiguity 1.84 0.64 Family social support 3.71 0.87 Family-to-work support 3.60 0.89 Age 42.68 9.85 Race 1.96 0.57 Education 6.87 1.43 Gender 0.66 0.47 Wk fam. centrality 2.05 0.76
n 697 696 697 696 688
1 0.42 * * 0.37 * * 0.22 * * 0.37 * *
2
0.17 * * 0.39 * * 0.04
3
4
5
6
7
8
2 0.21 * * 0.42 * * 20.21 * * 0.08 * 20.03
696
0.16 * *
0.23 * *
0.10 *
0.26 * *
20.06
0.10
2 0.10 * *
696
0.28 * *
0.14 * *
0.08 *
0.08 *
0.13 * *
0.03
2 0.03
2 0.15 * *
2 0.11 * * 20.04
2 0.06
0.09 *
2 0.01
0.02
695 2 0.28 * * 2 0.14 * * 2 0.11 * * 2 0.02 2 0.13 * * 20.01 598 2 0.06 0.07 2 0.08 * 0.24 * * 2 0.19 * * 0.07 586
0.09 *
0.10 *
333 691
0.04 0.12 * *
695 694
0.26 * *
0.01
0.04
0.07 2 0.05 0.11 * * 2 0.01
0.02 0.06
0.09 0.00
0.22 * *
0.28 * *
0.10 *
0.30 * *
0.12 * *
0.20 * *
0.07
695 2 0.04
0.04
2 0.05
2 0.14 * * 2 0.01
0.04
12
0.47 * * 2 0.31 * * 2 0.37 * *
0.03 0.15 * * 2 0.22 * * 2 0.28 * * 2 0.27 * * 2 0.01 0.05 2 0.01 0.00
0.45 * * 0.01
0.05
0.00
0.05
0.04
0.00
0.06
0.00 20.06
0.09 2 0.02 0.15 * * 0.01
0.08 2 0.01
0.01 0.01
2 0.03 0.01
2 0.03 0.05
0.08 *
20.03
0.02
2 0.03
0.12 * *
0.18 * *
0.05
0.08 *
0.00
0.05
2 0.05
0.06
0.15 * * 2 0.01
2 0.17 * *
0.09 *
20.04
0.02
2 0.06
2 0.09 *
2 0.06 2 0.08 * 0.03 0.12 * * 2 0.09 *
2 0.11 * * 0.05 2 0.07 2 0.10 * 2 0.02 0.04 0.00 2 0.03 2 0.03 0.04
0.13 * * 2 0.02 0.02 0.04 2 0.06
2 0.03
2 0.03
2 0.07
0.13 * *
694 2 0.17 * * 2 0.20 * * 2 0.06 2 0.23 * * 0.08 * 0.04 0.11 * * 0.08 * 678 2 0.02 2 0.08 * 0.02 2 0.13 * * 0.00 20.14 * * 0.26 * * 0.04 684 0.11 * * 0.05 0.12 * * 2 0.02 0.10 * 20.03 0.18 * * 2 0.04 * * * * * * * * * * * * 691 0.23 0.16 0.19 0.04 0.34 20.16 0.48 2 0.03 684 2 0.14 * * 2 0.06 2 0.11 * * 0.01 2 0.37 * * 0.13 * * 2 0.50 * * 2 0.02 696
11
0.06
0.39 * *
0.03
10
0.17 * * 0.33 * * 2 0.07
693 2 0.11 * * 2 0.02 2 0.16 * * 0.03 634 0.15 * * 0.12 * * 0.21 * * 2 0.07 * * * 695 2 0.12 2 0.09 2 0.02 2 0.03
697 2 0.17 * *
9
0.10 * * 2 0.12 * *
0.19 * * 20.04
0.17 * *
0.06
0.00
0.00
2 0.01
2 0.08 * 2 0.19 * * 0.10 *
Notes: *p , 0:05; * *p , 0:01
Impact of work/ family demand
225
Table I. Means, standard deviations, and correlations
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Mean
SD
WIF 2.57 1.06 FIW 2.13 0.85 PWD 3.79 0.84 PFD 3.07 0.84 Hours worked 45.26 9.21 Supervisory status 1.51 0.50 Income 40,282 24,403 Autonomy 4.12 0.56 Work role conflict 2.87 0.95 Work role ambiguity 2.23 0.91 Work social support 3.76 0.99 Work-to-family support 3.98 0.76 Hours in care 33.46 31.39 Children at home 1.18 1.20 Dependents at home 0.39 0.62 Marital status 0.85 0.36 Family role conflict 2.78 0.84 Family role ambiguity 1.84 0.64 Family social support 3.71 0.87 Family-to-work support 3.60 0.89 Age 42.68 9.85 Race 1.96 0.57 Education 6.87 1.43 Gender 0.66 0.47 Wk fam. centrality 2.05 0.76
Notes: *p , 0:05; * *p , 0:01
n
13
14
15
16
17
18
19
20
21
22
JMP 23,3
7 8 9
226
Table II. Means, standard deviations, and correlations Variables 1 2 3 4 5 6
23
24
697 696 697 696 688 693 634 695 696 696 697 695 598 586
0.27 * *
333 691
0.07 0.09 *
0.18 * * 0.10 *
695
0.01
0.02
0.07 2 0.06
2 0.09 *
694 2 0.19 * * 2 0.09 *
2 0.06
2 0.10 *
0.10 *
0.02
695
0.03
694 678 684 691 684
2 0.01 2 0.13 * * 2 0.08 * 2 0.19 * * 0.25 * *
0.33 * *
0.24 * * 2 0.38 * * 2 0.24 * *
0.64 * * 2 0.01 2 0.01 0.16 * * 2 0.44 * * 2 0.18 * * 2 0.23 * * 0.10 0.11 * 0.06 0.00 2 0.03 2 0.09 * 2 0.03 0.02 0.09 * 2 0.01 0.00 0.03 0.01 0.09 * 0.08 2 0.03 0.09 * 0.03 0.03 0.08 * 0.08 * 2 0.12 * 2 0.05 2 0.20 * * 2 0.07 2 0.03 2 0.11 * * 2 0.16 * * 2 0.10 * 2 0.06
696 2 0.21 * * 2 0.19 * *
0.03
2 0.17 * *
0.27 * *
0.34 * * 2 0.12 * * 2 0.11 *
0.23 * *
0.05 2 0.08 * 2 0.32 * * 0.01
0.02
2 0.15 * *
Impact of work/ family demand
227
Figure 2. Indirect mediation model
The results for the interaction hypotheses are provided in Table V. Supporting H6a, the interaction term was significant (p ¼ 0:04; see Figure 3). Perceived work demand exhibited a stronger relationship with WIF for individuals who value family as more central than work. However, for H6b, the interaction coefficient was not significant. Discussion The findings of this study indicated that work domain and family domain variables influence perceived work and family demand, respectively. The two forms of demand partially mediated the effects of domain variables on conflict. Finally, the strength of the demand-conflict relationship was partly determined by values of work/family centrality. Taken together, these findings help clarify the determinants of WIF and FIW. Above all, this study establishes the importance of perceived work and family demand scales, measured directly, in clarifying the important variables leading to WIF and FIW. Based on this, we conclude that perceived work and family demand scales
JMP 23,3
228
Table III. Model statistics and standardized path coefficients for work domain
Measure 2
x df GFI AGFI RMSEA Hours worked-PWD Supervisory status-PWD Income-PWD Autonomy-PWD Work role conflict-PWD Work role ambiguity-PWD Work support-PWD Work-to-family support-PWD Hours worked-WIF Supervisory status-WIF Income-WIF Autonomy-WIF Work role conflict-WIF Work role ambiguity-WIF Work support-WIF Work-to-family support-WIF Perceived work demand (PWD)-WIF
Direct
Indirect
Saturated
634.95 * * 258 0.93 0.91 0.046 0.22 * * 20.11 * * 0.09 * 20.06 0.31 * * 20.12 * 0.14 * * 20.10 * 0.24 * * 20.04 0.03 20.11 * * 0.38 * * 20.01 0.12 * * 20.26 * *
766.53 * * 265 0.92 0.90 0.052 0.24 * * 20.11 * * 0.09 * 0.06 0.27 * * 20.09 0.13 * * 20.10 *
608.24 * * 257 0.94 0.91 0.044 0.23 * * 2 0.11 * 0.09 * 2 0.06 0.26 * * 2 0.09 0.13 * * 2 0.09 * 0.19 * * 2 0.01 0.01 2 0.09 * 0.28 * * 0.03 0.09 * 2 0.23 * * 0.22 * *
0.42 * *
Notes: *p , 0:05; * *p , 0:01
could be useful in future theoretical models of work-family conflicts. In the following sections, we discuss the research implications of the findings in more detail. Research implications This study provides a needed organization and clarification of the numerous antecedents of work-family conflict. For the work domain, all three of the categories had at least one significant finding. For family, two of three categories had significant variables. These findings generally support and extend the Judge and Colquitt (2004) three-category antecedent framework by adding specific examples of significant variables for each category. The direct relationships of work demand-WIF and family demand-FIW were strong and significant and should provide new key causal linkages for future research to build upon. The finding of significant direct effects for both FIW and WIF implies that increases in demand within each domain may contribute to WFC. The finding of partial rather than full mediation for the demand construct suggests that certain domain variables may affect WIF or FIW directly. This finding suggests that while some variables (e.g. supervisory status) may lead to work demand as posited, other variables (e.g. work role conflict) may lead to both work demand and WIF. These are important variables that impact employees’ attitudes about work and may be more easily identified and associated with levels of conflict experienced at work, yet also directly affect one’s level of work demand. The findings also suggests that some variables (e.g. children at home) may lead to family demand as posited, while other variables (e.g. marital status and family role conflict) may lead to both family demand and FIW.
Measure x2 df GFI AGFI RMSEA Hours in care-PFD Children at home-PFD Dependents at home-PFD Marital status-PFD Family role conflict-PFD Family role ambiguity-PFD Family social support-PFD Family-to-work support-PFD Hours in care-FIW Children at home-FIW Dependents at home-FIW Marital status-FIW Family role conflict-FIW Family role ambiguity-FIW Family social support-FIW Family-to-work support-FIW Perceived family demand (PFD)-FIW
Direct
Indirect
Saturated
634.41 * * 190 0.93 0.89 0.058 0.16 * * 0.20 * * 2 0.01 0.18 * * 0.50 * * 2 0.09 2 0.11 0.04 0.05 0.05 0.10 * 0.25 * * 0.44 * * 0.06 2 0.01 0.02
633.37 * * 197 0.93 0.90 0.056 0.18 * * 0.21 * * 20.02 0.15 * * 0.38 * * 20.03 20.09 20.04
586.57 * * 189 0.93 0.90 0.055 0.18 * * 0.21 * * 2 0.03 0.14 * * 0.37 * * 2 0.04 2 0.09 2 0.04 0.01 2 0.02 0.10 * 0.16 * * 0.15 * 0.14 * * 0.05 2 0.06 0.35 * *
0.46 * *
Notes: *p , 0:05; * *p , 0:01
The significant moderation effect for work-family centrality implies that those who hold family as far more central than work may be more likely to perceive increases in demand as causing work interference, consistent with the attributional explanation discussed above. It makes sense that such individuals may be more sensitive to increasing work demand. Other potential moderators that may affect attributions for cause of conflict include caregiving versus financial family responsibility and attributions for past work-family conflicts. Practical implications Work-family conflict has been shown to have a negative impact on organizational, family, and personal outcomes (Frone et al., 1992a, 1997a). Organizations can work to reduce WFC by adopting family-friendly programs that help employees balance work and family demands. Specifically, this study implies that organizations should find ways to hold constant or reduce perceptions of work and family demand, along with other direct antecedents of WIF and FIW. For instance, organizations can reduce workloads, limit the number of roles assigned to employees, and provide resources and encourage support from coworkers and supervisors in completing work assignments. Further, it might be beneficial to redesign jobs to make them more interesting and challenging (Hackman and Oldham, 1980); if employees enjoy their work and find it valuable, then they may not mind working harder. In fact, they may not perceive high levels of demand (Boyar et al., 2007). The idea that the work-family interface can result in positive outcomes, known as enrichment, and future research can explore the relationship between work and family demand and enrichment (Greenhaus and Powell,
Impact of work/ family demand
229
Table IV. Model statistics and standardized path coefficients for family domain
B
p 0.06 *
Step 1 Age Race Education Gender
20.004 0.071 * * 0.191 * 20.078 *
0.924 0.062 0.000 0.050
Step 2 Age Race Education Gender PWD PFD Work-family centrality Cumulative model
0.011 0.050 0.133 * 20.065 * * 0.290 * 0.157 * 0.000 0.179
0.767 0.168 0.001 0.085 0.000 0.000 0.997
Step 3 Age Race Education Gender PWD PFD Work-family centrality PWD £ work-family centrality PFD £ work-family centrality Cumulative model
0.003 0.059 * * 0.136 * 20.072 * 0.268 * 0.166 * 20.017 20.080 * 0.098 * 0.192
0.928 0.101 0.000 0.054 0.000 0.000 0.645 0.037 0.007
a
WIF R2
DR 2
F
B
10.527 *
0.119
0.066 * * 0.050 0.121 * 2 0.018
0.026 0.081 0.197 0.003 0.658
2 0.036 0.054 0.093 * 2 0.012 0.062 * * 0.389 * 0.071 *
0.329 0.128 0.015 0.757 0.094 0.000 0.055
31.668 *
17.262 *
DR 2
0.005 2 0.040 0.060 * * 0.097 * 2 0.015 0.042 0.393 * 0.082 * 2 0.063 * * 0.051 * * * 0.192
F
4.372 *
43.129 *
21.462 *
0.187
5.257 *
Notes: n ¼ 664; *p , 0:05; * *p , 0.10; * * *p , 0:05 for one-tailed test
FIW R2
0.162
20.427 *
0.179 0.013
p
JMP 23,3
230
Table V. Moderator hierarchical regression results for work-family centrality Regressiona
0.276 0.094 0.011 0.681 0.285 0.000 0.029 0.098 0.164 17.220 *
2.116
Impact of work/ family demand
231
Figure 3. Interaction effect for H6a
2006). The moderation finding implies that such efforts may work better with employees who value family as more central. Organizations may be able to survey work-family centrality and other potential moderators (e.g. family role responsibilities) to help assess how effective efforts to reduce demand will be. Limitations and conclusion This study is not without limitations. In particular, the study utilized only one sample with a relatively low response rate. While we have no specific reason to expect that the sample is not representative, these sampling factors limit the generalizability of these findings. Thus, the results must be regarded with caution as they await replication. Second, conducting a study with a cross-sectional design is problematic when examining relationships that occur over time. A related problem is that capturing all scales with a single survey poses the problem of common method bias, which may have inflated the predictive relationships. However, several aspects of the study do ameliorate this concern somewhat. First, our demographic and interactive findings were not subject to the biases inherent in using one survey to collect all the data. In other words, there is no reason to expect mono-method biases would inflate correlations involving demographics or interaction regression coefficients. Second, for direct effects, we were testing theoretically strong arguments in our hypotheses bolstered by supportive findings. For this reason, we were less concerned with effect sizes (which might have been inflated) than with causal relationships that were likely to be statistically significant. Also, in modeling studies where causal relationships are emphasized, common method bias is not as a serious concern. Despite these limitations, the current study successfully establishes the importance of explicitly including the constructs of perceived work and family demand in theoretical models and measuring them directly in empirical model tests. This study also provides a relatively comprehensive model of antecedents that can be useful in
JMP 23,3
232
future research. Hopefully, these contributions will help stimulate continued growth in the work-family literature. References Adams, G.A., King, L.A. and King, D.W. (1996), “Relationships of job and family involvement, family social support, and work-family conflict with job and life satisfaction”, Journal of Applied Psychology, Vol. 81, pp. 411-20. Aiken, M. and Hage, J. (1966), “Organizational alienation: a comparative analysis”, American Sociological Review, Vol. 31, pp. 497-507. Arkin, R.M., Appelman, A.J. and Burger, J.M. (1980), “Social anxiety, self-presentation, and the self-serving bias in causal attribution”, Journal of Personality and Social Psychology, Vol. 38, pp. 23-35. Baron, R.M. and Kenny, D.A. (1986), “The mediator-moderator variable distinction in social psychological research: conceptual, strategic, and statistical considerations”, Journal of Personality and Social Psychology, Vol. 51, pp. 1173-82. Beutell, N. and Wittig-Berman, U. (1999), “Predictors of work-family conflict and satisfaction with family, job, career, and life”, Psychological Reports, Vol. 85, pp. 893-903. Boyar, S.L., Maertz, C.P. Jr and Pearson, A. (2005), “The effects of work-family conflict and family-work conflict on non-attendance behaviors”, Journal of Business Research, Vol. 58, pp. 919-25. Boyar, S.L., Carr, J.C., Mosley, D.C. Jr and Carson, C.M. (2007), “The development and validation of scores on perceived work and family demand scales”, Educational and Psychological Measurement, Vol. 67, pp. 100-15. Brown, D. and Crace, R.K. (1996), “Values in life role choices and outcomes: a conceptual model”, Career Development Quarterly, Vol. 44, pp. 211-24. Burke, R.J. (1988), “Some antecedents and consequences of work-family conflict”, Journal of Social Behavior and Personality, Vol. 3, pp. 287-302. Carlson, D.S. and Kacmar, K.M. (2000), “Work-family conflict in the organization: do life role values make a difference?”, Journal of Management, Vol. 26, pp. 1031-54. Carlson, D.S. and Perrewe, P.L. (1999), “The role of social support in the stressor-strain relationship: an examination of work-family conflict”, Journal of Management, Vol. 25, pp. 513-40. Carlson, D.S., Kacmar, K.M. and Williams, L.J. (2000), “Construction and initial validation of a multidimensional measure of work-family conflict”, Journal of Vocational Behavior, Vol. 56, pp. 249-76. Cutrona, C.E. and Russell, D.W. (1987), “The provisions of social relationships and adaptation to stress”, Advances in Personal Relationships, Vol. 1, pp. 37-67. Dubin, R. (1956), “Industrial workers’ worlds: a study of the ‘central life interests’ of industrial workers”, Social Problems, Vol. 3, pp. 131-42. Dubin, R., Champoux, J.E. and Porter, L.W. (1975), “Central life interests and organizational commitment of blue-collar and clerical workers”, Administrative Science Quarterly, Vol. 20, pp. 411-21. Erwin, S.L. (2000), “Fourteen forecasts for an aging society”, The Futurist, Vol. 34 No. 6, pp. 24-8. Feldman, J.M. (1981), “Beyond attribution theory: cognitive processes in performance appraisal”, Journal of Applied Psychology, Vol. 66, pp. 127-48.
Frone, M.R. (2000), “Work-family conflict and employee psychiatric disorders: the national comorbidity survey”, Journal of Applied Psychology, Vol. 85, pp. 888-95. Frone, M.R., Russell, M. and Barnes, G.M. (1996), “Work-family conflict, gender, and health-related outcomes: a study of employed parents in two community samples”, Journal of Occupational Health Psychology, Vol. 1, pp. 57-69. Frone, M.R., Russell, M. and Cooper, M.L. (1992a), “Antecedents and outcomes of work-family conflict: testing a model of the work-family interface”, Journal of Applied Psychology, Vol. 77, pp. 65-78. Frone, M.R., Russell, M. and Cooper, M.L. (1992b), “Prevalence of work-family conflict: are work and family boundaries asymmetrically permeable?”, Journal of Organizational Behavior, Vol. 13, pp. 723-9. Frone, M.R., Russell, M. and Cooper, M.L. (1997a), “Relation of work-family conflict outcomes: a four-year longitudinal study of employed parents”, Journal of Occupational and Organizational Psychology, Vol. 70, pp. 325-35. Frone, M.R., Yardley, J.K. and Markel, K.S. (1997b), “Developing and testing an integrative model of the work-family interface”, Journal of Vocational Behavior, Vol. 50, pp. 145-67. Geurts, S., Kompier, M., Roxburgh, S. and Houtman, I. (2003), “Does work-home interference mediate the relationship between workload and well-being”, Journal of Vocational Behavior, Vol. 63, pp. 532-59. Goff, S.J., Mount, M.K. and Jamison, R.L. (1990), “Employer supported child care, work/family conflict, and absenteeism: a field study”, Personnel Psychology, Vol. 43, pp. 793-809. Greenhaus, J.H. and Beutell, N.J. (1985), “Sources of conflict between work and family roles”, Academy of Management Review, Vol. 10, pp. 76-88. Greenhaus, J.H. and Powell, G.N. (2006), “When work and family are allies: a theory of work-family enrichment”, Academy of Management Review, Vol. 31, pp. 72-92. Greenhaus, J.H., Bedeian, A.G. and Mossholder, K.W. (1987), “Work experiences, job performance, and feelings of personal and family well-being”, Journal of Vocational Behavior, Vol. 31, pp. 200-15. Gutek, B.A., Searle, S. and Klepa, L. (1991), “Rational versus gender role explanations for work-family conflict”, Journal of Applied Psychology, Vol. 76, pp. 560-8. Hackman, J.R. and Oldham, G.R. (1980), Work Redesign, Addison-Wesley, Reading, MA. Hepburn, C.G. and Barling, L. (1996), “Eldercare responsibilities, interrole conflict, and employee absence: a daily study”, Journal of Occupational Health Psychology, Vol. 1, pp. 311-8. House, R.J., Schuler, R.S. and Levanoni, E. (1983), “Role conflict and ambiguity scales: reality or artifacts?”, Journal of Applied Psychology, Vol. 68, pp. 334-7. Joreskog, K. and Sorbom, D. (1996), LISREL 8 User’s Guide, University of Uppsala, Uppsala. Judge, T.A. and Colquitt, J.A. (2004), “Organizational justice and stress: the mediating role of work-family conflict”, Journal of Applied Psychology, Vol. 89, pp. 395-404. Kahn, R.L., Wolfe, D.M., Quinn, R., Snoek, J.D. and Rosenthal, R.A. (1964), Organizational Stress, Wiley, New York, NY. Karasek, R.A. Jr (1979), “Job demands, job decision latitude, and mental strain: implications for job redesign”, Administrative Science Quarterly, Vol. 24, pp. 285-307. King, L.A., Mattimore, L.K., King, D.W. and Adams, G.A. (1995), “Family support inventory for workers: a new measure of perceived social support from family members”, Journal of Organizational Behavior, Vol. 16, pp. 235-58.
Impact of work/ family demand
233
JMP 23,3
234
Kinnunen, U. and Mauno, S. (1998), “Antecedents and outcomes of work-family conflict among employed women and men in Finland”, Human Relations, Vol. 51, pp. 157-78. Kossek, E.E., Colquitt, J.A. and Noe, R.A. (2001), “Caregiving decisions, well-being, and performance: the effects of place and provider as a function of dependent type and work-family climates”, Academy of Management Journal, Vol. 44, pp. 29-44. Martins, L.L., Eddleston, K.A. and Veiga, J.F. (2002), “Moderators of the relationship between work-family conflict and career satisfaction”, Academy of Management Journal, Vol. 45, pp. 399-409. Netemeyer, R.G., Boles, J.S. and McMurrian, R. (1996), “Development and validation of work-family conflict and family-work conflict scales”, Journal of Applied Psychology, Vol. 81, pp. 400-10. Parasuraman, S. and Greenhaus, J.H. (2002), “Toward reducing some critical gaps in work-family research”, Human Resource Management Review, Vol. 12, pp. 299-312. Parasuraman, S., Purohit, Y.S. and Godshalk, V.M. (1996), “Work and family variables, entrepreneurial career success, and psychological well-being”, Journal of Vocational Behavior, Vol. 48, pp. 275-300. Paulin, G.D. and Lee, Y.G. (2002), “Expenditures of single parents: how does gender figure in?”, Monthly Labor Review, July, pp. 16-37. Paullay, I.M., Alliger, G.M. and Stone-Romero, E.F. (1994), “Construct validation of two instruments designed to measure job involvement and work centrality”, Journal of Applied Psychology, Vol. 79, pp. 224-8. Rizzo, J.R., House, R.J. and Lirtzman, S.I. (1970), “Role conflict and ambiguity in complex organizations”, Administrative Science Quarterly, Vol. 15, pp. 150-63. Rothausen, T.J. (1999), “Family in organizational research: a review and comparison of definitions and measures”, Journal of Organizational Behavior, Vol. 20, pp. 817-36. Sapienza, H.J. and Korsgaard, M.A. (1996), “Procedural justice in entrepreneur-investor relations”, Academy of Management Journal, Vol. 39, pp. 544-74. Sims, H.P., Szilagyi, A.D. and Keller, R.T. (1976), “The measurement of job characteristics”, Academy of Management Journal, Vol. 9, pp. 195-212. Thomas, L.T. and Ganster, D.C. (1995), “Impact of family-supportive work variables on work-family conflict and strain: a control perspective”, Journal of Applied Psychology, Vol. 80, pp. 6-15. US Bureau of Labor Statistics (1999), “Employment characteristics of family summary”, USDL 99-146, May 25, US Government Printing Office, Washington, DC. US Bureau of Labor Statistics (2000), “Labor force statistics from the current population survey 2000”, available at www.bls.gov/ (accessed July 2004). Voydanoff, P. (1988), “Work role characteristics, family structure demands, and work/family conflict”, Journal of Marriage and the Family, Vol. 50, pp. 749-61. Weiner, B. (1985), “An attributional theory of achievement motivation and emotion”, Psychological Review, Vol. 92, pp. 548-73. Whitely, W. and England, G.W. (1977), “Managerial values as a reflection of culture and the process of industrialization”, Academy of Management Journal, Vol. 20, pp. 439-53. Yang, N., Chen, C.C., Choi, J. and Zou, Y. (2000), “Sources of work-family conflict: a Sino-US comparison of the effects of work and family demands”, Academy of Management Journal, Vol. 43, pp. 113-23.
About the authors Scott L. Boyar is an Assistant Professor of Management at Xavier University’s Williams College of Business. He received his PhD in Business Administration from Mississippi State University. Dr Boyar has conducted research in the areas of recruiting, selection, turnover, absenteeism, work-family conflict, and scale development. Dr Boyar has published in such journals as Educational and Psychological Measurement, Journal of Management, and Journal of Business Research. Scott L. Boyar is the corresponding author and can be contacted at:
[email protected] Carl P. Maertz Jr received his PhD in human resources management from Purdue University’s Krannert Graduate School of Management. He is currently an Associate Professor at the John Cook School of Business at Saint Louis University. His current research interests include voluntary turnover, expatriate management, and work-family conflict. Dr Maertz has published in such journals as Journal of Applied Psychology, Academy of Management Journal, Journal of Management, and Journal of Business Research. Donald C. Mosley Jr is an Assistant Professor of Management at the University of South Alabama’s Mitchell College of Business. He earned his PhD in Business Administration from Mississippi State University. His research interests include self-efficacy, staffing, work-family conflict, and emotional intelligence. Dr Mosley has published in such journals as Educational and Psychological Measurement, Journal of Applied Social Psychology, and Journal of Business Research. Jon C. Carr is Associate Professor of Management and a BAC Professor in the College of Business at the University of Southern Mississippi. He received his PhD in Business Administration from Mississippi State University. His research interests include organizational behavior in entrepreneurial settings, work-family conflict, and turnover behavior. Dr Carr has published in such journals as Journal of Management, Educational and Psychological Measurement, and Journal of Business Research.
To purchase reprints of this article please e-mail:
[email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints
Impact of work/ family demand
235
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0268-3946.htm
JMP 23,3
236
Effort-reward imbalance, over-commitment and work-life conflict: testing an expanded model Gail Kinman
Received April 2007 Revised August 2007 Accepted November 2007
University of Bedfordshire, Luton, UK, and
Fiona Jones University of Leeds, Leeds, UK Abstract Purpose – Although the effort-reward imbalance (ERI) model of job stress has gained support in predicting employee health, it has rarely been examined in the context of the work-home interface. This study aims to test an expanded ERI model in predicting work-life conflict (WLC) in university employees. Three hypotheses relating to the ERI are tested. It is also predicted that lower organisational support for work-life balance, less schedule flexibility and lower levels of separation between work and home life will lead to increased work-life conflict. Design/methodology/approach – In this cross-sectional study, 1,108 employees working in UK universities completed questionnaires assessing ERI, WLC, schedule flexibility, employer support and work-life separation/integration. Findings – Strong main effects of job-related efforts, rewards and over-commitment on WLC are found. A significant two-way interaction (effort £ reward) and some evidence for a three-way interaction effort £ reward £ over-commitment) are observed. Perceived schedule flexibility and work-life integration also make significant contributions to the variance in WLC. The final model explains 66 per cent of criterion variance. Research limitations/implications – As the study is cross-sectional, causal relationships cannot be established. Practical implications – This study extends knowledge of the ERI model as a predictor of WLC. More research is required into ways in which effort-reward inequity and over-commitment might threaten work-life balance, together with the working practices and organisational factors which might modify this threat. Originality/value – The ERI model has rarely been examined in the context of the work-home interface. The importance of effort-reward imbalance and over-commitment to WLC has been highlighted. Keywords Sociology of work, Role conflict, Stress, Personal health, Employees Paper type Research paper
Journal of Managerial Psychology Vol. 23 No. 3, 2008 pp. 236-251 q Emerald Group Publishing Limited 0268-3946 DOI 10.1108/02683940810861365
Introduction Several models of occupational stress have been formulated in an attempt to explain relationships between features of the working environment and employee wellbeing. One such model is the effort-reward imbalance model (ERI; Siegrist, 1996). The ERI model postulates that strain results from a perceived imbalance between the level of effort employees perceive that they put into their work and the rewards that they receive. The experience of effort-reward imbalance is considered to be more frequent in employees who are excessively committed (or over-committed) to their work.
The ERI model is gaining increasing attention by researchers in the field of occupational health psychology. It has advantages over many other models of work stress, such as the job demand-control (JDC) model, as it incorporates an individual difference component and acknowledges the importance of a wider range of employment conditions (such as pay, career opportunities and job security) to employee wellbeing. Owing to a general erosion of job security and status and enhanced opportunities for schedule flexibility and self-regulation, it has also been argued that the ERI model is a particularly appropriate model through which to investigate stress in contemporary organisational settings (De Jonge et al., 2000). The ERI model has had considerable success in predicting the health status of employees (Van Vegchel et al., 2005). It has been suggested that the ERI framework should be utilised to examine other outcomes of relevance to contemporary working life (Theorell, 2006). The potential for conflict between the work and home domains has increased amongst employees in most sectors of the economy (Lewis and Cooper, 2005); its negative impact on employee wellbeing has also been highlighted (Kinman and Jones, 2001). Managing the conflict between work demands and family responsibilities has been recognised as a critical challenge for employees and organisations (Kossek and Ozeki, 1999). For several reasons discussed later in the present paper, the ERI model has clear relevance to the work-home interface. Accordingly, the present study aims to enhance knowledge of the ERI model by examining its performance as a predictor of perceived conflict between work and home. As perceived support for work-life balance and individual working practices, such as integration/segmentation, have been associated with work-life conflict (Kinman and Jones, 2001), this study further aims to examine whether these factors account for additional variance in work-life conflict over and above that explained by the three ERI components (i.e. efforts, rewards and over-commitment). According to Siegrist (2001), between 10 and 40 per cent of the workforce experience some degree of effort-reward imbalance. Owing to the model’s emphasis on wider economic forces, however, some occupational groups might be more likely than others to perceive inequity between efforts expended at work and rewards received. The present study utilises a sample of employees working in UK universities. For a number of reasons explained later in this paper, this sector is currently experiencing working conditions whereby the ERI model might be a particularly salient predictor of strain. The effort-reward imbalance model The ERI model postulates that it is not merely effort (i.e. workload or other job demands) that leads to strain, but a perceived imbalance between the effort that employees believe they put into their jobs and the rewards that they receive (Siegrist, 1996). The model explicitly differentiates between extrinsic effort (i.e. situational factors that make work more demanding, such as heavy responsibilities and frequent interruptions) and intrinsic effort (an individual difference variable also termed over-commitment). Rewards are distributed to employees by three “transmitter” elements: (1) money (appropriate salary); (2) esteem (sufficient respect and support); and (3) security/career opportunities (adequate promotion prospects, job security and status consistency) (Siegrist, 1996).
Effort-reward imbalance
237
JMP 23,3
238
Whereas perceptions of fair and appropriate rewards are expected to promote employee wellbeing, the model predicts that perceived inequity in terms of costs (high efforts expended at work) and gains (low rewards received) are experienced as stressful and will compromise health and wellbeing over the long term (Siegrist, 2005). The ERI model further predicts that effort-reward imbalance at work will be experienced more frequently by employees who are excessively preoccupied with, and overly committed to, their work. Over-commitment is defined as “a set of attitudes, behaviours and emotions that reflect excessive striving in combination with a strong desire to be approved of and esteemed” (Siegrist, 2001, p. 55). According to Siegrist, an employee who is highly over-committed will respond inflexibly to situations of high effort and low reward at work and will, therefore, be more prone to strain than a person in the same situation who is less committed. Effort-reward imbalance and strain Although significant main effects of high effort and low reward on strain are anticipated, three hypotheses relating to the ERI model have been formulated: (1) the extrinsic ERI hypothesis, which maintains that high efforts combined with low rewards results in strain over and above that accounted for by efforts and rewards independently; (2) the intrinsic over-commitment hypothesis, which states the higher the level of over-commitment the greater the strain; and (3) the intrinsic interaction hypothesis, whereby the negative impact of effort-reward imbalance on employee wellbeing is stronger in employees who are over-committed (Siegrist, 1996; Van Vegchel et al., 2005). Cross-sectional and longitudinal research provides evidence for the extrinsic ERI hypothesis, whereby a co-manifestation of high effort at work and low reward has been found to predict cardiovascular risk factors and psychiatric disorders (e.g. Stansfeld et al., 1999; Siegrist, 2001). Other studies have related effort-reward imbalance to less serious outcomes such as psychosomatic symptomatology, sleep disturbances, fatigue, problem alcohol consumption, absenteeism and turnover (Bobak, 2005; Hasselhorn et al., 2004; Fahlen, 2006; Hanebuth et al., 2006; Van Vegchel et al., 2001; Siegrist, 2005). Research evidence for the role played by over-commitment in predicting strain is, however, mixed and inconclusive. Some evidence has been provided for the intrinsic over-commitment hypothesis, whereby employees who are more over-committed tend to report poorer physical and psychological health (Fahlen, 2006; Niedhammer, 2006; Van Vegchel et al., 2005), but other studies fail to find support this prediction (Ertel et al., 2005; Kuper et al., 2002; Hanebuth et al., 2006). As yet, the intrinsic interaction hypothesis has been little examined and the available studies have yielded contradictory findings (De Jonge et al., 2000). The present study aims to test all three of the ERI hypotheses in predicting work-life conflict, prior to testing the expanded model described below. Work-life conflict and effort-reward imbalance Work-life conflict is a form of inter-role conflict whereby the fulfilment of role demands emanating from one domain (i.e. work) interferes with fulfilling role demands in another domain (i.e. home or leisure activities) (Greenhaus and Beutell, 1985). Conflict
between work and other life domains may take several forms, but that derived from time devoted to the work role (known as time-based conflict) and that derived from the strains produced by this role (known as strain-based conflict) are thought to be of key importance (Netemeyer et al., 1996). The relevance of the ERI model to the work-home interface is clear. It is plausible that high levels of job-related effort and over-commitment to the job role might result in perceived conflict between work and home. It is also likely that employees who believe that their efforts and achievements at work are not counterbalanced by the rewards they receive may be less likely to tolerate intrusion into their home lives than those who work under more equitable conditions. Moreover, as previous studies have found that effort-reward imbalance can lead to negative affective reactions (Van Vegchel et al., 2005), perceived inequity could also manifest itself as strain-based work-life conflict. As yet, only one published study can be located that has examined the ERI construct in the context of the work-home interface. Franche et al. (2006) assessed the impact of specific working conditions (including the ratio of effort to reward) and work-life conflict on the mental health of female employees. Evidence was found for a mediating effect of work-to-family conflict on the relationship between high effort-reward imbalance and negative health status. The findings are promising, but more research is required that explicitly aims to test the performance of the full ERI model in the context of the work-home interface. Whilst the components of the ERI model are plausible predictors of conflict between work and other life domains, other variables are also likely to be important. The present study investigates whether specific working practices (i.e. work-life segmentation/integration), schedule flexibility and organisational support for work-life balance predict variance in work-life conflict over and above that accounted for by the ERI. Border theory posits that the boundaries between roles can be conceptualised and measured in terms of their flexibility and permeability. Flexibility refers to the extent to which a role can be “enacted in various settings and at various times”, whereas permeability represents the degree to which an individual can be “physically located in one role’s domain but psychologically and/or behaviourally involved in another role” (Ashforth et al., 2000, p. 474). When the boundary between two role domains is flexible and permeable these domains are considered to be integrated, whereas if the boundary is inflexible and impermeable they are segmented. Ashforth et al. (2000) maintain that work and family roles can be placed on a continuum ranging from high segmentation to high integration. Although subject to individual differences in context and individual preference, employees whose work and home roles are highly integrated frequently report more work-life conflict than those whose roles are more segmented (Greenhaus and Parasuraman, 1999; Olson-Buchanan and Boswell, 2006). To a large extent, employees are responsible for maintaining an acceptable degree of separation between the work and home domains. Nonetheless, organisations clearly have some responsibility for helping their employees achieve this aim. In general, the perception of a supportive organisational culture that aspires to promote work-life balance has been associated with lower levels of work-life conflict (Allen, 2001; Thompson et al., 1998). More specifically, however, it has been argued that control over where, when and how an employee works may be one of the most critical predictors of work-home conflict (Kossek and Lambert, 2004.) It is recognised that rigid work
Effort-reward imbalance
239
JMP 23,3
240
scheduling with low levels of timing control may undermine family functioning (Haas, 1999). Conversely, schedule flexibility (or the ability to adapt working hours to meet personal and/or family needs) may increase the ability of employees to cope with the competing demands of the work and non-work domains (Sullivan and Lewis, 2006). There is some evidence that employees who work flexibly report better work-life balance and less strain (Brough et al., 2005; Fox and Fallon, 2003). As yet, however, most studies tend to compare objectively defined flexible workers with those who work conventional hours rather than examine the impact of perceived schedule flexibility (Kossek et al., 2005). The adverse effects of perceived inequity between job-related efforts and rewards, together with a tendency towards over-commitment, might be compounded in employees who: . perceive less support from their employers to help them balance the demands of their work and home lives; . have less separation between work and home; and . perceive lower levels of flexibility to work where and when they choose. These issues will be examined in the present study. Work-life conflict, effort-reward imbalance, and university employees This study utilises a sample of university employees. For several reasons, this occupational group constitutes a particularly appropriate sample for the present study. Academic work is essentially “unbounded” and incorporates a wide range of roles – each with potentially competing demands (Fisher, 1994). These factors are likely to result in increased perceptions of work-life conflict in the sector (Wortman et al., 1991). Indeed, several studies conducted in the UK and other countries provide evidence that university employees encounter particular difficulties in maintaining an acceptable work-life balance (Doyle and Hind, 1998; Winefield et al., 2003). A review of the literature on the working conditions of university employees suggests that the ERI model is a particularly salient framework through which to examine job stress in this context. Research findings suggest that employees of UK universities generally perceive their jobs as having become increasingly more demanding as a result of expanded student numbers, restructuring and mergers, increased commercialisation, enhanced external scrutiny and reductions in funding (Kinman and Jones, 2003, Kinman et al., 2006). Heavy workload and time and resource constraints are frequently highlighted as the most stressful aspects of academic and academic-related work, but other more specific demands include: . long working hours; . too much administrative paperwork; . lack of support; . obtaining research funding and finding time for research; . frequent interruptions; . rapid change; . poor leadership and management; and
.
poor salary and lack of promotion prospects (e.g. Blix et al., 1994; Thorsen, 1996; Hogan et al., 2002; Fisher, 1994; Abouserie, 1996; Doyle and Hind, 1998; Kinman, 2001; Kinman and Jones, 2003; Tytherleigh et al., 2005).
Many of these stressors are key components of the extrinsic effort dimension of the ERI model (Siegrist, 2001). Previous studies conducted in the university sector in the UK have also highlighted low levels of many of the reward components of the ERI model (i.e. salary, esteem and job security/career opportunities) as problematic (Association of University Teachers, 2001; Tytherleigh et al., 2005). In particular, the central importance of professional recognition and respect to the wellbeing of academic employees has been emphasised (Cross and Carroll, 1990; Kinman et al., 2006; Gillespie et al., 2001; Winter and Sarros, 2002). Aims of this study This study aims to extend knowledge of the ERI model to the work-home interface. More specifically, the three ERI hypothesis outlined above will be tested as a predictor of work-life conflict in a sample of employees working in UK universities. It is also predicted that lower organisational support for work-life balance, less schedule flexibility and lower levels of separation between work and home life will result in increased work-life conflict. Method Participants Questionnaires were sent to a random sample of 5,000 academic and academic-related staff employed within universities in the UK. This sample was drawn at random from the membership database of the association that represents the largest proportion of university academic staff in the UK (The Association of University Teachers, or AUT). In total, 1,108 completed questionnaires were returned, representing a response rate of 22 per cent. Fifty-five per cent of respondents were male. The majority of respondents were in the older age groups 45-49 years (18 per cent), 50-54 years (17 per cent) and 55-59 (17 per cent). Analysis of staffing figures from the Higher Education Statistics Agency (2004) confirmed that the gender balance and age profile of respondents in the present survey corresponded broadly with that of the wider population of staff in the UK at that time[1]. The majority of the sample (90 per cent) was employed on a full-time basis and held permanent contracts (82 per cent). Measures Background information was obtained, including age and gender. Multi-item measures were used to assess the components of the ERI model and work-life conflict. Effort-reward imbalance. Scales from the effort-reward imbalance (ERI) questionnaire developed by Siegrist (1996) were used to measure extrinsic effort, rewards and over-commitment. A five-item measure of extrinsic efforts and a six-item measure of over-commitment were used. Examples of items measuring extrinsic effort and over-commitment are “I have constant time pressure due to a heavy workload” and “People close to me say I sacrifice too much for my job”, respectively. A ten-item scale was used to measure esteem, financial, and status rewards. An example of an item that assesses reward is “My job promotion prospects are poor”. These scales have been
Effort-reward imbalance
241
JMP 23,3
242
shown to have good psychometric properties (Siegrist et al., 2004). All items were scored on a five-point scale. Mean scores were taken across items with a higher score signifying higher efforts, rewards and over-commitment (Cronbach’s a: efforts ¼ 0:83; over-commitment ¼ 0:92; rewards ¼ 0:88). Work-life separation/integration. A single item assessed the extent to which respondents perceived their home and work lives to be separated/integrated. Respondents were asked to indicate on a nine-point scale (where 1 denoted total separation and 9 represented total integration) the extent to which their work and home lives were separated/integrated. Schedule flexibility. A three-item scale examined the extent of perceived schedule flexibility. An example of an item is “How well does your working schedule and the degree of flexibility in this schedule meet your personal needs?”. Responses were requested on a four-point scale where 1 represented “not at all” and 4 denoted “very much so”. Mean scores were computed across items with high scores representing more schedule flexibility (Cronbach’s a ¼ 0:87). Organisational support for work-life balance. A single item measured the level of support provided by employers to aid work-life balance. Responses were requested on a four-point scale where 1 represented no support and 4 represented high support. Work-life conflict. A seven-item scale adapted from a measure developed by Netemeyer et al. (1996) was utilised. Although this scale encompasses aspects of time-based and strain-based conflict, it was designed as a unidimensional measure. Responses were requested on a seven-point scale (1 ¼ “strongly disagree” to 7 ¼ “strongly agree”). Mean scores across items were computed with higher scores denoting higher levels of work-life conflict (Cronbach’s a ¼ 0:92). Analysis The majority of studies that have tested the ERI model have utilised categories or ratios of various types in order to calculate high/low effort, high/low reward and/or develop an index of effort-reward imbalance. A review of the literature on the ERI model by Van Vegchel et al. (2005) highlights the risks in utilising arbitrary cut-off points and dichotomising continuous variables. These authors recommend that future research should utilise continuous variables and employ hierarchical linear regression techniques in order to test the interactions implied in the ERI model. They further argue that utilising an interaction term (rather than the commonly used ratio term) to assess effort-reward imbalance is more likely to yield significant effects. This study follows these recommendations. Results Table I shows the correlations between the components of the ERI model (i.e. efforts, rewards and over-commitment), work-life separation/integration, schedule flexibility, employers support for work-life balance and the outcome variable work-life conflict. Job-related efforts were negatively associated with rewards and positively associated with over-commitment (both p , 0:001). Significant relationships were observed between the three components of the ERI model and work-life conflict (all p , 0:001): i.e. respondents who reported higher efforts, lower rewards and greater over-commitment reported higher levels of work-life conflict. Significant negative relationships were also observed between work-life conflict and schedule flexibility
Variable 1. 2. 3. 4. 5. 6. 7.
Efforts Rewards Over-commitment Work-home separation/integration Schedule flexibility Employer support Work-life conflict
Means
SD
1
2
3
4
5
6
7
2.67 2.16 2.73 5.53 2.71 1.53 4.43
0.89 0.94 0.59 2.27 0.77 1.03 1.45
0.00 20.45 * * * 0.60 * * * 0.30 * * * 20.26 * * * 20.11 * * * 0.63 * * *
0.00 20.37 * * * 20.18 * * * 0.23 * * * 0.17 * * * 20.43 * * *
0.00 0.46 * * * 2 0.29 * * * 2 0.16 * * * 0.73 * * *
0.00 2 0.18 * * * 2 0.11 * * * .51 * * *
0.00 0.23 * * * 2 0.35 * * *
0 20.20 * * *
0.00
Note: One-tailed correlations: *p , 0:05; * *p , 0:01; * * *p , 0:001
Effort-reward imbalance
243
Table I. Descriptive statistics and correlations between ERI components and other study variables
JMP 23,3
244
and employer support (both p , 0:001). In terms of working practices, respondents whose work and home lives were more integrated tended to report significantly higher levels of work-life conflict (p , 0:001). In order to examine the predictors of work-life conflict, a hierarchical multiple regression equation was computed whereby the mean-centred independent variables were entered in seven steps. At the first step, gender was entered to control for any effects. Job-related efforts and rewards were entered in step two to examine their main effects. The third step tested the extrinsic ERI hypothesis by entering a two-way interaction between efforts and rewards. The fourth step tested the intrinsic over-commitment hypothesis, by entering over-commitment in order to examine its direct effects. In the fifth step, the three-way interaction term efforts £ rewards £ over-commitment was entered in order to test the intrinsic interaction hypothesis. Employer support for work-life balance, perceived schedule flexibility and work-life separation/integration were entered in the sixth and final step. Table II presents the results of these analyses. Gender, entered in Step 1 of the equation, failed to account for any significant variance in work-life conflict. The main effects of efforts and rewards entered in Step 2 accounted for 42 per cent of the variance in work-life conflict. An examination of the betas indicated that both job-related efforts and rewards were significant (p , 0:001) but efforts made the strongest contribution. A significant two-way interaction between efforts and rewards (Step 3) was observed that explained 2 per cent of variance in work-life conflict, thus supporting the extrinsic ERI hypothesis. The intrinsic over-commitment hypothesis was supported by the main effects of over-commitment making a unique contribution to the variance of 17 per cent (Step 4). Some evidence was also found to support the intrinsic interaction hypothesis of the ERI model with a significant three-way interaction between efforts, rewards and over-commitment but at 0.4 per cent its contribution to the variance was minimal (Step 5). Employer support for work-life balance, schedule flexibility and work-life integration (entered in the sixth and final step) together accounted for just over 5 per cent of variance. An examination of the betas indicated that schedule flexibility and work-life separation/integration were significant predictors (both p , 0:001) and that the latter variable was made the strongest contribution. The final model explained a total of 66 percent of the variance in work-life conflict. Discussion This study aimed to extend knowledge of the ERI construct as a predictor of work-life conflict – an issue of considerable relevance to contemporary organisations. Three hypotheses relating to the model were tested suggested by Siegrist (1996) prior to examining an extended model that encompassed organisational support for work-life balance, schedule flexibility and work-life separation/integration. Findings indicate that the components of the ERI model are powerful predictors of work-life conflict. University employees whose reward expectancies are not fully met tended to report a poorer work-life balance than those who worked under conditions of greater equity, on average, reported more conflict between their work and home lives. As hypothesised by the model, some evidence was found that the experience of effort-reward imbalance at work is potentially more damaging to work-life balance in employees who exhibit a particular pattern of coping with work demands characterised by excessive
Predictors Gender Step 1 R 2 Gender Efforts Rewards Step 2 R 2 change Gender Efforts Rewards Efforts £ rewards Step 3 R 2 change Gender Efforts Rewards Efforts £ rewards Over-commitment Step 4 R 2 change Gender Efforts Rewards Efforts £ rewards Over-commitment Efforts £ rewards £ over-commitment Step 5 R 2 change
R 2 change
Betas
Effort-reward imbalance
0.00 0.000
0.417 * * *
2 0.00 0.54 * * * 2 0.18 * * *
245
2 0.01 0.88 * * * 2 0.63 * * * 0.69 * * * 0.022 * * * 2 0.01 0.46 * * * 2 0.40 * * * 0.42 * * * 0.53 * * * 0.017 * * * 2 0.01 0.34 * * * 2 0.26 * * 0.19 0.64 * * * 0.46 * * * 0.004 * *
Gender Efforts Rewards Efforts £ rewards Over-commitment Efforts £ rewards £ over-commitment Employer support Schedule flexibility Work-home separation/integration Step 6 R 2 change
2 0.00 0.34 * * * 2 0.25 * * 0.07 0.49 * * * 0.34 * * 2 0.03 2 0.09 * * * 0.22 * * * 0.048 * * *
Total R 2
0.66
Notes: n ¼ 1; 108; p , 0:05; * *p , 0:01; * * *p , 0:001
commitment to work. The study also found evidence that lack of perceived schedule flexibility and higher levels of work-life integration are independent risk factors that further compromise work-life balance. Strong main effects of job-related efforts and rewards on work-life conflict were observed. The significant two-way interaction between efforts and rewards also supports the extrinsic ERI hypothesis in the context of the work-home interface. Two hypotheses relating to the intrinsic effort component of the model in predicting work-life conflict were also supported in this study. Previous research that has
Table II. Hierarchical multiple regression model examining the incremental variance of the ERI components and other study variables on work-life conflict
JMP 23,3
246
examined the role played by intrinsic effort, or over-commitment, in employee wellbeing has yielded inconsistent findings (Van Vegchel et al., 2005). Although it has been suggested that over-commitment might not be a risk factor for all occupational groups (Van Vegchel et al., 2001), the findings of this study suggest that it is a robust predictor of work-life conflict in university employees that might also exacerbate the negative effects of effort-reward imbalance. Evidence to support the intrinsic over-commitment hypothesis and the intrinsic interaction hypotheses as predictors of work-life conflict was therefore found. Although the variance accounted for by the three-way interaction between efforts, rewards and over-commitment was minimal, it has been suggested that even small contributions to the observed variance may be an indication that there are substantial effects for those at the extremes of a population (Frese, 1985, p. 314) The cognitive, emotional and motivational pattern that constitutes over-commitment (i.e. excessive striving combined with a strong motivation to seek esteem and approval) signifies high engagement with the work role. The potential for over-commitment to threaten work-life balance is therefore clear. In the present study, over-commitment emerged as one of the strongest predictors of work-life conflict. Furthermore, university employees who were more over-committed to their jobs perceived less separation between their work and home lives, lower levels of schedule flexibility and less support from their employers to facilitate work-life balance. Future research should investigate the mechanisms by which over-commitment manifests itself in the home domain. Employees who are more over-committed to their work may have less time and energy available to engage fully in home life and leisure activities; they may also be more likely to import strain engendered by work into the home environment. It is also possible that employees who more over-committed perceive their job roles to be more salient than their family roles. Nonetheless, the fact that they report significantly higher levels of work-life conflict suggests that an over-committed employee might perceive this tendency as threatening his or her wellbeing and family functioning. This study found that work-life integration and lack of schedule flexibility were independent risk factors for work-life conflict beyond that accounted for by the components of the ERI model. In accordance with previous research findings (e.g. Desrochers et al., 2005), respondents with firmer boundaries between their work and home lives were more likely to have achieved a better work-life balance. Organisations might help their employees to maintain an acceptable balance between work and home by offering guidance on strategies to help them maintain greater separation between the two domains. Schedule flexibility (or the level of control over where, when, and how an employee works) was also found to be a key predictor of work-life conflict. This implies that the ability to adapt working hours to meet personal and/or family needs may help employees cope with the competing demands of the work and non-work domains. Based on previous research (Allen, 2001), it was anticipated that an organisational culture that supported work-life balance might protect employees from work-life conflict. Nonetheless, although participants who perceived greater support from their institutions tended to report significantly lower levels of conflict, support did not emerge as a significant predictor in the regression analysis. The mean level of satisfaction with organisational support for work-life balance was, however, extremely low (i.e. 1.5 on a four-point scale), suggesting that support in the sector should be enhanced.
Interventions to minimise work-life conflict based on the ERI model would involve restoring the balance between efforts expended and rewards received – thus improving employees’ sense of fairness and reciprocity. In the university sector, such balance could be achieved by reducing extrinsic efforts and/or enhancing rewards such as esteem, promotion prospects and job security. It is acknowledged that enhancing rewards without simultaneously reducing efforts might not necessarily reduce work-life conflict and might further threaten quality of life outside the working environment – especially for employees who are more over-committed to the job role. The findings of this study also suggest that attempts to modify over-commitment might improve the work-life balance of employees beyond those afforded by structural improvements. However, finding ways to accomplish this is likely to be a challenge for occupational health researchers. Siegrist’s (2001, 2005) view that over-commitment is an intrinsic characteristic of the individual implies that it might not be easily modifiable. Nonetheless, the stability of over-commitment has not yet been established; it is possible that this tendency might, to some extent, be encouraged by exposure to specific working conditions or occupational cultures. Longitudinal research is needed to investigate the extent to which over-commitment is a state or a trait, and to establish ways by which it might be modified. Whether such interventions would be successful under current working conditions in the UK university sector is open to question. A high degree of commitment to the job role may be required if academics are to meet personal and professional standards of performance in the face of high levels of demand emanating from the different aspects of the job role. There are some limitations inherent in this study. The cross-sectional nature of the data precludes making firm conclusions about causality. Although considerably less plausible, it is nonetheless possible that negative assessments of work-life balance are the cause (rather than the result) of a perception of imbalance between efforts and rewards and a tendency towards over-commitment. Future studies should adopt a longitudinal design to more confidently establish the direction of causation between the ERI components and work-life conflict. The risk of common method variance in inflating relationships between some of the independent and dependent variables utilised in this study is also acknowledged. Despite these limitations, this study has undoubted strengths. The negative implications of effort-reward imbalance and over-commitment for work-life balance have been confirmed, as has the value of extending the model to encompass different working practices and perceptions of support. More research is necessary to further examine the implications of the ERI model for the work-home interface. There is evidence that perceptions of working conditions and work-life conflict are subject to daily variation (Jones and Fletcher, 1996; Butler et al., 2005). Future research might utilise daily diaries to examine ways in which inequity between job-related efforts and rewards and a tendency towards over-commitment might lead to work-life conflict (differentiating between time-based and strain-based conflict) and the working practices and organisational factors which might modify or mitigate this threat. Note 1. The mean age of the sample utilised in the present study was slightly higher than that of the UK population working in the higher education sector at the time the study was conducted.
Effort-reward imbalance
247
JMP 23,3
248
References Abouserie, R. (1996), “Stress, coping strategies and job satisfaction in university academic staff”, Educational Psychology, Vol. 16 No. 1, pp. 49-56. Allen, T.D. (2001), “Family-supportive work environments: the role of organizational perceptions”, Journal of Vocational Behaviour, Vol. 58, pp. 414-35. Ashforth, B.E., Kreiner, G.E. and Fugate, M. (2000), “All in a day’s work: boundaries and micro role transitions”, Academy of Management Review, Vol. 25, pp. 472-91. Association of University Teachers (2001), Implementation of the Fixed-term Work Directive: Department of Trade and Industry Consultation – A Preliminary AUT Response, AUT Publications, London. Blix, A.G., Cruise, R.J., Mitchell, B.M. and Blix, G.G. (1994), “Occupational stress among university teachers”, Educational Research, Vol. 36 No. 2, pp. 157-69. Bobak, M. (2005), “The association between psychosocial characteristics at work and problem drinking: a cross-sectional study of men in three Eastern European urban populations”, Occupational and Environmental Medicine, Vol. 62 No. 8, pp. 546-50. Butler, A.B., Grzywacz, J.G., Bass, B.L. and Linney, K.D. (2005), “Extending the demands-control model: a daily diary study of job characteristics, work-family conflict, and work-family facilitation”, Journal of Occupational and Organizational Psychology, Vol. 78, pp. 155-69. Cross, G. and Carroll, D. (1990), Goodwill Under Stress: Morale in UK Universities, Association of University Teachers, London. De Jonge, J., Bosma, H., Peter, R. and Siegrist, J. (2000), “Job strain, effort-reward imbalance and employee well-being: a large-scale cross sectional study”, Social Science and Medicine, Vol. 50, pp. 1317-27. Desrochers, S., Hilton, J. and Larwood, L. (2005), “Preliminary validation of the work-family integration blurring scale”, Journal of Family Issues, Vol. 26, pp. 442-66. Doyle, C. and Hind, P. (1998), “Occupational stress, burnout and job status in female academics”, Gender, Work and Organisation, Vol. 5 No. 2, pp. 67-82. Ertel, M., Pech, E., Ullsperger, P., Von Dem Knesebeck, O. and Siegrist, J. (2005), “Adverse psychosocial working conditions and subjective health in freelance media workers”, Work and Stress, Vol. 19 No. 3, pp. 293-9. Fahlen, G. (2006), “Effort-reward imbalance, sleep disturbances and fatigue”, International Archives of Occupational and Environmental Health, Vol. 79 No. 5, pp. 371-8. Fisher, S. (1994), Stress in Academic Life: The Mental Assembly Line, Open University Press, Buckingham. Franche, R., Williams, A., Ibrahim, S., Grace, S., Mustard, C., Minore, B. and Stewart, D. (2006), “Path analysis of work conditions and work-family spillover as modifiable workplace factors associated with depressive symptomatology”, Stress and Health, Vol. 22 No. 2, pp. 91-103. Frese, M. (1985), “Stress at work and psychosomatic complaints: a causal interpretation”, Journal of Applied Psychology, Vol. 70 No. 2, pp. 314-28. Gillespie, N.A., Walsh, M., Winefield, A.H. and Stough, C. (2001), “Occupational stress in universities: staff perceptions of the causes, consequences and moderators of stress”, Work and Stress, Vol. 15 No. 1, pp. 53-72. Greenhaus, J.H. and Beutell, N.J. (1985), “Sources of conflict between work and family roles”, Academy of Management Review, Vol. 10, pp. 76-80.
Greenhaus, J.H. and Parasuraman, S. (1999), “Research on work, family and gender: current status and future directions”, in Powell, G.N. (Ed.), Handbook of Gender and Work, Sage Publications, London.
Effort-reward imbalance
Haas, L. (1999), “Families and work”, in Sussman, M., Steinmetz, S.K. and Peterson, G.W. (Eds), Handbook of Marriage and the Family, Plenum Press, New York, NY. Hanebuth, D., Meinel, M. and Fischer, J. (2006), “Health-related quality of life, psychosocial work conditions and absenteeism in an industrial sample of blue- and white-collar employees: a comparison of potential predictors”, Journal of Occupational and Environmental Medicine, Vol. 48 No. 1, pp. 28-37. Hasselhorn, H.M., Tackenberg, P. and Peter, R. and the NEXT Study Group (2004), “Effort-reward imbalance among nurses in stable countries and in countries in transition”, International Journal of Occupational and Environmental Health, Vol. 10, pp. 401-8. Higher Education Statistics Agency (2004), Resources of Higher Education Institutions, Higher Education Statistics Agency, London. Hogan, J.M., Carlson, J.G. and Dua, J. (2002), “Stress and stress reactions among university personnel”, International Journal of Stress Management, Vol. 9 No. 4, pp. 289-310. Jones, F. and Fletcher, B.C. (1996), “Taking work home: a study of daily fluctuations in work stressors, effects on mood and impacts on martial partners”, Journal of Occupational and Organizational Psychology, Vol. 69, pp. 89-106. Kinman, G. (2001), “Pressure points: a review of stressors and strains in UK academics”, Educational Psychology, Vol. 21 No. 4, pp. 473-92. Kinman, G. and Jones, F. (2001), “The work-home interface”, in Jones, F. and Bright, J. (Eds), Stress: Myth, Theory and Research, Prentice-Hall, London. Kinman, G. and Jones, F. (2003), “Running up the down escalator: stressors and strains in UK academics”, Quality in Higher Education, Vol. 9 No. 1, pp. 22-37. Kinman, G., Jones, F. and Kinman, R. (2006), “The wellbeing of the UK academy: 1998 to 2004”, Quality in Higher Education, Vol. 12 No. 1, pp. 15-27. Kossek, E. and Lambert, R. (2004), Work and Life Integration: Organizational, Cultural and Psychological Perspectives, Lawrence Erlbaum Associates, Mahwah, NJ. Kossek, E. and Ozeki, C. (1999), “Bridging the work-family and productivity gap: a literature review”, Community, Work and Family, Vol. 2 No. 1, pp. 7-32. Kuper, H., Singh-Manoux, A., Siegrist, J. and Marmot, M. (2002), “When reciprocity fails: effort-reward imbalance in relation to coronary heart disease and health functioning within the Whitehall II study”, Occupational and Environmental Medicine, Vol. 59, pp. 777-84. Lewis, S. and Cooper, C. (2005), Work-Life Integration: Case Studies of Organizational Change, Wiley, Chichester. Netemeyer, R.G., Boles, J.S. and McMurrian, R. (1996), “Development and validation of work-family conflict and family-work conflict scales”, Journal of Applied Psychology, Vol. 81, pp. 400-10. Niedhammer, I. (2006), “Psychosocial work environment and mental health: job-strain and effort-reward imbalance models in a context of major organisational changes”, International Journal of Occupational and Environmental Health, Vol. 12 No. 2, pp. 111-9.
249
JMP 23,3
250
Olson-Buchanan, J. and Boswell, W. (2006), “Blurring boundaries: correlates of integration and segmentation between work and nonwork”, Journal of Vocational Behavior, Vol. 68, pp. 432-45. Siegrist, J. (1996), “Adverse health effects of high effort-low reward conditions at work”, Journal of Occupational Health Psychology, Vol. 1, pp. 27-43. Siegrist, J. (2001), “A theory of occupational stress”, in Dunham, J. (Ed.), Stress in the Workplace: Past, Present and Future, Whurr Publishers, London. Siegrist, J. (2005), “Social reciprocity and health: new scientific evidence and policy implications”, Psychoneuroendocrinology, Vol. 30 No. 10, pp. 1033-8. Siegrist, J., Starke, D., Chandola, T., Godin, I., Marmot, M., Niedhammer, I. and Peter, R. (2004), “The measurement of effort-reward imbalance at work: European comparisons”, Social Science & Medicine, Vol. 58 No. 8, pp. 1483-99. Stansfeld, S., Fuhrer, R., Shipley, J.J. and Marmot, M. (1999), “Work characteristics predict psychiatric disorders: prospective results from the Whitehall II study”, Occupational and Environmental Medicine, Vol. 56 No. 5, pp. 302-7. Sullivan, C. and Lewis, S. (2006), “Relationships between work and home life”, in Jones, F., Burke, R. and Westman, M. (Eds), Managing the Work-Home Interface: A Psychological Perspective, Taylor & Francis, London. Theorell, T. (2006), “New directions for psychosocial work environment research”, Scandinavian Journal of Public Health, Vol. 34 No. 2, pp. 113-5. Thorsen, E.J. (1996), “Stress in academe: what bothers professors?”, Higher Education, pp. 471-89. Tytherleigh, M., Webb, C., Cooper, C. and Ricketts, C. (2005), “Occupational stress in UK higher education institutions: a comparative study of all staff categories”, Higher Education Research and Development, Vol. 24 No. 1, pp. 41-61. Van Vegchel, N., De Jonge, J., Bosma, H. and Schaufeli, W. (2005), “Reviewing the effort-reward imbalance model: drawing up the balance of 45 empirical studies”, Social Science and Medicine, Vol. 60 No. 5, pp. 1117-32. Van Vegchel, N., De Jonge, J., Meijer, T. and Hamers, J. (2001), “Different effort constructs and effort-reward imbalance: effects on employee well-being”, Journal of Advanced Nursing, Vol. 34 No. 1, pp. 128-37. Winefield, A.H., Gillespie, N., Stough, C., Dua, J., Hapuarachchi, J. and Boyd, C. (2003), “Occupational stress in Australian university staff”, International Journal of Stress Management, Vol. 10, pp. 51-63. Winter, R.P. and Sarros, J.C. (2002), “The academic work environment in Australian universities: a motivating place to work?”, Higher Education Research and Development, Vol. 21 No. 3, pp. 242-58. Wortman, C., Biernat, M. and Lang, E. (1991), “Coping with role overload”, in Frankenhaeuser, M., Lundberg, U. and Chesney, M. (Eds), Women, Work and Health: Stress and Opportunities, Plenum Press, London.
Further reading Daniels, K. and Guppy, A. (1994), “An exploratory study of stress in a British university”, Higher Education Quarterly, Vol. 48 No. 2, pp. 135-44. Goldberg, D. and Williams, P. (1988), A User’s Guide to the General Health Questionnaire, NFER-Nelson, Windsor.
About the authors Gail Kinman is Reader in Occupational Health Psychology in the Department of Psychology at the University of Bedfordshire, UK. She is a Chartered Psychologist who conducts research into aspects of job stress, work-life balance and the nature and impact of emotional demands in the workplace. She works mainly with public sector organisations including universities, nurses, teachers and police. Gail Kinman is the corresponding author and can be contacted at:
[email protected] Fiona Jones is a Senior Lecturer in Health and Occupational Psychology at the Institute of Psychological Sciences, University of Leeds, UK. She is a Chartered Health Psychologist and conducts research on the topic of work stress, work life balance and on aspects of health behaviour, including the impact of work stress on health behaviours.
To purchase reprints of this article please e-mail:
[email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints
Effort-reward imbalance
251
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0268-3946.htm
JMP 23,3
Examining couple agreement about work-family conflict
252
University of Tulsa, Tulsa, Oklahoma, USA
Michelle Streich Wendy J. Casper Received March 2007 Revised September 2007 Accepted November 2007
University of Texas at Arlington, Arlington, Texas, USA, and
Amy Nicole Salvaggio University of Tulsa, Tulsa, Oklahoma, USA Abstract Purpose – This study aims to explore the nature of couple agreement about work-family conflict, adding to previous research by explicitly testing the extent to which couples agree when rating work interference with family (WIF) and the influence of this agreement on other outcomes. Design/methodology/approach – In total, 224 dual-earner couples were surveyed to assess their own WIF, as well as what they believed to be their partner’s level of WIF. Participants also completed questions regarding their organizational commitment. Findings – Couples agreed when rating their own and their partners’ WIF more than they disagreed. As predicted, couples agreed more when rating the female partner’s WIF as compared to the male partner’s WIF. Finally, couple agreement about WIF moderated the relationship between female WIF and her continuance organizational commitment such that the relationship between the female partner’s WIF and her level of continuance commitment was stronger when agreement about her experienced WIF was low. Research limitations/implications – This was a convenience sample, and therefore caution should be used when generalizing to a broader population. Second, the research design was cross-sectional, prohibiting causal inferences and conclusions about couple agreement over time. Practical implications – Organizations should consider the perceptions and attitudes of both employees and their partners, as both have implications for work attitudes. Organizations might benefit from considering ways in which they can involve and engage employees’ spouses and partners, and could offer flexible schedules as a way to reduce employee work-to-family conflict and enhance both employee and partner attitudes toward the organization. Originality/value – This paper contributes to the literature by exploring both self and partner perceptions of work-family conflict and examining couple agreement about this conflict. Keywords Sociology of work, Family, Role conflict, Attitudes Paper type Research paper
Journal of Managerial Psychology Vol. 23 No. 3, 2008 pp. 252-272 q Emerald Group Publishing Limited 0268-3946 DOI 10.1108/02683940810861374
The number of women in the workforce has grown exponentially from 1970 to 2004, when women went from comprising 43 percent of the workforce to 59 percent (US Census Bureau, 2005). These changes resulted in an increase of dual-earner couples. For example, 58 percent of married couples reported income from both husband and wife in 2003, compared to 44 percent in 1967 (Bureau of Labor Statistics, 2003). Accordingly, research has increasingly examined how workers experience and manage conflict between their work and family roles (Eby et al., 2005).
Work-family conflict (WFC) is defined as “a form of interrole conflict in which the role pressures from work and family domains are mutually incompatible in some respect” (Greenhaus and Beutell, 1985, p. 77). Research suggests WFC is related to organizational commitment (Lyness and Thompson, 1997; Netemeyer et al., 1996), job performance (Frone et al., 1997), and turnover (Greenhaus et al., 1997). WFC also relates to non-work outcomes such as life satisfaction (Kossek and Ozeki, 1998), family involvement (Frone et al., 1992a), and marital satisfaction (Coverman, 1989). Thus, WFC has implications for both organizations and employees. WFC can occur in two directions: work can interfere with family (WIF) and family can interfere with work (FIW). The present study focuses on WIF for several reasons. Research has shown that WIF, but not FIW, predicts job dissatisfaction and organizational commitment (Casper et al., 2007; Lyness and Thompson, 1997). WIF is also more prevalent than FIW (McElwain et al., 2005; Frone et al., 1992b). Finally, we focus on WIF because it is appropriate for examining couple agreement. A spouse/partner has the opportunity to experience and observe when his or her partner’s work interferes with family, but may not have the opportunity to observe when family interferes with work. Extant research has tended to examine WFC from the individual level of analysis, failing to consider the perspectives of both members of a couple (Casper et al., 2007). Multiple studies call for research examining the couple as the level of analysis (Hammer et al., 1997; Moen and Yu, 2000; Casper et al., 2007; Eby et al., 2005; Hayden et al., 1998), and a recent review of the work-family literature found that only 5 percent of studies examined the couple level of analysis (Casper et al., 2007). Given an individual’s perception of WFC has been found to relate to his or her partner’s experiences (Hammer et al., 1997, 2003; Matthews et al., 2006), to completely understand an individual’s WFC, partner perceptions are critical. The present study examines both self and partner perceptions of WIF among dual-earner couples. We explore the extent to which individuals within a couple agree about their WIF and their partner’s WIF, and whether this agreement moderates the relationship between WIF and individual-level organizational commitment. In what follows we delineate a theoretical rationale for examining couple agreement about WIF and its potential moderating influence on the relationship between WIF and individual-level organizational commitment. Systems theory Systems theory (e.g. Bronfenbrenner, 1979) argues that individuals exist in an ecological environment comprised of nested systems. The two-person system, or dyad, is considered the “innermost level of the ecological schema” (p. 5). Research on both members of a dyad provides a richer and more dynamic picture of the relationship between dyad members (Bronfenbrenner, 1979). Thus, family systems theory can be utilized to conceptualize the couple in the WFC literature (Hammer et al., 2003). Family systems theory suggests that a person’s attitudes are affected by his or her family members’ attitudes and behaviors (Hammer et al., 2003, 2005). Accordingly, each individual within a couple influences his or her partner’s decisions (Doumas et al., 2003; Gareis et al., 2003). For instance, both husbands and wives report spending more hours at work the day after they reported low marital warmth from their spouse (Doumas et al., 2003).
Examining couple agreement about WFC 253
JMP 23,3
254
Failure to consider the interdependence of couples’ experiences has implications for organizations, given research finds that employees’ partners can and do influence their attitudes and behaviors at work. For example, Eby and Russell (2000) found that spouses’ willingness to move was the strongest predictor of employees’ willingness to relocate for jobs. Similarly, Rosen and Durand (1995) found that the strongest predictor of US Army officer retention was the spouse’s attitude toward re-enlistment. Dornstein and Matalon (1989) found that family attitudes toward Army service were one of the biggest predictors of commitment to the Israeli Army. Several other studies have found that the adjustment of expatriates’ spouses plays a critical role in expatriates’ adjustment (Caligiuri et al., 1998) and withdrawal from assignments (Shaffer and Harrison, 1998). Crossover effects Family systems theory posits that the attitudes and experiences of one family member influence those of another family member. Crossover effects occur when one person’s stressors and strains result in elevated levels of stress or strain for his or her partner (Westman and Etzion, 1995). For example, Hammer et al. (1997) found that within dual-earner couples, one person’s WFC accounted for significant variance in his or her partner’s WFC. Hammer et al. (2003) found that wives’ tardiness to work was related to husbands’ level of FIW, and that husbands’ absence from work was related to wives’ FIW. Crossover effects have also been found with respect to husband-to-wife and wife-to-husband burnout (Westman and Etzion, 1995, 2001). Matthews et al. (2006) studied crossover effects among dual-earner couples. This study explored what they called direct crossover effects in which an individual’s experienced work-to-relationship conflict was expected to relate to his or her partner’s perceptions of this conflict. Thus, direct crossover effects, as defined by Matthews et al. (2006), reflect what we refer to as couple agreement. Findings from this study support the notion that couples agree reasonably well when rating work-to-relationship conflict. In addition, feeling like one’s partner’s work interfered with the relationship was positively related to relationship tension, which in turn was related to relationship satisfaction and negative health outcomes (Matthews et al., 2006). We extend Matthews et al.’s (2006) study in two ways. First, we examine the degree to which level of agreement regarding work-to-family conflict (WIF) depends on who is being rated in the dyad – the male or female partner. As we describe below, personal experiences, socialization, and gender roles may result in higher agreement for women’s WIF than for men’s. Second, we expand upon Matthews et al.’s (2006) results by studying job attitudes rather than relationship outcomes. Specifically, we investigate the possible role of couple agreement in the relationship between WIF and organizational commitment. Couple agreement Because individuals within a couple influence one another (Hammer et al., 2003; Westman and Etzion, 1995, 2001), it is important to consider the extent to which two members of a dyad agree regarding their perceptions. That is, some couples may perceive the level of WFC experienced similarly (e.g. both think the woman has high conflict), whereas other couples may disagree about the level of conflict experienced (e.g. the woman thinks the man has high conflict, but the man thinks he has low conflict). Because individuals within a couple have the information available and the
motivation to accurately perceive each other’s attitudes (Matthews et al., 2006; Kenny and Acitelli, 2001), examining couple agreement is warranted. Researchers have begun to investigate couple agreement about perceptions that pertain to a couple’s relationship. For example, Rusbult et al. (1998) examined couple agreement by using self and partner ratings of relationship accommodation behaviors. They found that, when rating the husband’s accommodation, the wife’s and husband’s ratings were fairly highly correlated (from 0.35 to 0.74). In contrast, when examining the wife’s accommodation, the husband’s and wife’s ratings were positively correlated but smaller in magnitude (from 0.14 to 0.57). That is, couples agreed more when rating the husband’s level than the wife’s level of accommodation. Matthews et al. (2006) examined individuals’ self ratings of work-to-relationship conflict as it relates to their partner’s rating of their conflict. As expected, self ratings of work-to-relationship conflict were positively related to partner ratings of their work-to-relationship conflict for both male and female targets. Following Matthews et al. (2006), we hypothesized that self ratings and partner ratings of a target’s WIF would be positively and significantly related. In other words, couples were expected to exhibit significant agreement when evaluating a target’s WIF. H1. There will be a positive and significant relationship between self ratings and partner ratings (i.e. self-partner agreement) regarding the level of WIF experienced by a member of a dual-earner couple. Gender differences in agreement We also explore whether couple agreement is higher when rating the male target’s WIF or the female target’s WIF. This is important because different levels of agreement about a partner’s experiences may influence the degree to which partner perceptions are of interest to organizations. For instance, if there tends to be high agreement between partners about female’s work-family experiences, it makes sense for an organization to consider the perspectives and attitudes of the female’s partner, given high agreement suggests the female employee’s attitudes toward her organization are influenced by her partner. In contrast, if there is low agreement about a male employee’s work-family experiences, his partner’s perceptions may be of less interest to his organization. That is, if he and his partner perceive his organization and his work-family issues quite differently, his partner’s perceptions may have little direct impact on his work attitudes. Because a key goal of this study is to explore partner perceptions and the degree to which it makes sense for organizations to care about partner perceptions, it is useful to know whether partner agreement differs based on employee gender. Research has found that women generally report more WIF than men, which some suggest is because women deem work as more of an imposition on family than men, given women feel greater family responsibility (McElwain et al., 2005). This is consistent with the fact that women spend approximately twice as much time as men performing household and family duties (Lundberg and Frankenhaeuser, 1999). Theories regarding gender stereotypes partly explain why women report higher levels of WIF, and predict that male and female partners may perceive each other’s WIF differently. Gender stereotypes are both descriptive (i.e. they contain information about what women are like) and proscriptive (i.e. they describe what women should be like) (Heilman, 2001). The gender stereotype for women includes care-giving (Eagly and Johannesen-Schmidt, 2001; Edwards and Hamilton, 2004), a communal orientation,
Examining couple agreement about WFC 255
JMP 23,3
256
and nurturance (Heilman and Okimoto, 2007). Women are simply expected to express more interest in (and, in turn, are more focused on) their close personal relationships (Cross and Madson, 1997). This expectation is internalized through the types of activities and interests parents encourage and cultivate in their children (Lytton and Romeny, 1991; Whiting and Edwards, 1988). Thus, a woman’s interference from work to family may be more salient for the woman (who has internalized the norm that she, as a woman, should be focused on family care-giving), as well as for her male partner (who is also exposed to gender stereotypes and thus has similar expectations for her). Other aspects of gender stereotypes, in addition to descriptions and proscriptions about nurturance, may predict different levels of agreement regarding male and female partner’s WIF. Women tend to be more expressive than men (Kring and Gordon, 1998) and individuals who express complaints to their partners are more likely to have partners that understand their experiences (Sillars et al., 1984). Thus, when women experience WIF they may express it more openly, and increase the likelihood that their partner will understand their conflict. Moreover, men are permitted less “latitude” in their gender roles – it is not deemed socially acceptable for them to express as many cross-sex attributes as compared to women (McCreary, 1994). In other words, heterosexual men may experience more negative consequences for displaying female attributes than women do for displaying male attributes. Accordingly, if men feel work is interfering with their family (for example, if long work hours interfere with care-giving activities), they may not express it, even to their partners, as such expressions and valuing of family in this way may be perceived by men as “feminine.” Hence, couples may experience more agreement about the woman’s WIF because women are allowed to, and do, communicate about it more. This should lead to higher agreement about the female’s level of WIF. H2. Couple agreement will be higher when rating the female target’s WIF than when rating the male target’s WIF. Work-to-family conflict and organizational commitment It is important to consider whether agreement within couples about WIF relates to other aspects of work or family functioning. Research suggests that couple agreement is related to family outcomes. For example, O’Brien and Peyton (2002) found that couple agreement about child-rearing was related to wives’ increased perception of marital intimacy. However, no research we are aware of has examined couple agreement about WIF as an antecedent of work outcomes, despite the fact that such agreement could influence individual perceptions of work and work attitudes. One potentially important attitude is organizational commitment. Meyer and Allen (1991) propose three components of organizational commitment: (1) affective; (2) continuance; and (3) normative. Affective commitment refers to employees’ genuine felt loyalty toward an organization. Continuance commitment reflects employees’ perceptions that the costs of leaving the organization are too high. Normative commitment reflects employees’ sense that they should remain with an organization, and is highly related to affective
commitment. The current study focuses on WIF as it relates to the two most distinct aspects of organizational commitment, affective and continuance commitment. WFC is related to both affective and continuance commitment (Casper et al., 2002; Good et al., 1988; Lyness and Thompson, 1997; Meyer et al., 2002). Research has consistently found a moderate negative relationship between general WFC and affective commitment (Allen et al., 2000; Good et al., 1988; Lyness and Thompson, 1997; Siegel et al., 2005). More specifically, although most studies have found a negative relationship between WIF and affective commitment (Good et al., 1988; Lyness and Thompson, 1997; Netemeyer et al., 1996), Casper et al. (2002) found that WIF was unrelated to affective commitment among employed mothers. Discrepancies in past findings suggest it may be relevant to explore moderators of these relationships. Our study explores influences from the family system as moderators of this relationship. Family systems theory argues that what one member of a family (or unit) does or thinks influences other members of the family system. When considering work-to-family conflict, it may be the case that when individuals within a couple agree about the conflict experienced, this agreement serves to reinforce each other’s attitudes. For instance, the negative relationship between a woman’s WIF and her affective commitment may be amplified if her partner agrees and commiserates with the fact that her work is interfering with family. It may be that improved communication or shared understanding among the couple enhances the association between her perceived conflict and job attitudes. Thus, if she is experiencing reduced affective commitment due to high WIF, knowing her partner agrees with her about the high level of WIF experienced may reinforce any negative attitudes that result from WIF. Accordingly, the negative relationship between WIF and affective commitment is expected to be stronger among couples with high agreement about WIF. H3. Couple agreement will moderate the relationship between a target’s WIF and that target’s affective commitment. Specifically, the negative relationship will be stronger when couple agreement is higher than when it is lower (see Figure 1). Previous research has found that general WFC is positively related to continuance commitment (Meyer et al., 2002; Thompson et al., 1998), such that those with more conflict report a greater sense of being “stuck” in their organization. WIF specifically is also positively related to continuance commitment (Casper et al., 2002; Good et al., 1988; Lyness and Thompson, 1997). Based again on the rationale drawn from family systems theory, high agreement about WIF is expected to reinforce any negative attitudes (i.e. high continuance commitment) that result from WIF. Thus, the relationship between WIF and continuance commitment is expected to be greater when agreement is high. H4. Couple agreement about WIF will moderate the relationship between a target’s WIF and that target’s continuance commitment. Specifically, the positive relationship will be stronger when couple agreement is higher than when it is lower (see Figure 1). In sum, the aims of this research are: . to examine couple agreement when rating their own and each other’s WIF; . to determine if couples agree more when rating the female than the male target’s WIF; and
Examining couple agreement about WFC 257
JMP 23,3
258
Figure 1. Predicted relationships between couple agreement, WIF, and organizational commitment
.
to determine if couple agreement when rating WIF moderates the relationship between his/her WIF and his/her affective and continuance commitment.
Method Participants The sample included 224 heterosexual dual-earner couples who were married, cohabiting, or in a serious relationship and both parents and non-parents. To participate, each partner had to work a minimum of 20 hours per week (cf. Frone et al., 1992b, 1997) and could not be self-employed. The average participant age was 36 years (range ¼ 18-67, SD ¼ 12.3). Of the participants, 51 percent had at least a bachelor’s degree. Of the couples, 73 percent were married, 14 percent were in a serious relationship but not cohabitating, 7 percent were cohabitating, and 6 percent were engaged. Of those that reported race, 78 percent were Caucasian, 7 percent were Hispanic, 5 percent were Asian, 5 percent were African American, 2 percent were Native American, and 3 percent reported their race as “Other”. On average, participants spent 41 hours a week in paid
employment (range ¼ 20-80 hours, SD ¼ 14.42) and spent 20 hours each week in family activities (range ¼ 0.5-80 hours, SD ¼ 14.01). Procedure Two methods were used to recruit couples. First, e-mails were sent to working adults through networking to recruit subjects across the USA. Second, students at a Midwestern university received extra credit for each couple recruited. The combined response rate (percentage of those who agreed to participate where both partners completed the survey) was 71 percent. The survey was administered online. Each couple that agreed to participate was provided with the website address and two code numbers which were used to match the surveys. Participants completed measures about their own attitudes as well as their perceptions of their partner’s attitudes.
Examining couple agreement about WFC 259
Measures The survey assessed perceptions and attitudes of each member of the couple. To assess perceptions of partner’s attitudes, the items for each variable were adapted by the researchers to reflect “my partner” rather than “I” statements. Work interfering with family. WIF was assessed with five items from Netemeyer et al. (1996). Responses were provided on a five-point Likert scale from 1 (strongly disagree) to 5 (strongly agree). Participants rated their own WIF as well as their perceptions of their partners’ WIF. A sample self-rated WIF item is “My job produces strain that makes it difficult to fulfill family duties”. A sample item to rate their partners’ WIF is “Due to work-related duties, my partner has to make changes to his/her plans for family activities”. Coefficient alpha for self-rated WIF was 0.90 and for partner-rated WIF was 0.91. Organizational commitment. Continuance and affective commitment were assessed using Meyer et al.’s (1993) scale. Each component was measured with six items on a five-point Likert scale with responses ranging from 1 (strongly disagree) to 5 (strongly agree). A sample continuance commitment item is “I feel that I have too few options to consider leaving this organization”. A sample affective commitment item is “This organization has a great deal of personal meaning for me”. Coefficient a was 0.82 for affective commitment and 0.70 for continuance commitment. Results Table I depicts descriptive statistics and correlations. H1 was tested with the correlation between self-rated and partner-rated WIF. The correlation was positive and significant (r ¼ 0:48, p , 0:01), supporting H1. H1 was also tested by calculating rwgj between self and partner ratings. The rwgj is a measure of perceptual agreement, Variable 1. 2. 3. 4.
WIF self ratinga WIF partner ratinga Affective commitmenta Continuance commitmenta a
Mean
SD
2.56 2.39 3.43 3.03
0.96 0.93 0.83 0.72
1 (0.90) 0.48 * * 216 * * 0.30 * *
2 (0.91) 218 * * 0.17 * *
Notes: Work interference with family (WIF), n ¼ 446: *p # 0:05; * *p # 0:01
3
(0.82) 2 26 * *
4
(0.70)
Table I. Correlations, means, standard deviations and reliabilities for study variables
JMP 23,3
260
which, conceptually, indicates the degree to which ratings are the same (e.g. both raters put “1”) (Bliese, 2000). The average rwgj across the sample was 0.79 (the median was 0.95), suggesting reasonable agreement between self and partner ratings of WIF. H2: gender of target and couple agreement about WIF H2 posited that couples would agree more when rating the female target’s WIF than the male target’s WIF. This hypothesis was tested using MANOVA to account for the relationships among self-partner ratings and gender as a set (Edwards, 1995). Male and female ratings of WIF were the dependent variables and rating source (i.e. self versus partner) was the predictor variable. As shown in Table II, the omnibus multivariate test for self and partner ratings indicated that gender of the target was related to the self-partner ratings of WIF (Wilks’ l ¼ 0:98, Fð2; 447Þ ¼ 4:33, p , 0:01). Univariate tests indicated a significant effect for self versus partner ratings on male WIF. As can be seen in Table III, the male’s average self-rating of WIF was significantly higher than his female partner’s rating for him (M male ¼ 2:72 versus M female ¼ 2:45, respectively). In contrast, there was no difference between self-rated and partner-rated WIF for the female target (M self ¼ 2:47; M partner ¼ 2:41). Thus, results support H2; there was more agreement in rating WIF for female partners than for male partners. H3 and H4: couple agreement as a moderator H3 posited that couple agreement about the target’s WIF would moderate the relationship between the target’s self-rated WIF and affective commitment. To test this hypothesis, couple agreement about the target’s WIF was operationalized by the difference between the self- and partner-rated WIF. Moderated regression was used with the target’s self-rated affective commitment as the dependent variable, the target’s self-rated WIF entered into Block 1, couple agreement (i.e. the difference between self-rated and partner-rated WIF) about the target’s WIF in Block 2, and the interaction of self-rated WIF and the difference score in Block 3. Support was not found for H3; agreement did not
df Overall Wilks’ l Canonical correlation
0.98 0.14
Self-partner comparisons F
2,449
4.33 *
Self Table II. Multivariate analysis of variance for self and partner ratings of WIF
Table III. Means for self and partner ratings of WIF by target gender
Gender Error Total
df 1 448 449
Partner F 8.66 *
df 1 448 449
F 0.52
Note: *p , 0:05
Male partner Female partner
Mean self rating
Mean partner rating
2.72 2.47
2.45 2.40
moderate the relationship between self-rated WIF and affective commitment for females (R 2 ¼ 0:00, p . 0:05; see Table IV) or males (R 2 ¼ 0:00, p . 0:05; see Table V). However, self-rated WIF was negatively related to self-rated affective commitment for both females (R 2 ¼ 0:02, p , 0:05, b ¼ 20:15, p , 0:05) and males (R 2 ¼ 0:02, p , 0:05, b ¼ 20:15, p , 0:05). In other words, higher WIF was associated with lower affective commitment for both male and female members of the couple. H4 specified that the relationship between the target’s WIF and continuance commitment would be stronger and more positive when couple agreement was higher. H4 was tested using moderated regression, with the target’s self-rated continuance commitment as the dependent variable, and the entry of variables the same as those used to examine H3. Self-rated WIF was positively related to self-rated continuance commitment for both females and (R 2 ¼ 0:10, p , 0:05, b ¼ 0:31, p , 0:05; see Table VI) males (R 2 ¼ 0:08, p , 0:05, b ¼ 0:22, p , 0:05; see Table VII). That is, for both females and males those experiencing higher WIF reported greater continuance commitment. This relationship was not moderated by couple agreement for males, in opposition to H4. In contrast, for females, the relationship between self-rated WIF and continuance commitment was moderated by couple agreement about female WIF, with the interaction accounting for an increment of 2 percent in the variance in continuance commitment (see Table VIII). Aiken and West’s (1991) procedures were used to graph and interpret the interaction. Figure 2 reveals that the relationship between the female self-rated WIF and her continuance commitment was weaker for couples with high agreement (b ¼ 0:14, p , 0:05) and stronger for couples with low agreement (b ¼ 0:31, Variable
B
SEB
b
2 0.13
0.06
2 0.15 *
Step 2 Couple agreement about the female partner’s WIF
0.10
0.07
0.12
Step 3 Interaction of WIF self and agreement
0.06
0.05
0.21
Step 1 WIF self-rated
R2
DR 2
0.02 *
0.02 *
0.03
0.01
0.04
0.00
Notes: WIF self-rated is the female WIF self-report rating. Couple agreement is the female difference score. *p # 0:05; n ¼ 218
Variable
B
SEB
b
Step 1 WIF self-rated
2 0.14
0.06
20.15 *
Step 2 Couple agreement about the male partner’s WIF
2 0.02
0.07
20.02
Step 3 Interaction of WIF self and agreement
2 0.05
0.06
20.18
R2
DR 2
0.02 *
0.02 *
0.02
0.00
0.03
0.00
Notes: WIF self-rated is the male WIF self-reported rating. Couple agreement is the male difference score. *p # 0:05; n ¼ 218
Examining couple agreement about WFC 261
Table IV. Moderated regression analysis for couple agreement about the female partner’s WIF as a moderator of the relationship between self-rated female WIF and her affective commitment
Table V. Moderated regression analysis of couple agreement about the male partner’s WIF as a moderator of the relationship between the male partner’s WIF and his affective commitment
JMP 23,3
262
Table VI. Moderated regression analysis for couple agreement about the female partner’s WIF as a moderator of the relationship between self-rated female WIF and her continuance commitment
Table VII. Moderated regression analysis for couple agreement about the male partner’s WIF as a moderator of the relationship between male self-rated WIF and his continuance commitment
Table VIII. Post hoc comparisons of direction of disagreement and WIF
p , 0:05). That is, although females who experience WIF have greater continuance commitment, this relationship is ameliorated when couple agreement about female WIF is high. Thus, the nature of the interaction was opposite of what was hypothesized. A post hoc analysis was conducted to examine the nature of the low agreement relationship (i.e. when the difference between self-rated and partner-rated WIF was high). We were interested in understanding circumstances in which females rated their WIF higher than their male partners rated it and situations in which male partners rated their wife’s/partner’s WIF higher than she rated it. Thus, the female difference score was dummy-coded. Scores where the female’s ratings of her WIF were higher than her partners were coded as 0, whereas scores where her partner’s rating of her Variable
B
SEB
Step 1 WIF self-rated
0.23
0.05
0.31 * *
Step 2 Couple agreement about the female partner’s WIF
0.03
0.06
0.04
2 0.09
0.04
20.36 * *
Step 3 Interaction of WIF self-rated and agreement
b
R2
DR 2
0.10 * *
0.10 * *
0.10
0.00
0.12 *
0.02 *
Notes: WIF self-rated is the female WIF self-reported rating. Couple agreement is the female difference score. *p # 20:05; * *p # 0:01; n ¼ 218
Variable
B
SEB
Step 1 WIF self-rated
0.22
0.05
2 0.03
0.06
20.03
0.05
0.05
0.22
Step 2 Couple agreement about the male partner’s WIF Step 3 Interaction of WIF self and agreement
R2
DR 2
0.08 * *
0.08 * *
0.08
0.00
0.08
0.01
b 0.28 * *
Notes: WIF self-rated is the male WIF self-reported rating. Couple agreement is the male difference score. * *p # 0:01; n ¼ 218
Variable
B
SEB
WIF self Dummy-coded agreement (1 ¼ self rating is lower than partner) Interaction of WIF self and agreement
0.14
0.06
20.40 0.21
0.26 0.11
B 0.14 0.35
SE 0.06 0.09
Simple slopes WIF where self is higher than partner WIF where self is lower than partner
Note: Dependent variable ¼ female continuance commitment
R2
DR 2
0.08
0.08
20.28 0.32
0.08 0.10
0.00 0.02
t 2.30 4.07
p 0.02 0.00
b 0.20 *
WIF were higher than her ratings were coded as 1. The relationship was tested with moderated regression using the dummy coded variable (male rating higher versus female rating higher), WIF, and their interaction as predictors with continuance commitment as the dependent variable. Results (see Table VIII) indicate that the interaction was significant (DR 2 ¼ 0:02, p , 0:05). When a female rated her own WIF low relative to her partner, the relationship between WIF and continuance commitment was stronger (b ¼ 0:35, t ¼ 4:07, p , 0:05), whereas when a female rated her WIF higher relative to her partner the relationship between WIF and continuance commitment, while still significant, was lower in magnitude (b ¼ 0:14, t ¼ 2:30, p , 0:05). This interaction is depicted in Figure 3.
Examining couple agreement about WFC 263
Discussion This study expands our knowledge of work-to-family conflict by examining the role of both self and partner perceptions of WIF. Specifically, we examined couple agreement
Figure 2. Couple agreement about female WIF as a moderator of the relationship between female’s self-reported WIF and her continuance commitment
Figure 3. Low agreement moderator: male and female rating of female’s WIF
JMP 23,3
264
when rating WIF. The current study contributes to the literature in several ways. First, we found that couples exhibit substantial agreement when rating the WIF experienced by both male and female partners. Second, we found that couple agreement was higher when rating the female target’s WIF than when rating the male target’s WIF. Finally, couple agreement about the female target’s WIF moderated the relationship between her self-rated WIF and her continuance commitment, such that this relationship was weaker when couple agreement was high. Couple agreement about WIF Across two distinct operationalizations of couple agreement (i.e. correlations and rwgj), our findings reveal agreement about the level of WIF each individual experiences. These findings are consistent with Matthews et al. (2006), who found substantial agreement between members of a couple regarding the degree to which work-to-relationship conflict was experienced. The degree to which one member of a couple could articulate his or her partner’s level of WIF suggests that couples that participated in this study may communicate about work-family concerns. Given that employees’ partners appear to be aware of the impact that work can have on family, it is also likely that individuals possess attitudes toward not only their own organizations, but their partners’ employers as well. Although agreement about WIF in rating both male and female targets was reasonably high, couples reported more similar ratings of the female target’s WIF than the male target’s WIF. This finding was consistent with our hypothesis. Because women are more communicative than men (Edwards and Hamilton, 2004), they may be more likely to share their WIF with their partners, better enabling their male partners to understand the level of conflict they experience. It may also be the case that because women are responsible for more family tasks (Fu and Shaffer, 2001), when work interferes with family, it is more evident to their male partner. In contrast, if men are responsible for fewer family tasks, or there is disagreement among the two members of the couple regarding the degree to which the male partner should be responsible for duties at home, less agreement about his WIF would be expected. We suspect that sex-role socialization around the roles men and women have in the family system may underlie the greater agreement about the female partner’s WIF. Future research might seek to understand whether the greater couple agreement regarding the female target’s WIF is due to her greater tendency to communicate, her greater responsibility for family, both these phenomena, or some other factor entirely. WIF, organizational commitment, and couple agreement as a moderator Consistent with prior research, higher WIF was associated with lower levels of affective commitment among both women and men. This is consistent with previous research findings of a negative relationship between WFC and affective commitment (Allen et al., 2000; Siegel et al., 2005). Thus, when work interferes with family employees feel less emotionally attached to their organizations. Self-reported WIF and continuance commitment were also positively related as in previous studies (Casper et al., 2002; Lyness and Thompson, 1997; Good et al., 1988). Since affective commitment is positively correlated with job performance, continuance commitment is negatively correlated with performance (Meyer et al., 1989, 2002), and WIF is associated with lower affective commitment and higher continuance commitment, organizational
efforts to support employees’ family responsibilities may weaken the effects of WIF and, through influencing distinct types of commitment, improve employee performance. A unique contribution of the current study was the examination of couple agreement about the level of WIF as a moderator in an attempt to discern whether such agreement about this important experience (i.e. WIF) might reinforce the effects of WIF on work attitudes. This variable was explored as a moderator of four relationships: (1) the WIF-affective commitment relationship for women; (2) the WIF-affective commitment relationship for men; (3) the WIF-continuance commitment relationship for women; and (4) the WIF-continuance commitment relationship for men. Of these, the only significant moderating effect was for the relationship between female self-rated WIF and her continuance commitment. Interestingly, although couple agreement about the female target’s WIF did indeed moderate this relationship, the relationship between female WIF and her continuance commitment was stronger for low agreement couples and weaker for couples with high agreement, counter to the hypothesis. Thus, when a woman perceived her level of WIF to be similar to what her partner believed it to be, the relationship between her WIF and her continuance commitment was weaker. This finding may reflect a buffering effect such that, higher couple agreement regarding the female’s WIF occurs among women who have a greater sense of social support from their partners. Couples would be expected to agree more when the female partner shares her stresses and her male partner listens effectively and hears what she says. These same circumstances are also likely to facilitate a sense of partner support given attentive listening, validation, and expressiveness are indicators social support (cf. Julien et al., 2003). Thus, when couples communicate effectively about the woman’s work-family conflict, this is likely to facilitate both high agreement regarding the level of her WIF and the woman’s sense that her partner supports her. However, given that we did not measure perceived partner support directly, future research is needed to determine if agreement about the level of WIF experienced is indeed associated with increased perceived support as we propose. The notion that these findings are based on a social support phenomenon is consistent with research that suggests that spouse support can reduce work-family conflict and ameliorate any negative effects of such conflict (Beutell and Greenhaus, 1982; Carlson and Perrewe, 1999; Holohan and Gilbert, 1979). These findings also parallel models that depict social support as a moderator of the stressor-strain relationship (Ganster et al., 1986; Greenhaus and Parasuraman, 1986; Ray and Miller, 1994) and findings that support can reduce the deleterious effects of WFC on organizational commitment (Casper et al., 2002; Boumans and Landeweerd, 1992). Post hoc analyses of the moderated relationship indicated that when couples disagreed about the level of WIF the female partner experienced, and the male perceived her WIF to be higher, the female partner experienced a greater sense of being trapped in her organization (i.e. continuance commitment). In contrast, when the disagreement entailed the female perceiving her WIF as higher than her male partner perceived it, the relationship between WIF and continuance commitment, although still
Examining couple agreement about WFC 265
JMP 23,3
266
positive, was much weaker in magnitude. One plausible explanation for this finding is attribution theory, which asserts that individuals make sense of situations they experience, and develop causal explanations based on characteristics of the person (internal attributions) or the environment (external attributions) (Kelley, 1973). Continuance commitment reflects an attribution about “why I stay” in an organization. In other words, a woman who has high continuance commitment attributes her continued tenure to necessity rather than desire. Since women are socialized to place great value on family (Cooper et al., 1994), when her partner perceives her as allowing work to interfere with family she may get pressure from him to attend more to family and less to work. Under such circumstances, women who have been socialized to put their family above their work may need to attribute their reason for remaining with the organization to necessity (e.g. financial reasons, lack of alternatives) rather than desire in order to ameliorate their guilt. This is consistent with Casper et al.’s (2002) proposition that working mothers that experienced WIF and yet remained with their organizations may attribute their continued employment to need (continuance commitment) to ameliorate any guilt associated with WIF. Practical implications and strengths In addition to contributing to theory, this study has a number of practical implications that should be considered by organizational decision makers who are determining how to deal with employees’ work-family concerns. First, our results suggest that organizations that require employees to work long hours should realize that WIF has negative consequences to the organization. Specifically, WIF contributes to decreased affective commitment and increased continuance commitment. Given employees who remain with their organizations because they want to have higher performance and those that remain because they believe they have to have lower performance (Meyer et al., 2002), employees with high WIF are likely to have poorer job performance compared to those with less conflict. At a general level, our findings also suggest that employees do not operate in isolation, but that partnered employees function within the context of a dyadic unit. Our findings indicate that employees’ partners are fairly aware of the degree to which their work interferes with family. Thus, the negative attitudes that exist toward an organization as a result of WIF are likely to exist not only among employees but also their partners. Given findings that spouse attitudes play a critical role in employee decisions about their continuing employment (Eby and Russell, 2000; Rosen and Durand, 1995), partners’ attitudes toward the organization clearly have important consequences. In fact, our findings indicate that when female employees and their male partners agree that WIF is problematic, this agreement reinforces her negative attitudes toward the organization. Such findings suggest it is important for organizations to consider not only how their actions are perceived by employees, but also by employees’ partners, and this appears to be more true when considering the male partner of female employees. Accordingly, organizations might take opportunities that are available to them to engage employees’ spouses and partners in a positive way. For example, organizations can incorporate employees’ partners into formal employee assistance programs, or include them in organization-wide social activities such as employee picnics or office holiday parties. An organization that is noted for its success at engaging spouses is the military. For example, the US Army
has implemented Family Support Groups which provide social and emotional support, as well as information, for soldiers and their families (Shumm et al., 2000). The partner perceptions that appeared to have the greatest potential for a deleterious effect on employee attitudes were male partners’ perceptions that their wife’s/female partner’s WIF was higher than she believed it was. Thus, efforts to enhance employee attitudes through reducing partner perceptions of WIF may be most successful when targeted at male partners of female employees. These findings are consistent with the notion that sex-role stereotypes still govern how men and women manage their work and family roles (Cleveland et al., 2000). That is, female employees may face more difficulty from their husbands due to work demands than male employees face from their wives. Accordingly, organizations that are committed to the career development and promotion of women should be aware that organizational supports to ameliorate WIF may be important to help women handle the pressures and expectations they face at both work and home. Such supports may take the form of individual manager support or organizational policies and practices. In terms of implications for manager behavior, findings suggest that managers should be attentive when employees discuss concerns surrounding their partner and family. For instance, if an employee shares concerns with a supervisor that her husband is uncomfortable with her excessive business travel or irregular or long hours, managers should consider these legitimate concerns which could lead to negative work attitudes if not resolved. Accordingly, managers who are attentive to such employees’ family concerns and resolve them whenever possible may benefit by facilitating affective rather than continuance commitment among their employees. With respect to organizational policies and practices, results suggest there may be value to organizations in providing alternative work arrangements such as flexible work schedules and telecommuting options. Alternative schedules could decrease the perception that work is interfering with family for employees as well as their partners, thereby decreasing backlash against the organization, i.e. less affective and more continuance commitment. The current study has a number of strengths that enhance its ability to contribute to the literature. First, answering calls to examine both members of a couple in studying work-family phenomena (Casper et al., 2007), we obtained self ratings and partner ratings of WIF. This enabled us to explore couple agreement regarding each member’s experience of WIF, which has received limited attention in past studies. We also sampled a diverse group of working adults, including both married couples as well as couples in partnerships, which did not reflect legal marriages. Since more couples today are cohabiting without marrying (Rhoades et al., 2006), it is important to expand our conception of “couple” in order to reflect the partnerships that exist in today’s world. Thus, this study helps advance research by exploring work-family conflict within the context of diverse dyadic relationships using measures with good reliability and validity evidence. Limitations and directions for future research As with all research, there are limitations of the present study that should be noted. Because convenience sampling was used, the sample may not represent couples in the population and caution is urged in generalizing. Future studies should solicit organizations that allow data to be collected from employees and their significant
Examining couple agreement about WFC 267
JMP 23,3
268
others. Our study also focused exclusively on heterosexual couples, so conclusions about work-family conflict among same-sex couples are not possible. Furthermore, because couple agreement may change over time (Yogev and Brett, 1985), future longitudinal research is warranted. Future research should also explore couple agreement about other important issues such as work-family values and job attitudes. It is also important to note that this study examined work-family conflict from a single direction – work interfering with family. This was appropriate given “other” ratings were given by a romantic partner who experienced and observed their partners’ WIF. However, research clearly suggests that family can also interfere with work (FIW) and future research should consider the importance of the degree of agreement between self and other perceptions of FIW. Such studies using supervisor/subordinate dyads would be very relevant, as disagreements regarding FIW might have important implications for supervisor-subordinate relationships. Conclusions The current study highlights the importance of evaluating couple agreement with respect to employee and partner levels of WIF. Findings suggest that individuals within a couple tend to perceive the level of WIF each partner experiences more similarly than differently. Couple agreement about the female partner’s WIF buffers the impact of her self-rated WIF on her attitudes toward her organization. Clearly, partnered employees do not operate in a vacuum and we should not underestimate the degree to which employees’ partners influence the employer-employee relationship. This is especially true when considering a female employee and her male partner. Thus, when employers consider not only employee perceptions, but also the perspective of employees’ partners, they may develop higher quality social exchange relationships. References Aiken, L.S. and West, S.G. (1991), Multiple Regression: Testing and Interpreting Interactions, Sage Publications, Thousand Oaks, CA. Allen, T., Herst, D., Bruck, C. and Sutton, M. (2000), “Consequences associated with work-to-family conflict: a review and agenda for future research”, Journal of Occupational Health Psychology, Vol. 5, pp. 278-308. Beutell, N.J. and Greenhaus, J.H. (1982), “Interrole conflict among married women: the influence of husband and wife characteristics on conflict and coping behavior”, Journal of Vocational Behavior, Vol. 21, pp. 99-110. Bliese, P.D. (2000), “Within-group agreement, non-independence, and reliability: implications for data aggregation and analysis”, in Klein, K.J. and Kozlowksi, S.W.J. (Eds), Multilevel Theory, Research, and Methods in Organizations: Foundations, Extensions, and New Directions, Jossey-Bass, San Francisco, CA. Boumans, N.P. and Landeweerd, J.A. (1992), “The role of social support and coping behaviour in nursing work: main or buffering effect?”, Work and Stress, Vol. 6, pp. 191-202. Bronfenbrenner, U. (1979), The Ecology of Human Development: Experiments by Nature and Design, Harvard University Press, Cambridge, MA. Bureau of Labor Statistics (2003), Women in the Work Force: A Databook, p. 10, available at: www.bls.gov (accessed January 10, 2007).
Caligiuri, P.M., Hyland, M.M., Joshi, A. and Bross, A.S. (1998), “Testing a theoretical model for examining the relationship between family adjustment and expatriates’ work adjustment”, Journal of Applied Psychology, Vol. 83, pp. 598-614. Carlson, D.S. and Perrewe, P.L. (1999), “The role of social support in the stressor-strain relationship: an examination of work-family conflict”, Journal of Management, Vol. 25, pp. 513-40. Casper, W.J., Martin, J.A., Buffardi, L.C. and Erdwins, C.J. (2002), “Work-family conflict, perceived organizational support, and organizational commitment among employed mothers”, Journal of Occupational Health Psychology, Vol. 7, pp. 99-108. Casper, W.J., Eby, L.T., Bordeaux, C., Lockwood, A. and Lambert, D. (2007), “A review of research methods in IO/OB work-family research”, Journal of Applied Psychology, Vol. 92, pp. 28-43. Cleveland, J.N., Stockdale, M. and Murphy, K.R. (2000), Women and Men in Organizations: Sex and Gender Issues at Work, Lawrence Erlbaum Associates, Mahwah, NJ. Cooper, S.E., Arkkelin, D.L. and Tiebert, M.J. (1994), “Work-relationship values and gender roles differences in relationship to career-marriage aspirations”, Journal of Counseling and Development, Vol. 73, pp. 63-78. Coverman, S. (1989), “Role overload, role conflict, and stress: addressing consequences of multiple role demands”, Social Forces, Vol. 67, pp. 965-82. Cross, S.E. and Madson, L. (1997), “Models of the self: self-construals and gender”, Psychological Bulletin, Vol. 122, pp. 5-37. Dornstein, M. and Matalon, Y. (1989), “A comprehensive analysis of the predictors of organizational commitment: a study of voluntary army personnel in Israel”, Journal of Vocational Behavior, Vol. 34, pp. 192-203. Doumas, D.M., Margolin, G. and John, R.S. (2003), “The relationship between daily marital interaction, work, and health-promoting behaviors in dual-earner couples”, Journal of Family Issues, Vol. 24, pp. 3-20. Eagly, A.H. and Johannesen-Schmidt, M.C. (2001), “The leadership styles of women and men”, Journal of Social Issues, Vol. 57, pp. 781-97. Eby, L.T. and Russell, J.E. (2000), “Predictors of employee willingness to relocate for the firm”, Journal of Vocational Behavior, Vol. 57, pp. 42-61. Eby, L.T., Casper, W.J., Lockwood, A., Bordeaux, C. and Brinley, A. (2005), “Work and family research in IO/OB: content analysis and review of the literature (1980-2002)”, Journal of Vocational Behavior, Vol. 66, pp. 124-97. Edwards, J.R. (1995), “Alternatives to difference scores as dependent variables in the study of congruence in organizational research”, Organizational Behavior and Human Decision Processes, Vol. 64, pp. 307-24. Edwards, R. and Hamilton, M.A. (2004), “You need to understand my gender role: an empirical test of Tannen’s model of gender and communication”, Sex Roles, Vol. 50, pp. 491-504. Frone, M., Russell, M. and Cooper, M.L. (1992a), “Antecedents and outcomes of work-family conflict: testing a model of the work-family interface”, Journal of Applied Psychology, Vol. 77, pp. 65-78. Frone, M.R., Russell, M. and Cooper, M.L. (1992b), “Prevalence of work-family conflict: are work and family boundaries asymmetrically permeable?”, Journal of Organizational Behavior, Vol. 13, pp. 723-9. Frone, M.R., Yardley, J.K. and Markel, K.S. (1997), “Developing and testing an integrative model of the work-family interface”, Journal of Vocational Behavior, Vol. 50, pp. 145-67.
Examining couple agreement about WFC 269
JMP 23,3
270
Fu, C.K. and Shaffer, M.A. (2001), “The tug of work and family: direct and indirect domain-specific determinants of work-family conflict”, Personnel Review, Vol. 30, pp. 502-22. Ganster, D.C., Fusilier, M.R. and Mayes, B.T. (1986), “Role of social support in the experience of stress at work”, Journal of Applied Psychology, Vol. 71, pp. 102-10. Gareis, K.C., Barnett, R.C. and Brennan, R.T. (2003), “Individual and crossover effects of work schedule fit: a within-couple analysis”, Journal of Marriage and Family, Vol. 65, pp. 1041-54. Good, L.K., Sisler, G.F. and Gentry, J.W. (1988), “Antecedents of turnover intentions among retail management personnel”, Journal of Retailing, Vol. 64, pp. 295-314. Greenhaus, J. and Parasuraman, S. (1986), “A work-nonwork interactive perspective of stress and its consequences”, in Ivancevich, J.M. and Ganster, D.C. (Eds), Job Stress: From Theory to Suggestion, Haworth Press, New York, NY. Greenhaus, J.H. and Beutell, N.J. (1985), “Sources of conflict between work and family roles”, Academy of Management Review, Vol. 10, pp. 76-88. Greenhaus, J.H., Collins, K.M., Singh, R. and Parasuraman, S. (1997), “Work and family influences on departure from public accounting”, Journal of Vocational Behavior, Vol. 50, pp. 249-70. Hammer, L.B., Allen, E. and Grigsby, T.D. (1997), “Work-family conflict in dual-earner couples: within individual and crossover effects of work and family”, Journal of Vocational Behavior, Vol. 185, p. 203. Hammer, L.B., Bauer, T.N. and Grandey, A.A. (2003), “Work-family conflict and work-related withdrawal behaviors”, Journal of Business and Psychology, Vol. 17, pp. 419-36. Hammer, L.B., Neal, M.B., Newsom, J.T., Brockwood, K.J. and Colton, C.L. (2005), “A longitudinal study of the effects of dual-earner couples’ of family-friendly workplace supports on work and family outcomes”, Journal of Applied Psychology, Vol. 90, pp. 799-810. Hayden, L.C., Schiller, M., Dickstein, S., Seifer, R., Sameroff, A.J., Miller, I., Keitner, G. and Rasmussen, S. (1998), “Levels of family assessment: I. Family, marital, and parent-child interaction”, Journal of Family Psychology, Vol. 12, pp. 7-22. Heilman, M.E. (2001), “Description and prescription: how gender stereotypes prevent women’s ascent up the organizational ladder”, Journal of Social Issues, Vol. 57, pp. 657-74. Heilman, M.E. and Okimoto, T.G. (2007), “Why are women penalized for success at male tasks? The implied communality deficit”, Journal of Applied Psychology, Vol. 92, pp. 81-92. Holohan, C.K. and Gilbert, L.A. (1979), “Conflict between major life roles: women and men in dual career couples”, Human Relations, Vol. 326, pp. 451-67. Julien, D., Chartrand, E., Simard, M.C., Bouthillier, D. and Begin, J. (2003), “Conflict, social support, and relationship quality: an observational study of heterosexual, gay male, and lesbian couples’ communication”, Journal of Family Psychology, Vol. 17, pp. 419-28. Kelley, H.H. (1973), “The processes of causal attribution”, American Psychologist, Vol. 28, pp. 107-28. Kenny, D.A. and Acitelli, L.K. (2001), “Accuracy and bias in the perception of the partner in a close relationship”, Journal of Personality and Social Psychology, Vol. 80, pp. 439-48. Kossek, E.E. and Ozeki, C. (1998), “Work-family conflict, policies, and the job-life satisfaction relationship: a review and directions for future organizational behavior-human resources research”, Journal of Applied Psychology, Vol. 83, pp. 139-49. Kring, A.M. and Gordon, A.H. (1998), “Sex differences in emotion: expression, experience, and physiology”, Journal of Personality and Social Psychology, Vol. 74, pp. 686-703.
Lundberg, U. and Frankenhaeuser, M. (1999), “Stress and workload of men and women in high-ranking positions”, Journal of Occupational Health Psychology, Vol. 4, pp. 142-51. Lyness, K.S. and Thompson, D.E. (1997), “Above the glass ceiling? A comparison of matched samples of female and male executives”, Journal of Applied Psychology, Vol. 82, pp. 359-75. Lytton, H. and Romeny, D.M. (1991), “Parents’ differential socialization of boys and girls: a meta-analysis”, Psychological Bulletin, Vol. 109, pp. 267-96. McCreary, D.R. (1994), “The male role and avoiding femininity”, Sex Roles, Vol. 31, pp. 517-31. McElwain, A.K., Korabik, K. and Rosin, H.M. (2005), “An examination of gender differences in work-family conflict”, Canadian Journal of Behavioural Science, Vol. 37, pp. 283-98. Matthews, R.A., Del Priore, R.E., Acitelli, L.K. and Barnes-Farrell, J.L. (2006), “Work-to-relationship conflict: crossover effects of dual-earner couples”, Journal of Occupational Health Psychology, Vol. 11, pp. 228-40. Meyer, J.P. and Allen, N.J. (1991), “A three-component conceptualization of organizational commitment”, Human Resource Management Review, Vol. 1, pp. 61-98. Meyer, J.P., Allen, N.J. and Smith, C.A. (1993), “Commitment to organizations and occupations: extension and test of a three-component conceptualization”, Journal of Applied Psychology, Vol. 78, pp. 538-51. Meyer, J.P., Stanley, D.J., Herscovitch, L. and Topolnytsky, L. (2002), “Affective, continuance, and normative commitment to the organization: a meta-analysis of antecedents, correlates, and consequences”, Journal of Vocational Behavior, Vol. 61, pp. 20-52. Meyer, J.P., Paunonen, S.V., Gellatly, I.R., Goffin, R.D. and Jackson, D.N. (1989), “Organizational commitment and job performance: it’s the nature of the commitment that counts”, Journal of Applied Psychology, Vol. 74, pp. 52-156. Moen, P. and Yu, Y. (2000), “Effective work/life strategies: working couples, work conditions, gender, and life quality”, Social Problems, Vol. 47, pp. 291-326. Netemeyer, R.G., Boles, J.S. and McMurrian, R. (1996), “Development and validation of work-family conflicts and work-family conflict scales”, Journal of Applied Psychology, Vol. 81, pp. 400-10. O’Brien, M. and Peyton, V. (2002), “Parenting attitudes and marital intimacy: a longitudinal analysis”, Journal of Family Psychology, Vol. 16, pp. 118-27. Ray, E.B. and Miller, K.I. (1994), “Social support, home/work stress and burnout: who can help?”, Journal of Applied Behavioral Science, Vol. 30, pp. 357-73. Rhoades, G.K., Stanley, S.M. and Markman, H.J. (2006), “Pre-engagement cohabitation and gender asymmetry in marital commitment”, Journal of Family Psychology, Vol. 20, pp. 553-60. Rosen, L.N. and Durand, D.B. (1995), “The family factor and retention among married soldiers deployed in Operation Desert Storm”, Military Psychology, Vol. 7, pp. 221-34. Rusbult, C.E., Bissonnette, V.L., Arriaga, X.B. and Cox, C.L. (1998), “Accommodation processes during the early years of marriage”, in Bradbury, T.N. (Ed.), The Developmental Course of Marital Dysfunction, Cambridge University Press, New York, NY. Shaffer, M.A. and Harrison, D.A. (1998), “Expatriates’ psychological withdrawal from international assignments: work, nonwork, and family influences”, Personnel Psychology, Vol. 51, pp. 87-118. Siegel, P.A., Post, C., Brockner, J., Fishman, A.Y. and Garden, C. (2005), “The moderating influence of procedural fairness on the relationship between work-life conflict and organizational commitment”, Journal of Applied Psychology, Vol. 90, pp. 13-24.
Examining couple agreement about WFC 271
JMP 23,3
272
Shumm, W.R., Bell, D.B., Milan, L.M. and Segal, M.W. (2000), The Family Support Group (FSG) Leaders’ Handbook, Tech Report 2000-02, US Army Research Institute for the Behavioral and Social Sciences, Alexandria, VA. Sillars, A.L., Pike, G.R., Jones, T.S. and Murphy, M.A. (1984), “Communication and understanding in marriage”, Human Communication Research, Vol. 10, pp. 317-50. Thompson, C.A., Beauvais, L.L. and Lyness, K.S. (1998), “When work-family benefits are not enough: the influence of work-family culture on benefit utilization, organizational attachment, and work-family conflict”, Journal of Vocational Behavior, Vol. 54, pp. 392-415. US Census Bureau (2005), Statistical Abstract of the United States: 2005, 125th ed., US Census Bureau, Washington, DC. Westman, M. and Etzion, D. (1995), “Crossover of stress, strain and resources from one spouse to another”, Journal of Organizational Behavior, Vol. 16, pp. 169-81. Westman, M. and Etzion, D. (2001), “Job insecurity and crossover of burnout in married couples”, Journal of Organizational Behavior, Vol. 22, pp. 467-81. Whiting, B.B. and Edwards, C.P. (1988), Children of Different Worlds: The Formation of Social Behavior, Harvard University Press, Cambridge, MA. Yogev, S. and Brett, J. (1985), “Patterns of work and family involvement among single- and dual-earner couples”, Journal of Applied Psychology, Vol. 70, pp. 754-68. About the authors Michelle Streich is employed by Edward Jones as a Performance Development Specialist, where she conducts training and leadership development programs. She received her PhD in Industrial/Organizational Psychology from the University of Tulsa. She has worked in areas such as recruitment, selection, training, and organizational development. Her primary research areas are work-family issues, workplace romance, and gender-related studies. Michelle Streich is the corresponding author and can be contacted at:
[email protected] Wendy J. Casper is an Assistant Professor of Management at the University of Texas at Arlington. She received her PhD in Industrial/Organizational Psychology from George Mason University. She worked for a number of years as a human resource practitioner and has consulted to numerous organizations in both the public and private sector regarding human resource management and work-family issues. She has published 25 academic and professional papers on work-family issues and human resource practices and presented over 50 papers at professional conferences. Her work has appeared in numerous outlets, including Journal of Applied Psychology, Journal of Vocational Behavior, Journal of Occupational Health Psychology, Human Resource Management, Human Performance, and Applied Psychology: An International Review. Her research has also been discussed by numerous media sources, including The Washington Post and MSNBC. Amy Nicole Salvaggio is an Assistant Professor at the University of Tulsa. She received her PhD in Industrial/Organizational Psychology from the University of Maryland. She has consulted for organizations in areas related to customer service and organizational climate. Her diverse areas of research include gender bias within organizations, workplace romance, organizational climate, and work-family conflict. She has published her research in journals such as the Journal of Applied Psychology, Educational and Psychological Measurement, and International Journal of Aviation Psychology.
To purchase reprints of this article please e-mail:
[email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0268-3946.htm
Positive effects of nonwork-to-work facilitation on well-being in work, family and personal domains
Non-work to work facilitation
Pam Allis and Michael O’Driscoll
Received March 2007 Revised October 2007 Accepted November 2007
University of Waikato, Hamilton, New Zealand
273
Abstract Purpose – The paper seeks to examine whether spillover from “nonwork” to work contributes to individuals’ well-being. Design/methodology/approach – An online survey was administered to New Zealand local government employees. Positive (facilitation) and negative (conflict) spillover from two “nonwork” domains (family and personal benefit activities) to work were investigated. The survey also assessed psychological involvement (in work, family and personal benefit activities), time devoted to each domain, and self-reported well-being in each area. Findings – Levels of nonwork-to-work facilitation were moderate, and significantly higher than nonwork-to-work conflict, and well-being was moderately high (although greater for the family and personal benefit domains than for work). There were significant positive relationships between psychological involvement in the nonwork domains and levels of facilitation from these domains to work, and nonwork-to-work facilitation was associated with higher well-being. Time invested in family and personal activities was not linked with greater nonwork-to-work conflict. Mediation analyses indicated that psychological involvement (in family and personal activities) was associated with increased facilitation, which in turn enhanced well-being. Practical implications – Engagement in family and personal benefit activities yields positive outcomes for individuals, in terms of their psychological well-being and facilitation of work-related outcomes. Encouragement to engage in these areas can therefore be beneficial for both individuals and their employing organizations. Originality/value – The main contribution of this research is that involvement in personal benefit activities (as another component of the “nonwork” domain, in addition to family activities) can have positive outcomes for individuals, resulting in facilitation of work outcomes and positive well-being. Keywords Sociology of work, Family, Individual psychology Paper type Research paper
The growing interest in work-life balance may originate from changes in many global and local domains. For example, the traditional family consisting of the husband going out to work and the wife staying at home to look after the children is now less common (Greenhaus and Powell, 2006), and statistics demonstrate a greater participation of women in paid work, family structures altering, an aging workforce and the desire for some workers not to be “married to the job”. More families are juggling work and family and having greater difficulty in doing so (Hammer et al., 2005). These changes have affected both work and “nonwork” life and their interaction (Frone, 2003). Consequently, work-life balance is becoming increasingly important but also perhaps increasingly complex.
Journal of Managerial Psychology Vol. 23 No. 3, 2008 pp. 273-291 q Emerald Group Publishing Limited 0268-3946 DOI 10.1108/02683940810861383
JMP 23,3
274
Many researchers have noted the predominance of studies investigating conflict between (predominantly) two domains – work and family. The present study aims to extend this existing knowledge on two fronts. First, Seligman and Csikszentmihalyi (2000) discussed the growing importance of positive psychology, of what makes things better for individuals. Following this suggestion, the present study examined facilitation between domains, with less attention to inter-domain conflict. That is, our aim was primarily to explore the benefits rather than the burdens of inter-domain relationships, and to examine the positive effect of these inter-relationships on psychological well-being. Secondly, Eby et al. (2005) have argued the need to extend research beyond the family component of the “nonwork” domain. With this in mind, the present research explored two dimensions of the nonwork domain – family and personal benefit activities. Our interest was in the potential effects of facilitation between each of these sub-domains and the individual’s work experiences. Although the primary focus of this paper is on the benefits of facilitation, it is recognized that time demands from each specific domain can impinge upon the fulfilment of responsibilities in another domain; hence, we also examine time-based conflict between domains. In particular, we assess the relationships between time demands emanating from the two nonwork domains (family and personal benefit activities) and conflict between these roles and work. Researchers have suggested that the two primary domains for an individual are the work and “nonwork” domains (Frone, 2003). The work domain clearly refers to paid employment (whether full-time or part-time). As mentioned above, in the current study we subdivided the nonwork domain into two distinct components – family life and personal benefit activities. The latter refer to activities which a person undertakes for their own personal benefit. These can fit under several different categories, for instance leisure (e.g. physical activities, sport and hobbies), personal development (e.g. private study, new challenges), spiritual involvement (e.g. religious activities, meditation), and voluntary work. Personal benefits activities are those which benefit the individual. To date, research has predominantly assessed the relationship between work and family (as the presumed major nonwork domain), and studies investigating the relationship between work and specific nonwork roles other than family are sparse. There has been considerable research on the work-leisure interface. The present study does overlap with this interface but personal benefit activities extend beyond the leisure concept. Recent reviews of the literature have noted a preoccupation with inter-domain conflict and have suggested greater recognition of the benefits of multiple roles as well as expanding the knowledge of more nonwork domains (Kirchmeyer, 1992; Frone, 2003; Eby et al., 2005; Greenhaus and Powell, 2006). Perhaps part of the reason for the lack of research on nonwork domains other than family is the challenge of distinguishing the boundaries between the various nonwork domains and the difficulty of positioning discrete nonwork activities in a particular nonwork domain. For example, going for a walk with a friend or kayaking with a club member include clearly a physical element, but also a social element and a personal benefit element. In summary, the present exploratory study not only investigated family-to-work facilitation but expands research with another nonwork variable, by examining the relationship between people taking some time to relax and care for themselves and facilitation from that personal benefit activities variable to the work domain. The Collins Compact Thesaurus (1999) lists as antonyms for work “ease, leisure, rest, play,
recreation” (p. 814). Personal benefit activities attempt to encapsulate a range of activities that include these and more. These activities have been linked to feelings of freedom, intrinsic satisfaction, and positive mood and having beneficial consequences for well-being (Eden, 2001). In the current paper we label this nonwork variable as “personal benefit activities”. This variable generally relates to activities an individual undertakes to take care of themselves. They are a cluster of activities for overall benefit rather than stand-alone individual activities. They may include physical fitness, maintaining one’s health, spiritual commitments, hobbies, craft work, reading, and study. It is proposed that there is a positive flow-on effect if individuals take care of themselves and participate in activities they enjoy because this allows for restoration of personal resources and recovery from work. However, it is also interesting to know whether the time demands these activities require induce conflict with the work domain. The scarcity theory (see Marks and MacDermid, 1996) suggests that there would be conflict due to limited time resources available. Facilitation between domains Individuals place values on roles which determine their level of importance (Clark, 2000). As argued by the scarcity theory, active participation in nonwork domains such as family, community and recreation has been viewed historically as reducing the time available for work, as well as individuals’ feelings of commitment to their job (Marks and MacDermid, 1996). On the other hand, some research has identified involvement in multiple roles as protecting or buffering individuals from the effects of negative experiences in any one role (Barnett and Hyde, 2001). Rather than draining energy from work and reducing performance, workers’ commitments in other domains may provide multiple opportunities for satisfaction and resource gaining and may energize them for work (Ruderman et al., 2002). This “expansion” model considers personal resources to be abundant and expandable (Crouter, 1984) and provides the theoretical basis for facilitation (Hammer et al., 2005). From this perspective, high levels of participation in nonwork domains may be associated with an individual experiencing facilitation to work (Kirchmeyer, 1995). Psychological involvement and investment of time are two variables that can be used to reflect a person’s level of commitment to and participation in a domain (O’Driscoll, 1996). Some research suggests that work-family facilitation (which is one manifestation of positive spillover) may be a crucial component of work-family balance (Frone, 2003) and balance may be improved by increasing facilitation levels. Facilitation represents the extent to which participation in one domain is made easier because of the experiences, skills and resources gained or developed in a different domain. Balance would be demonstrated by not only low conflict, but also high facilitation levels between the different domains (Frone, 2003; Grzywacz and Bass, 2003; Keene and Quadagno, 2004). For instance, Grzywacz and Bass (2003) found that most positive outcomes accrued almost exclusively from low levels of work-family conflict and high levels of work-family facilitation. Facilitation rests on the expansion theory (Hammer et al., 2005), along with Seiber’s (1974) classification of the positive outcomes of role accumulation into four types: (1) role privileges; (2) status security;
Non-work to work facilitation
275
JMP 23,3
276
(3) status enhancement; and (4) enrichment of personality. Seiber noted that certain role privileges or rights are institutionalised in each role, for example, the buying privileges of a salesman or the free game tickets for a chief executive. Therefore the greater number of roles accumulated the greater number of privileges that can be enjoyed. Status security stems from the idea that role strain in one domain can be buffered by participation in other roles. Status enhancement occurs when the by-products of one role, such as personal contacts, are utilized in other roles. Seiber’s fourth type of outcome, which is highly salient in the context of the present study, is personal enrichment or development. Skills and attitudes acquired in one domain can be valuable in a different domain. Kirchmeyer (1992) argued that Seiber’s classification accounts for the positive spillover experiences identified in her study, and Greenhaus and Powell (2006) suggested that this classification captures the concept of work-family enrichment. Hypotheses Nonwork to work spillover The enhancement hypothesis discussed earlier proposes facilitation between domains, that is, the extent to which participation in one domain is made easier because of the experiences, skills and resources developed in a different domain. Facilitation is a form of synergy in which the resources in one role enhance or make easier the participation in another role and may occur via mood, values, skills and behaviors, as suggested by a positive correlation between family and work (Frone, 2003; Voydanoff, 2004; Kirchmeyer, 1992; Grzywacz and Marks, 2000). While facilitation can be bi-directional, the current research only investigated the outcomes of facilitation from nonwork (family and personal benefit activities) to the work domain. This decision was made because the focus of this research was on the potential implications of nonwork experiences for work. As previously mentioned, research to date has focused predominantly on the impact of work on nonwork experiences (usually family). This paper explores the following routes for each of the nonwork areas being investigated (family life and personal benefit activities) and the work domain: . the relationship between facilitation and psychological involvement; . the relationship between psychological involvement and positive well-being; . the relationship between time demands and inter-domain conflict; and . whether facilitation plays a mediating role between domain involvement and positive well-being (see Figure 1). Each of these routes is discussed in turn below. Psychological involvement Psychological involvement, also referred to as “commitment” or “investment”, is the level of engagement an individual in the domain (Greenhaus et al., 2003; O’Driscoll, 1996). It is thought that increasing levels of psychological involvement can lead to a preoccupation with the responsibilities and demands of a particular role. Organizations often assume that commitment to family compromises work performance (Kofodimos,
Non-work to work facilitation
277
Figure 1. Research model
1995), but some research suggests that this same commitment can enhance work performance (Ruderman et al., 2002). Kirchmeyer (1992) assessed the impact of nonwork experiences on the work domain and found that high nonwork involvement was associated more with facilitation than with conflict between domains. Consistent with Kirchmeyer’s findings, a more recent study by Graves et al. (2007) found that high family commitment levels were associated with facilitation rather than conflict with the work domain. In a nonwork domain with high levels of psychological involvement, the benefits, rewards and pleasures of that domain should become more obvious, enhancing other realms of life such as work. To assess if high psychological involvement in a nonwork domain produces facilitation to the work domain, the following hypotheses were examined: H1a. There will be a positive relationship between psychological involvement in family life and family to work facilitation. H1b. There will be a positive relationship between psychological involvement in personal benefit activities and facilitation from the personal benefit activities domain to the work domain. Time demands Voydanoff (2004) among others has found that time-based work demands show a positive relationship with work-to-family conflict. The present research investigated whether time demands in the two nonwork areas (family and personal benefit activities) can also induce nonwork-to-work conflict. Research on work-family dynamics illustrates that the demands of work and family roles compete for the individual’s time and energy (Frone et al., 1992; Voydanoff, 2004). Similarly, we predicted that greater time spent in personal benefit activities will also be associated with nonwork-to-work conflict. Consequently we hypothesised that: H2a. There will be a positive relationship between the number of hours spent in the family life domain and levels of family to work conflict. H2b. There will be a positive relationship between the number of hours spent in the personal benefit activities domain and levels of personal benefit activities to work conflict.
JMP 23,3
278
Well-being Well-being is a broad concept that includes a variety of affects and aspects of satisfaction and mental health (Sonnentag, 2001). Warr’s (1992) measure places different feelings on two principal axes: (1) anxiety-contentment; and (2) depression-enthusiasm. These two axes correspond with positive and negative well-being. Given the thrust of the present research, in particular our primary interest in facilitation between nonwork and work domains, in this study we focused solely on positive well-being. Following enhancement theory, being engaged in multiple roles is generally thought to promote well-being and is synergistic for an individual (Ruderman et al., 2002). Satisfactory role engagement between domains is expected to be associated positively with individuals’ well-being because it can reduce inter-domain conflict and stress, both of which detract from well-being (Greenhaus et al., 2003). Conflict and stress should be minimal in the presence of facilitation, resulting in higher well-being levels. These cross-domain effects are supported by Ford et al. (2007) in their meta-analysis, which reported that reduced stressors and sources of support that are specific to one domain were positively related to satisfaction in another domain. We therefore predicted that if there is facilitation from a nonwork domain to work, individuals will also score high on positive well-being in all three domains (family, work and personal benefit activities) investigated here: H3a. There will be a positive correlation between family-to-work facilitation and positive family well-being. H3b. There will be a positive correlation between family-to-work facilitation and positive work well-being. H3c. There will be a positive correlation between family-to-work facilitation and positive personal benefit activities well-being. H4a. There will be a positive correlation between personal activities-to-work facilitation and positive family well-being. H4b. There will be a positive correlation between personal activities-to-work facilitation and positive work well-being. H4c. There will be a positive correlation between personal activities-to-work facilitation and positive personal benefit activities well-being. Positive events are associated with increased well-being, especially if the individual enjoys the activity and puts a high value rating on the activity, possibly due to the influence it has on mood (Eden, 2001; Haworth, 1997; Sonnentag, 2001). The personal benefit activities domain encompasses events that an individual enjoys participating in and should have high psychological involvement. We predicted that individuals who score high in psychological involvement in personal benefit activities would also score high on well-being in that domain:
H5a. Individuals who score high on psychological involvement in the personal benefit activities domain will also score high on positive well-being within the personal benefit activities domain. We also propose that the positive aspects of engagement in the personal benefit activities domain will have a positive consequence for the individual at work. Experiences outside of the workplace contribute to the value an individual brings to the workplace and may lead to enhancement for individuals’ well-being and organizational productivity (Kofodimos, 1995; O’Driscoll, 1996). We predicted that individuals with high psychological involvement in the personal benefit activities domain would correspondingly score highly on positive work well-being, due to the experiences an individual brings to work when they engage in personal benefit activities: H5b. Individuals who score high on psychological involvement in the personal benefit activities domain will also score high on positive well-being for the work domain. Finally, expansion theory suggests that synergies which prevail between personal benefit activities and family life should generate facilitation effects between these two domains, and hence that high commitment to personal benefit activities would be linked with positive well-being in the family domain: H5c. Individuals who score high on psychological involvement for the personal benefit activities domain will also score highly on well-being for the family domain. Mediated relationships In addition to the above relationships, we also propose that facilitation (between nonwork and work domains) can function as a mediator of relationships between psychological involvement and well-being. That is, when the skills, experiences and resources gained in a nonwork domain are utilised in the workplace, well-being may be significantly improved. Following from the above hypotheses, it is predicted that high domain involvement will be associated with higher levels of facilitation, which in turn will be related to high well-being. The following hypotheses were investigated: H6a. Family-to-work facilitation will mediate the relationship between family involvement and positive family well-being. H6b. Family-to-work facilitation will mediate the relationship between family involvement and positive work well-being. H6c. Family-to-work facilitation will mediate the relationship between family involvement and positive personal benefit activities well-being. H7a. Personal activities-to-work facilitation will mediate the relationship between personal benefit activities involvement and positive family well-being. H7b. Personal activities-to-work facilitation will mediate the relationship between personal benefit activities involvement and positive work well-being. H7c. Personal activities-to-work facilitation will mediate the relationship between personal benefit activities involvement and positive personal benefit activities well-being.
Non-work to work facilitation
279
JMP 23,3
280
Method Sample and procedure Participants included 938 New Zealand local government employees working in a variety of jobs and positions of responsibilities in 86 different local government workplaces. The average age was 43 years and 54 percent female. Participants had been in their current employment for an average of seven years; 13 percent were in senior management positions. Seventy-two percent of the participants had children. The data for this study were collected via an online questionnaire, which was distributed to over 3,000 employees, and had an estimated response rate of 31 percent (the estimation is due to the unknown number of employees with internet access). Measures Survey respondents were asked to reflect on activities related to their family life and the activities they participate in that are part of their self-care and of benefit to themselves. Two forms of facilitation were separately assessed: (1) family-to-work; and (2) personal activities-to-work. Interrole facilitation was assessed using Kirchmeyer’s (1992) measure. Fifteen positive statements (for each form of facilitation) were presented for participants to indicate the extent of agreement with each statement using a six-point Likert scale. Sample items included “being involved with family life activities helps me to forget the problems at work” and “being involved in personal benefit activities provides me with contacts that are helpful at work”. Response options for each of these items ranged from 1 (strongly disagree) to 6 (strongly agree). A total score for each direction of facilitation was obtained for each individual by averaging their item scores (Cronbach’s a ¼ 0:90 for family-to-work, a ¼ 0:91 for personal activities-to-work). Negative spillover (interrole conflict) was assessed separately for family to work and personal benefit activities to work utilising Kirchmeyer’s (1992) measure. The participants were required to respond to eight statements on a six-point Likert scale as indicated above. Sample items included “being involved with family life tires me out that I feel drained at work” and “being involved in personal benefit activities demands time from me that could be spent at work”. A total score for each direction of conflict was obtained for each participant by averaging their item scores (Cronbach’s a ¼ 0:88 for family-to-work, a ¼ 0:89 for personal activities-to-work). Domain psychological involvement in the three areas was assessed utilising Kanungo’s (1982) nine-item measure. Respondents were asked to indicate the extent of their agreement with each statement for work, family and personal benefit activities domains on a seven-point Likert scale with anchors 1 ¼ strongly disagree, 7 ¼ strongly agree. Sample questions included “I consider my job to be very central to my existence” and “many of my personal life goals are family orientated”. Responses to three negatively worded items were reverse-scored. In each of the three domains, item scores were averaged to derive an overall work, family and personal benefit activities involvement score for each person (Cronbach’s a ¼ 0:79 for work involvement, a ¼ 0:89 for family involvement, and a ¼ 0:89 for personal benefit activities involvement). Time commitment was measured by asking the participant to report the number of hours they would typically spend per week on the activities in each of the three
domains, i.e. work domain, family domain and personal benefit activities domain. This allowed for a great deal of subjective assessment as the respondent was able to choose which activities belonged in what domain, along with estimating the number hours. The average time spent was 37.03 hours in the family domain, 13.22 hours in the personal benefit activities domain, and 43.58 hours in the work domain. Well-being was assessed via Warr’s (1990) measure of well-being. Respondents were asked to respond to 15 adjectives on a six point Likert scale for each domain, with response categories ranging from 1 ¼ strongly disagree to 6 ¼ strongly agree, indicating whether they had experienced a particular feeling in the last three months in each of the three domains. Sample items included tense, contented, optimistic and anxious. Principal components factor analysis resulted in two factors, corresponding with positive and negative well-being. However, as the emphasis of this exploratory study was on positive attributes, the negative well-being data were omitted from further analysis. Responses to the eight positive well-being items were averaged to derive an overall positive work well-being, positive family well-being and positive personal benefit activities well-being score for each person (Cronbach’s a ¼ 0:92 for positive work well-being, 0.96 for positive family well-being, and 0.95 for positive personal benefit activities well-being).
Non-work to work facilitation
281
Results Descriptive statistics for all variables are presented in Table I. The mean scores indicate that, on average, moderate levels of family-to-work facilitation (M ¼ 4:12) and personal activities-to-work facilitation (M ¼ 4:11) were reported by respondents. Means of the nonwork-to-work conflict measures were lower than those for nonwork-to-work facilitation, suggesting that on average interference between nonwork domains and work was relatively low. The means for psychological involvement in the nonwork domains were moderate, signifying a moderate level of involvement in these domains (family involvement M ¼ 5:65 and personal benefit activities involvement M ¼ 4:49). Well-being means were M ¼ 3:56 for positive work well-being, M ¼ 4:19 for positive family well-being and M ¼ 4:24 for positive personal Variable a
Family-work facilitation Family-work conflicta Personal benefit activities-work facilitationa Personal benefit activities-work conflicta Family involvementb Personal benefit activities involvementb Family hours Work hours Personal benefit activities hours Positive work well-beinga Positive family well-beinga Positive personal benefit activities well-beinga
n
M
SD
Skew
a
975 975 969 967 938 934 912 937 907 939 937 931
4.11 2.19 4.12 1.72 5.65 4.49 37.03 43.58 13.22 3.56 4.19 4.24
0.84 0.92 0.86 0.76 1.03 1.19 25.45 8.62 9.81 0.93 0.95 0.97
20.61 * 0.81 * 20.62 * 1.63 * 21.17 * 20.20 * 1.51 * 0.77 * 2.34 * 0.01 * 20.57 * 0.54 *
0.90 0.88 0.91 0.89 0.89 0.89
0.92 0.96 0.95
Notes: aResponse ratings were made on a six-point scale; bresponse ratings were made on a seven-point scale; *p , 0:05
Table I. Descriptive statistics for all variables
JMP 23,3
282
benefit activities well-being, suggesting that participants generally experienced positive well-being in all three domains. Correlations Due to the sample size (. 900), the Bonferroni correction (Collis and Rosenblood, 1985) was applied to reduce the likelihood of correlations being significant due to chance. Correlations between all variables are displayed in Table II. H1a and H1b predicted high psychological involvement in a nonwork domain would be related to facilitation to the work domain. The correlations in Table II indicate a positive relationship between family psychological involvement and family to work facilitation (r ¼ 0:37, p , 0:05) as well as between psychological involvement in personal benefit activities and personal activities-to-work facilitation (r ¼ 0:27, p , 0:05). H1a and H1b were therefore supported. H2a and H2b considered the relationship between time demands in the nonwork domains and the conflict between those domains and work. Analysis for both these hypotheses produced non-significant results, suggesting that time demands from the family or personal benefit activities did not relate significantly with family-to-work conflict (r ¼ 0:04) or personal benefit activities-to-work conflict (r ¼ 0:03). H3a, H3b, H3c, H4a, H4b and H4c predicted that if there is facilitation from a nonwork domains to the work domain, individuals would also score high on positive well-being in each of the domains. Correlations between these variables were all significant, supporting a positive relationship between facilitation and the well-being variables. H5a, H5b and H5c investigated the relationship between involvement in personal benefit activities and positive well-being in the three domains studied. H5a was supported, with a significant correlation of 0.35, indicating that psychological involvement in personal benefit activities did have a positive relationship with positive well-being in this same domain. H5b was not supported (r ¼ 0:04), signifying that psychological involvement in personal benefit activities was not related to positive work well-being. Although the correlation between psychological involvement in personal benefit activities and family well-being was low (r ¼ 0:11), it was significant (p , 0:05), and therefore H5c was supported (see Figure 2). Mediation effects The Baron and Kenny (1986) mediated regression procedure was utilized to examine the mediational effects of family-to-work and personal activities-to-work facilitation. This procedure contains three phases: (1) regression of the mediator on the predictor variable; (2) regression of the criterion on the predictor variable; and (3) regression of the criterion variable on the predictor and mediator variables simultaneously. Baron and Kenny (1986) proposed that mediation is indicated when the following conditions are met: . there is a significant relationship between the predictor and mediator variables (step 1);
Variable 1 2 3 4 5 6 7 8 9 10 11 12
Family hours Personal hours Work hours F-W conflict F-W facilitation PB-W facilitation PB-W conflict PB involvement Family involvement Positive WWB Positive FWB Positive PBWB
1
2
3
4
5
6
7
8
9
10
11
20.00 0.10 * * 0.04 0.09 * * 20.04 20.07 * * 20.16 * 0.26 * * 0.02 0.18 * * 0.01
20.01 0.04 20.01 0.1 0.03 20.01 20.00 20.02 20.01 20.01
0.05 20.03 20.04 0.12 * * 0.7 * 0.08 * * 0.1 0.6 20.04
20.08 * * 0.03 0.60 * * 0.12 20.17 * * 20.23 * * 20.48 * * 20.24 * *
0.62 * * 20.03 0.12 0.37 * * 0.15 * * 0.37 * * 0.16 * *
0.4 0.27 * * 0.08 * * 0.16 * * 0.16 * * 0.22 * *
0.02 20.15 * * 20.18 * * 20.31 * * 20.29 * *
20.08 * * 0.04 0.11 * 0.35 * *
0.08 0.50 * * 0.13 * *
0.44 * * 0.43 * *
2 0.60 * *
Notes: *p , 0:05; * *p , 0:01
Non-work to work facilitation
283
Table II. Correlations between variables
JMP 23,3
284
Figure 2. Diagram to illustrate H1a-H5c
.
. .
there is a significant relationship between the predictor and criterion variables (step 2); the mediator is significantly related to the criterion variable (step 3); and the relationship of the predictor with the criterion variable is less at step 3 than at step 2.
Full mediation occurs when the predictor variable has no significant relationship with the criterion variable when the predictor and mediator are entered into the equation simultaneously at step 3. Partial mediation is indicated when the predictor-criterion association is reduced in magnitude, but remains significant when the predictor and mediator are entered simultaneously. The Sobel test (McKinnon et al., 2002; Sobel, 1983) was used to assess the statistical significance of mediated relationships. Table III illustrates that family-to-work facilitation functioned as a mediator between family involvement and the positive well-being measures, which supports H6a, H6b, and H6c. Greater family involvement was associated with higher levels of family-to-work facilitation, which in turn was linked to more positive family well-being, work well-being and personal benefit activities well-being. The Sobel test confirmed that each of these mediation effects was significant. Similarly, Table IV indicates that personal activities-to-work facilitation was a mediator between involvement in personal benefit activities and positive well-being in that domain. When facilitation occurred between the personal benefit activities domain and the work domain, levels of positive family well-being and personal benefit activities well-being were significantly higher. Again the Sobel test indicated that these mediations were significant, confirming H7a and H7c. However, because involvement
Hypothesis 6a
6b
6c
Equation 1 2 3
1 2 3
1 2 3
b
R 2 change
Family involvement Family involvement Family involvement F-W facilitation Sobel test
0.37 * * 0.50 * * 0.40 * 0.18 * z ¼ 6:52
0.14 0.25 0.27
Family involvement Family involvement Family involvement F-W facilitation Sobel test
0.37 * * 0.08 *
Criterion
Predictor
F-W facilitation Positive FWB Positive FWB
F-W facilitation Positive WWB Positive WWB
F-W facilitation Positive PBWB Positive PBWB
Family Involvement Family Involvement Family Involvement F-W facilitation Sobel test
0.03 0.13 * z ¼ 3:63 0.37 * * 0.13 * 0.08 * 0.13 * z ¼ 3:59
p ¼ 0:00
p ¼ 0:00 0.14 0.12 0.03 p ¼ 0:00
PBWB, personal benefit activities well-being
7a
7b
7c
Equation 1 2 3
b
R 2 change
PB involvement PB involvement PB involvement PB-W facilitation Sobel test
0.27 * * 0.35 * * 0.31 * 0.13 * z ¼ 3:65
0.07 0.12 0.14
Criterion
Predictor
PB-W facilitation Positive PBWB Positive PBWB
1 2 3
PB-W facilitation Positive WWB Positive WWB
PB involvement PB involvement PB involvement PB-W facilitation
0.27 * *
1 2 3
PB-W facilitation Positive FWB Positive FWB
PB involvement PB involvement PB involvement PB-W facilitation Sobel test
0.27 * * 0.11 * 0.06 0.16 * z ¼ 2:80
0.04 0.00 0.16 *
285
0.14 0.01 0.02
Notes: *p , 0:05; * *p , 0:01. F-W, family to work; FWB, family well-being; WWB, work well-being;
Hypothesis
Non-work to work facilitation
Table III. Regression analysis of facilitation mediating the relationship between family involvement and well-being
p ¼ 0:00 0.07 0.00 0.03 0.07 0.01 0.04 p ¼ 0:01
Notes: *p , 0:05; * *p , 0:01. F-W, family to work; FWB, family well-being; WWB, work well-being; PBWB, personal benefit activities well-being; PB, personal benefit activities; PB-W, personal benefit activities to work
in personal benefit activities and work well-being were not significantly interrelated, the mediation effect proposed in H7b was not supported. In summary, facilitation was found to act as a mediator between psychological involvement in personal benefit activities and family and the three forms of positive well-being, supporting H6a, H6b, H6c, H7a and H7c. Involvement (in personal benefit activities and family) enhanced individuals’ experiences, not only in the specific domain, but also in other life domains, resulting in elevated levels of well-being.
Table IV. Regression analysis of facilitation mediating the relationship between personal benefit activities involvement and well-being
JMP 23,3
286
Discussion A major aim of the present research was to explore the relationship between experiences in two nonwork domains and the work domain, firstly family life to work, and secondly personal benefit activities to work. As noted earlier, the changing nature of individuals’ lives makes it important to identify factors that contribute to well-being. Of particular interest in this context was the relationship between personal benefit activities and well-being. While the results involving the family life domain concur with the limited extant research (e.g. Graves et al., 2007, Kirchmeyer, 1992), the present study explored new territory by also investigating the relationship between personal benefit activities and work-related well-being. Psychological involvement We found that high psychological involvement in a nonwork domain correlated positively with facilitation to the work domain (as predicted by H1a and H1b), illustrating that high involvement levels can have implications for nonwork-to-work facilitation. This should not be surprising, as the rewards, benefits and pleasures that can accompany high involvement can lead to facilitation to another domain (Kirchmeyer, 1992). Our findings correspond with those reported by Kirchmeyer (1992) and Grzywacz and Marks (2000), whose analysis showed family factors (such as parental status, spouses support and disagreement) are the primary sources of family-to-work facilitation. However, this study ventured into a new domain – personal benefit activities. When experiences promote enhanced functioning (such as putting to use new skills) in another domain, facilitation occurs (Grzywacz, 2000). For example, when a committed surf lifesaver (high psychological involvement) transfers her ability in handling stressful situations in her nonwork domain into better handling stress in her workplace, spillover is evident. An important contribution of this study is the demonstration that psychological involvement in two nonwork domains (family and personal benefit activities) is associated with positive facilitation to work. The importance of involvement in personal benefit activities is further illustrated when the results from H5a, H5b and H5c are considered. H5a and H5c were supported but not H5b, indicating that involvement in personal benefit activities has a positive relationship with positive personal benefit activities well-being and positive family life well-being, but not positive work well-being. The role of the border an individual may have placed around a domain (as discussed below) may influence these relationships. A positive correlation supporting H5a is not surprising. One would expect well-being in relation to positive personal benefit activities to be enhanced by high involvement in this domain, as these activities that the individual enjoys. They may be challenging or relaxing, but the literature on leisure extols its virtues for well-being (e.g. Eden, 2001; Haworth, 1997). As many personal benefit activities could also be classified as leisure activities, the present study also demonstrates that there are benefits of these for well-being. H5c predicted that individuals who have high psychological involvement in personal benefit activities would also have high positive family well-being. This hypothesis was supported, which confirms the expansion theory of synergies between multiple roles (Marks, 1977), but also questions the nature of the borders individuals may have placed around a domain. Family life and personal benefit activities may be
very integrated, allowing for permeability, flexibility and blending between the domains (Clark, 2000). Many activities may be located in both domains simultaneously. On the other hand, H5b was not supported; involvement in personal benefit activities did not significantly relate with positive work well-being. This appears to be counter-intuitive, as it might be expected that if an individual actively engages in personal benefit activities that they should also enjoy positive work well-being. The present lack of support for this hypothesis may be due to the borders between personal benefit activities and work being relatively impermeable. Where family life and personal benefit activities may have integrated, blended borders as discussed above, the personal benefit activities domain and work domain may be more segmented, resulting in less opportunity for spillover of positive experiences across these two domains. The present findings suggest that, generally speaking, individuals may maintain a relatively impermeable boundary between their work life and their “personal life”. Time demands H2a and H2b predicted a positive relationship between the amount of time spent in nonwork domains (family and personal activities) and conflict with work. However, these hypotheses were not supported. As in some previous studies (e.g. O’Driscoll et al., 1992), time demands from personal benefit activities and family commitments did not appear to induce conflict with work responsibilities Although time is not expandable, the scarcity theory of time in one domain robbing time in another domain is not supported by the present research. Active participation in family life and personal benefit activities may actually energize people for work and not create conflict flowing to the work domain (Kirchmeyer, 1992). Time and involvement in family relationships, while time-consuming, can create other benefits and synergies such as satisfaction and psychological energy, which carry over to the work environment (Grzywacz, 2000; Marks, 1977). It may not be the amount of time spent in a domain, but whether the time spent in the domain drains or energises the individual (Marks, 1977; Voydanoff, 2004). Conversely, the relationship between work time demands and work-to-family conflict has been frequently documented (e.g. Voydanoff, 2004). However, it has been observed that there is an asymmetry in the impact of time demands, indicating that the interference of the work with family commitments and responsibilities is not the same as interference in the other direction (Frone, 2003). Both the scarcity and enhancement theories may work simultaneously, with the scarcity theory operating in the work-to-nonwork relationship and the enhancement theory operating in the nonwork-to-work relationship. The effects of time demands may very much depend on the domain in which these demands originate. While nonwork can revitalise and energise an individual (Marks, 1977; Seiber, 1974), for many people work may be viewed more as a drain on time and resources, creating work-to-nonwork conflict (Voydanoff, 2004). Facilitation Our research supported hypotheses concerning the relationship between facilitation from nonwork domains to the work domain and positive well-being (H3a, H3b, H3c and H4a, H4b, H4c). This is not surprising, given that facilitation encompasses how the nonwork domain supports and enhances well-being (Kirchmeyer, 1992). A “ripple”
Non-work to work facilitation
287
JMP 23,3
288
effect can occur, with positive events having a beneficial impact on well-being (Eden, 2001) through the transmission processes in which nonwork domains are associated with psychological responses, which are transmitted into attitudes and behaviors at work (Voydanoff, 2004). Increases in facilitation have been related to favourable physical and psychological health outcomes, improved health and well-being outcomes (Grzywacz, 2000) and less depression (Hammer et al., 2005). Recent results have indicated that there are protective effects from facilitation, and that family-to-work facilitation is associated with better psychological well-being, independent of the effects of interrole conflict (Grzywacz and Bass, 2003). This study extends these positive effects to the personal benefit activities domain, illustrating that involvement in both family and personal benefit activities may contribute to higher well-being at work, as well as in life off the job. Well-being Positive well-being incorporates affects and aspects of mental health and satisfaction. These include feelings such as enthusiasm, contentedness and motivation. Participating in nonwork activities that encourage rest and reflection and promote enjoyment may replenish depleted physical and emotional resources, and thus improve well-being (Eden, 2001; Haworth, 1997). Although well-being and work-life balance are not conceptually or empirically identical, they may overlap and operate concurrently. Facilitation has been shown in this study to have a positive relationship with positive well-being and has previously been identified as an integral part of work-life balance, with increasing levels of facilitation improving balance (Frone, 2003; Grzywacz and Bass, 2003; Keene and Quadagno, 2004). Consistent with Frone (2003), who suggested that work-life balance is a combination of low conflict and high facilitation, an increase in positive well-being levels may also indicate greater facilitation of work-life balance. Mediated relationships One of the primary aims of this research was to investigate if facilitation acts as a mediator between psychological involvement and well-being. Individuals who are highly committed to a nonwork domain may gain skills, experiences and resources that can raise their positive well-being levels in the work, family and personal benefit activities domains. These positive well-being levels are significantly higher when facilitation acts as a mediator. The implications that arise include finding ways to advance facilitation in order to raise well-being levels. This is an important consideration for both individuals and organizations. Individuals need to develop strategies to enhance their facilitation levels and subsequently raise their positive well-being levels across different domains. In raising well-being levels, individuals were more likely to have elevated levels of mental health and satisfaction. These results are of value, as improving the quality of life, which includes raising well-being, is an emerging area of positive psychology (Seligman and Csikszentmihalyi, 2000). Future research One limitation of research in this field, referred to previously, is the lack of a clear conceptualization of the expression work-life “balance”. To progress our understanding of the role of this construct in promoting overall life satisfaction and
psychological well-being, research needs to investigate both the meaning of “balance” for individuals and its association with other (related) constructs, such as interrole facilitation and conflict. In addition, further development of appropriate measures of work-life balance is required (see also Greenhaus et al., 2003). More research investigating inter-relationships between the nonwork domains could also be enlightening, as the perceived lack of balance may be due to conflict between nonwork domains as well between these domains and work. However, typically workplaces are identified as the principal cause of imbalance. Society labels the issue as “work-life” balance, resulting in a focus on the “work” portion of life. A methodological limitation in the present study is that we did not assess the role of choice and control over work and nonwork activities and demands, which clearly may influence individuals’ affective experiences in these areas of life. Another potential limitation is that respondents were given freedom to choose exactly which activities they included within, for example, the family domain and the personal activities domain, which may have created considerable variability in time allocations, depending on how activities were classified. On the other hand, this approach did enable respondents to identify family and personal benefit activities in terms of their own schema, rather than one imposed by the researchers. A final limitation is participants in this research were all local government employees, which means that the findings cannot necessarily be generalized to other industry sectors and types of employment. However, we have no reason to believe that the current respondents’ perceptions and attitudes concerning work, family and personal benefit activities would be substantially different from those held by other workers. Practical implications The results emerging from the present study have several practical implications. If organizations embrace these results and findings from other related research, both the employee and the organization may benefit. Employers should encourage employees to enjoy rich nonwork activities, as the benefits of engagement in these off-the-job activities may spill over into the work context. There are good reasons to encourage facilitation – improved health and well-being and less depression. Increased attention needs to be given to how facilitation can be developed and cultivated. Finally, our findings should also encourage individuals to pursue nonwork activities that are important to them and which can enhance their lives. Not only will they benefit personally from these activities, but facilitation may create positive well-being at work and in their family life. References Barnett, R.C. and Hyde, J.S. (2001), “Women, men, work, and family: an expansionist theory”, American Psychologist, Vol. 56 No. 10, pp. 781-96. Baron, R.M. and Kenny, D.A. (1986), “The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations”, Journal of Personality and Social Psychology, Vol. 51, pp. 1173-82. Clark, S.C. (2000), “Work/family border theory: a new theory of work/family balance”, Human Relations, Vol. 53 No. 6, pp. 747-70. Collins Compact Thesaurus (1999), Collins Compact Thesaurus, HarperCollins, London.
Non-work to work facilitation
289
JMP 23,3
290
Collis, B.A. and Rosenblood, L.K. (1985), “The problem of inflated significance when testing individual correlations from a correlation matrix”, Journal of Research Mathematics Education, Vol. 16 No. 1, pp. 52-5. Crouter, A.C. (1984), “Spillover from family to work: the neglected side of the work-family interface”, Human Relations, Vol. 37, pp. 425-42. Eby, L.T., Casper, W.J., Lockwood, A., Bordeaux, C. and Brindley, A. (2005), “Work and family research in IO/OB: content analysis and review of the literature (1980-2002)”, Journal of Vocational Behavior, Vol. 66, pp. 124-97. Eden, D. (2001), “Vacation and other respites: studying stress on and off the job”, in Cooper, C.L. and Robertson, I.T. (Eds), Well-being in Organizations, Wiley, Chichester, pp. 305-30. Ford, M.T., Heinen, B.A. and Langkamer, K.L. (2007), “Work and family satisfaction and conflict: a meta-analysis of cross domain relations”, Journal of Applied Psychology, Vol. 92 No. 1, pp. 57-80. Frone, M.R. (2003), “Work-family balance”, in Quick, J.C. and Tetrick, L.E. (Eds), Handbook of Occupational Health, American Psychological Association, Washington, DC, pp. 143-62. Frone, M.R., Russell, M. and Cooper, M.L. (1992), “Antecedents and outcomes of work-family conflict: testing a model of the work-family interface”, Journal of Applied Psychology, Vol. 77, pp. 65-78. Glazer, S. and Beehr, T.A. (2005), “Consistency of implications of three role stressors across four countries”, Journal of Organizational Behavior, Vol. 26 No. 5, pp. 467-87. Greenhaus, J.H., Collins, K.M. and Shaw, J.D. (2003), “The relation between work-family balance and quality of life”, Journal of Vocational Behavior, Vol. 63 No. 3, pp. 510-31. Grzywacz, J.G. (2000), “Toward a theory of work-family facilitation”, paper presented at the 32nd Annual Theory Construction and Research Methodology Workshop, Houston, TX. Grzywacz, J.G. and Bass, B.L. (2003), “Work, family, and mental health: testing different models of work-family fit”, Journal of Health and Marriage, Vol. 65 No. 1, pp. 248-62. Grzywacz, J.G. and Marks, N.F. (2000), “Reconceptualizing the work-family interface: an ecological perspective on the correlates of positive and negative spillover between work and family”, Journal of Occupational Health Psychology, Vol. 5, pp. 111-26. Hammer, L.B., Cullen, J.C., Neal, M.B., Sinclair, R.R. and Shafiro, M.V. (2005), “The longitudinal effects of work-family conflict and positive spillover on depressive symptoms among dual-earner couples”, Journal of Occupational Health Psychology, Vol. 10 No. 2, pp. 138-54. Haworth, J.T. (1997), Work, Leisure and Well-being, Routledge, London. Keene, J.R. and Quadagno, J. (2004), “Predictors of perceived work-family balance: gender difference or gender similarity?”, Sociological Perspectives, Vol. 47 No. 1, pp. 1-23. Kirchmeyer, C. (1992), “Perceptions of nonwork-to-work spillover: challenging the common view of conflict-ridden domain relationships”, Basic and Applied Psychology, Vol. 13 No. 2, pp. 231-49. Kirchmeyer, C. (1995), “Managing the work-nonwork boundary: an assessment of organizational boundaries”, Human Relations, Vol. 48 No. 5, pp. 515-37. Kofodimos, J.R. (1995), Beyond Work-family Programs. Confronting and Resolving the Underlying Causes of Work-personal Life Conflict, Center for Creative Leadership, Greensboro, NC. McKinnon, D.P., Lockwood, C.M., Hoffman, J.M., West, S.G. and Sheets, V. (2002), “A comparison of models to test mediation and other intervening variable effects”, Psychological Methods, Vol. 7, pp. 83-104.
Marks, S.R. (1977), “Multiple roles and role strain: some notes on human energy, time and commitment”, American Sociological Review, Vol. 42, pp. 921-36. Marks, S.R. and MacDermid, S.M. (1996), “Multiple roles and the self: a theory of role balance”, Journal of Marriage & Family, Vol. 58 No. 2. O’Driscoll, M., Iglen, D.R. and Hildreth, K. (1992), “Time devoted to job and off-job activities, interrole conflict, and affective experiences”, Journal of Applied Psychology, Vol. 77 No. 3, pp. 272-9. O’Driscoll, M.P. (1996), “The interface between job and off-job roles: enhancement and conflict”, in Cooper, C.L. and Robertson, I.T. (Eds), Well-being in Organizations: A Reader for Students and Practitioners, Wiley, Chichester, pp. 149-76. Ruderman, M.N., Ohlott, P.J., Panzer, K. and King, S.N. (2002), “Benefits of multiple roles for managerial women”, Academy of Management Journal, Vol. 45 No. 2, pp. 369-86. Seiber, S.D. (1974), “Towards a theory of role accumulation”, American Sociological Review, Vol. 39, pp. 567-78. Seligman, M.E.P. and Csikszentmihalyi, M. (2000), “Positive psychology: an introduction”, American Psychologist, Vol. 55, pp. 5-14. Sobel, M.E. (1983), “Asymptotic confidence intervals for indirect effects in structural equation models”, in Leinhardt, S. (Ed.), Sociological Methodology, American Sociological Association, Washington, DC, pp. 290-312. Sonnentag, S. (2001), “Work, recovery activities, and individual well-being: a diary study”, Journal of Occupational Health Psychology, Vol. 6 No. 3, pp. 196-210. Voydanoff, P. (2004), “The effects of work demands and resources on work-to-family conflict and facilitation”, Journal of Marriage and Family, Vol. 66 No. 2, pp. 398-412. Further reading Burke, R.J. (2004), “Work and personal life integration”, International Journal of Stress Management, Vol. 11 No. 4, pp. 299-304. Mackavey, M.G. and Levin, R.J. (1998), Shared Purpose. Working Together to Build Strong Families and High Performance Companies, AMACOM, New York, NY. Corresponding author Michael O’Driscoll can be contacted at:
[email protected]
To purchase reprints of this article please e-mail:
[email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints
Non-work to work facilitation
291
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0268-3946.htm
JMP 23,3
The influence of family responsibilities, career fields and gender on career success
292
An empirical study
Received April 2007 Revised December 2007 Accepted December 2007
Wolfgang Mayrhofer Interdisciplinary Unit of Management and Organisational Behaviour, Wirtschaftsuniversita¨t Wien, Vienna, Austria
Michael Meyer Unit of Nonprofit Management, Wirtschaftsuniversita¨t Wien, Vienna, Austria
Michael Schiffinger Interdisciplinary Unit of Management and Organisational Behaviour, Wirtschaftsuniversita¨t Wien, Vienna, Austria, and
Angelika Schmidt Institute of Change Management and Management Development, Wirtschaftsuniversita¨t Wien, Vienna, Austria Abstract Purpose – The paper seeks to analyze empirically the consequences of family responsibilities for career success and the influence of career context variables and gender on this relationship. Design/methodology/approach – The sample consists of 305 business school graduates (52 percent male) from a major Central European university who finished their studies around 2000 and who were in their early career stages (i.e. third and fourth career years). Findings – The paper reports a negative relationship between family responsibilities and objective and subjective career success via work centrality. There is also substantive support for the effect of contextual factors on the relationship between family situations and career success, emphasizing the importance of a multi-level perspective. Finally, evidence of gender effects exists. Research limitations/implications – The empirical generalizability of the results is limited by the structure of the sample. Qualitative in-depth studies are needed to further understand the relationships found. Practical implications – The results underscore the importance of the work-family-interface for employee retention measures. Tailored HR policies are crucial. Originality/value – Theoretically, the paper develops a multi-level causal model of specific aspects of work-family relations including variables ranging from meso (career context) to more micro (family, individual). Empirically, the study focuses on young business professionals prior to having a family or in the early stages of their family life. Keywords Sociology of work, Family, Careers, Gender Paper type Research paper Journal of Managerial Psychology Vol. 23 No. 3, 2008 pp. 292-323 q Emerald Group Publishing Limited 0268-3946 DOI 10.1108/02683940810861392
Introduction Work-family relations are central to individuals, organizations and policy makers. For individuals, they relate to issues such as life satisfaction, work-life balance, or career
success (e.g. Greenhaus et al., 2003; Kirchmeyer, 2006; Ford et al., 2007). For organizations, work-family relations touch on individual and organizational performance, scope of flexibility in terms of work capacity or organizational programs allowing employees to better combine private and occupational demands (e.g. Hall, 1990; Osterman, 1995; Kossek and Lambert, 2005). For policy makers, they lead to policies such as regulations about better combining family life with paid work, support for individuals re-entering work life after familial leaves of absence or legislative frameworks for working time and conditions (e.g. Esping-Andersen, 2000; Poelmans and Sahibzada, 2004). Overall, results of the substantial body of work-family studies are quite sobering. While a wide range of views exists, it seems fair to say that far more suggest a negative spillover in terms of work-family conflict than the other way round, i.e. that work and family can positively influence one another (see, for example, Eby et al., 2005). In addition, some aspects of this relationship are clearly less researched. Among them is, first, the context of work careers. Work-family relations are frequently analyzed without relating them to the specific career context relevant to individuals. While the macro-context of legal regulations, population demography or economic development is regularly taken into account (e.g. Poelmans and Sahibzada, 2004), the importance of the specific career context, i.e. factors such as characteristics of the profession or the kind of careers individuals are pursuing, is less acknowledged (for notable exceptions see, for example, Greenhaus et al., 2001; Perry-Smith and Blum, 2000). Second, the importance of gender for work-family relations is not yet fully developed. While a gender perspective is prominent in work-family and career research and many results point towards gender based differences in family relations and work careers, calls for more fine-grained analyses of different aspects of gender effects abound (e.g. Higgens et al., 2000; Stoner et al., 1990). The current paper responds to these deficits by taking up a classical issue of work-family studies – effects of the family situation on careers, analyzing here the effects of family responsibilities on career success – and developing it further through two differentiations cutting across various levels of analysis: first, by analyzing the effects in different types of career contexts, and second by differentiating along the gender dimension. Specifically, this paper addresses three major questions: (1) What effects do individuals’ family responsibilities have on their objective and subjective career success? (2) How do core characteristics of individuals’ career context influence these effects? (3) Are these effects different for men and women? In answering these questions, the current paper proposes a causal model with crucial linkages between family responsibility (family level), career context (meso-level) and career success (individual level). Using a sample of young business professionals, the paper investigates empirically the proposed relationships using partial least squares analyses. The paper contributes to the theoretical as well as empirical advancement of the field. Regarding the former, the paper further develops specific aspects of work-family relations, presents a model based on current literature and includes variables from different conceptual levels ranging from meso (career context) to more micro (family, individual). Empirically, the study focuses on a specific group: young
Responsibilities, career fields and gender 293
JMP 23,3
business professionals prior to having a family or in the early stages of their family life. Thus, insight into this specific group, which is important for many businesses, is generated and possible generalizations are discussed.
294
Conceptual background Research on the work-family interface, careers and the importance of work provides essential insight into the link between family responsibilities and career outcomes. It points towards the importance of work centrality as a key linking factor and emphasizes gender as a differentiating variable. Work-family interface The past decades have seen the emergence of new configurations of “work and family” with more women in the labor force and more dual-earner families (Tharenou, 1999). Dual-earner households are supplying more working hours to the labor market than ever before (Edwards and Wajcman, 2005, p. 47). In the last 50 years partnership ideas have changed for both men and women (Jacobs and Gerson, 2001; Erler, 1996; Peuckert, 1996; Herlth et al., 1994; Kaufmann, 1990). From the perspective of the early twenty-first century in industrialized countries, fertility is a choice variable and cohabitation, separation and divorce are commonplace. Because of these changing and separating views of roles, the supporting activities of (female) partners can no longer be taken for granted. The forms of partnerships and families have differentiated (Schmidt, 2001): rather traditional forms (living together in the same household with or without marriage), couples living and working in various locations (so called commuters or long-distance marriages), or couples living in the same area but who have decided to have separate households (so called living apart together couples). Within all these forms we can find various constellations with no child, one child or more children. In addition, it varies whether the children are living with their biological parents or not (patchwork families or stepfamilies). Parental demands are a function of the number and ages of the children, and the age of the youngest child (Voydanoff, 1988; Lewis and Cooper, 1987; Greenhaus and Parasuraman, 1986). They are highest for persons with infants and pre-school children, lower for those with school-age children, and lowest for those with adult children not living at home (Osherson and Dill, 1983). Family involvement refers to the importance of the family to an individual and the extent of psychological investment in the family. As in the case of job involvement, family involvement is likely to generate internal pressures to invest increased effort and energy in the family domain to fulfill family role demands (Parasuraman and Simmers, 2001, p. 555). Beside the number of children, the work life of the partners and their career orientations (e.g. single breadwinner orientation versus dual earning constellations or dual career couples) are important factors influencing the form of family responsibility (e.g. Blossfeld and Drobnic, 2001; Lewis and Cooper, 1987). Models of the family concern the allocation of roles and responsibilities, role specialization and the division of labor among adult family members (Hakim, 2005, p. 56). Numerous studies have examined characteristics of the family domain as predictors of work-family conflict. Most of the research at least implicitly assumes that work-family conflict negatively affects the psychic and physical condition of individuals and their enjoyment of work and life. As a consequence, disadvantages for
companies may result because of reduced productivity and increased turnover (Greenhaus et al., 2001; Cooper and Williams, 1994; Ganster and Schaubroeck, 1991). These studies have found that conflict is higher among those who: . have children at home (Behson, 2002; Carlson, 1999); . are concerned or troubled about child care (e.g. Fox and Dwyer, 1999); . have greater time demands from family (e.g. Parasuraman and Simmers, 2001); . have disagreements with their family or partner (e.g. Day and Chamberlain, 2006; Williams and Alliger, 1994); and . have less family support (Grzywacz and Marks, 2000). Family responsibilities are one important factor influencing the amount of time and energy that individuals are able and willing to devote to work. Consistent with gender-based normative expectations, women still generally bear primary responsibility for home maintenance and childcare irrespective of their employment status. Although employed married women spend less time on housework and childcare than non-employed women, they devote considerably more time to home and family in fulfilling their family role responsibilities than men (Parasuraman and Simmers, 2001; Pleck, 1985). Researchers have argued that in a workforce that is increasingly composed of individuals in “post-traditional families”, work-family conflict may have a significant impact on how individuals view their career outcomes (Kirchmeyer, 2006; Schneer and Reitman, 2002). Under conditions of new “protean” careers (Hall, 1996), self-employment and increasingly flexible employment contracts, work-family conflict has not become obsolete. On the contrary, the demands are even higher for the self-employed (Jennings and McDougald, 2007; Parasuraman and Simmers, 2001; Loscocco, 1997). Characteristics of the work context are important predictors for work-family conflict, too. In particular, work variability (Fox and Dwyer, 1999) and forms of working hours such as working weekends (Schneider, 2005) are related to higher conflict. Conflict seems to be higher among those who work a greater number of hours (Grzywacz and Marks, 2000; Carlson and Perrewe, 1999; Greenhaus et al., 1987). This indicates that the work and career contexts of individuals (i.e. factors such as demands and pressures from work, degree of turbulence, etc.) constitute important influencing factors when analyzing the effects of family situations on career success. Careers Career success. Career success is an integral part of career research (for a recent overview, see, for example, Gunz and Heslin, 2005). Despite the huge body of literature on factors influencing career success, little scholarly attention has been devoted to analyzing the concept of career success itself (Greenhaus, 2003; Heslin, 2003; Sturges, 1999). One framework that is widely accepted in career research is Hughes’s (1937, 1951) distinction between objective and subjective career success. The former is defined as directly observable, measurable, and verifiable by an impartial third party when looking at attainments such as pay, promotions, or occupational status. The latter is only experienced directly by the person and defined by an individual’s reactions to his or her unfolding career experiences (Hughes, 1937, 1951). It heavily depends on individuals’ (re)construction of career success according to subjective and
Responsibilities, career fields and gender 295
JMP 23,3
296
individualized patterns. Objective and subjective views on careers constitute a “two-sidedness” inherent in the career concept. The subjective-objective career duality expresses these two dimensions as unique, empirically distinct constructs (Arthur et al., 2005) showing different patterns of correlations with commonly used predictor variables (Ng et al., 2005). Arthur and Rousseau (1996) found that more than 75 per cent of the career-related articles published in major interdisciplinary journals between 1980 and 1994 focused on objective perspectives. Over the last decade, however, subjective criteria have increasingly been adopted (see, for example, the literature review by Arthur et al., 2005). Importance of career context. Work careers are embedded in the broader economic and institutional environment. In career terms, contributions from labor economics as well as sociology for a long time have specifically dealt with issues like labor market segmentation, stratification and dual labor markets (Tolbert, 1982; Edwards, 1975; Piore, 1975; Doeringer and Piore, 1971) and their relationship to various aspects of careers, especially earnings/wages (Theodossiou, 1995) and mobility (Bernhardt et al., 2001). The career literature in the more narrow sense takes a more restricted and a greatly diverse perspective. It includes contributions such as the specific characteristics of occupations and their labor market consequences (Tolbert, 1996) or the analysis of specific aspects such as the dual labor market hypothesis (Leontaridi, 2002). Based on the work of French sociologist Pierre Bourdieu (e.g. Bourdieu, 1977), Iellatchitch et al. (2003) present a field and habitus perspective of careers emphasizing the role of the context for individual careers. For Bourdieu, a social field is a patterned set of practices. In this playground various actors with a field-relevant capital try to advance their position by following individual strategies. Careers as the sequence of positions influenced by work related individual efforts are not a field, but unfold within a field. These career fields are the social context within which individual members of the work force make their moves. Individuals try to maintain or improve their place in the given and unfolding network of work related positions through their field-relevant career capital and a patterned set of practices. The latter are constrained by their career capital and the rules of the field and, in turn, contribute to the shaping of these rules. Career fields can be differentiated along two core dimensions. Coupling focuses on the closeness of relationships and the degree of mutual influence between the focal actor and the other actor(s) in the field. Configuration focuses on changes in the configuration of relationships between the focal actors and other relevant actors over a longer period of time. Importance of work Work is an important activity for individuals. Particularly in industrialized societies, work is not only a means of survival, but also a major way of expressing and developing oneself and a source of social recognition. Hinting at the importance of work for individuals (Harpaz, 1986), work centrality is conceptualized and measured in three major ways: (1) as a set of integrated and interrelated beliefs; (2) as individuals’ preference in terms of work; and (3) as related to central life interests (i.e. the attachment and commitment to work in general; Pryor and Davies, 1989).
Compared to job involvement (i.e. the extent of preoccupation and immersion in present jobs; Paullay et al., 1994) and organizational centrality (i.e. the extent to which an employee is integrated into the network of interpersonal relationships within the work system; (O’Hara et al., 1994), work centrality is the broader concept (Diefendorff et al., 2002). Work centrality is related to a number of outcomes at the individual level. This includes a positive relationship on career outcome variables such as organizational commitment, career planning, and wages (Mannheim et al., 1997), the hours spent on work (Snir and Harpaz, 2006), attitudes and job performance (Peterson and Ruiz-Quintanilla, 2003), workaholism (Harpaz and Snir, 2003), job satisfaction and participation in decision making (Kanungo, 1982), and job tenure (Dubin et al., 1975). Gender plays an important role when analyzing the effects of work centrality. Research indicates that gender moderates some of the effects, e.g. in the area of job involvement and organizational citizenship behavior (Diefendorff et al., 2002), workaholism (Harpaz and Snir, 2003) or determining factors of work centrality such as socio-economic status, work values or socialization (Mannheim, 1993). A causal model: family responsibility, career fields and career success Based on the research outlined briefly above, the following issues related to the three research questions emerge. From the discussion about the importance of work, work centrality evolves as crucial construct. Hence, this variable plays a central role in the causal model as a “linking pin” between the contextual variables, i.e. the family and work situation, and individual career outcome variables. The literature on the work-family interface points towards family responsibility as a core variable when looking at effects of the family situation on career success. It influences – among other things – the amount of time and energy that individuals are able and willing to devote to work, i.e. their work centrality. In turn, work centrality is an important factor for both objective and subjective career success. Research on career success clearly points towards these two distinctive dimensions of career success with objective careers success influencing subjective success. Also stemming from the extant career literature and belonging to the more confined work on career context (for an overview see Mayrhofer et al., 2007) are the core characteristics of the respective career fields. In particular, changeability and job alternatives directly influence work centrality. In addition, changeability has a moderating effect on the link between family responsibility and work centrality. Fourth, gender as a major “cross-sectional” phenomenon has an effect on the link between family responsibilities and work centrality. Hence, we propose a model linking three conceptual levels (i.e. individual, family, and career context) and relating the central variables of our research (i.e. family responsibility, work centrality, career fields and career success) to each other (see Figure 1). It is explained in more detail below. Family responsibility and its effects A overview of the European Foundation for the Improvement of Living and Working Conditions (2003, p. 48) showed that on average across 16 European countries, men would prefer to work a 36.5-hour week and women a 30-hour week. Family responsibilities are the main reason mentioned for this interest in reducing actual working hours. Therefore, family responsibilities and the wish or need for other
Responsibilities, career fields and gender 297
JMP 23,3
298
Figure 1. Family responsibility, career fields and career success: a causal model
working time arrangements reduce the perceived job alternatives. In addition, a growing expectation of mobility exists (Schneider, 2005). Jobs in high qualification segments are especially linked with a growing amount of business travel (see, for example, Harris et al., 2005). Hence, family responsibilities reduce the number of jobs available. The assumption that workers can choose the amount of overall working hours freely is only partly correct. Bo¨heim and Taylor (2004) state that rigidities in the labor market still remain especially for jobs with flexible work hours. Family responsibilities also influence the importance of work. Family status has been found to play an important role in individuals work experiences (e.g. Stroh et al., 1996; Tharenou et al., 1994; Schneer and Reitman, 1993). Resources such as time, attention, and energy are finite, and those expended in one domain are unavailable for other domains. This constraint yields a negative direct relationship between family and work resources available for the other domain (Edwards and Rothbard, 2000). Gender plays an important role for the work related effects of family responsibilities. Gendered roles affect the conditions and consequences of the work-family conflict (Carlson and Kacmar, 2000). Women’s job satisfaction depends on family member’s emotional state (King et al., 1995), and the determinants of job satisfaction change essentially after the birth of a child (Holtzman and Glass, 1999). The presence of children, the employment status of one’s partner, the form of partnership and household – all these factors influence the perception of job centrality and have negative effects on the experienced work centrality. In particular, female managers who experienced high levels of family role salience and long work hours (also) experienced the highest of work-family conflict (Stoner et al., 1990).
Career fields and work centrality As mentioned above, we focus on two characteristics of career fields: (1) job alternatives constitute a core indicator for tight/loose coupling: the fewer job alternatives an actor has in their evoked set, the tighter coupled they are (e.g. with an employer) and the more the current work situation is in danger of being a dead-end position; and (2) changeability of both work content and professional relationships represent the dimension of configuration. Prior research is barely sufficient to profoundly assess the impact of both dimensions of career fields on work centrality. However, there are some results on related issues. Increasing dynamics generated new forms of vocational identities, especially “new career negotiators” in the telecommunications industry (Dif, 2004), which indicates a higher importance of job-related issues. Within a sample of technical contractors, a majority worked longer hours and rarely scheduled their time in a flexible way despite conditions of new temporal flexibility and dynamics (Evans et al., 2004). Several scholars argue that increasing flexibility and changeability of work conditions do not indispose the importance of psychological contracts (Marsden, 2004; McGovern et al., 1996). New contracts might imply higher work centrality. As changeability is concerned, we assume that actors within a more turbulent work context tend towards a higher work centrality, as these situations command more attention than stable ones and attract more time and energy, detracting it from the leisure and family sphere. In general, we suppose that the overall relationship between changeability and work centrality has an inverse U-shape. However, our sample does not reach the point where turbulence at work detracts attention. Thus, we assume a positive relation. The same direction of relationship is assumed for job alternatives and work centrality: the more alternatives actors have and/or perceive to have in labor markets and career fields, the more important work will get for them and the more hours they will work. Having alternatives and opportunities is a clear indicator of an individual’s value and position on labor markets, thus strengthening work-specific self-efficacy and self confidence. Contrarily, a loss of career perspectives and job alternatives results in locked-in effects (van Ours, 2004), thus reducing not only job search activities, but also work centrality and job-involvement. This direction of impact is not supported by economic research, which shows a countercyclical characteristic of working hours and vacation time (e.g. Altonji and Usui, 2007) and negative impacts of labor market rigidities on individually preferred reduction of working hours (Bo¨heim and Taylor, 2004). For our sample of young business professionals, we nevertheless assume that the first effect is stronger than the latter. At least in early career phases we assume a strong reinforcing cycle between the perceptions of one’s professional attractiveness and an actor’s investment of time and energy into work (success to the successful; Senge, 1990, Appendix 2). Work centrality and career success Work centrality is positively related to both dimensions of career success. For objective career success, research suggests that the time and energy devoted to work has positive effects on career outcomes such as income or hierarchical advancement (see, for example, Mannheim et al., 1997; for similar effects of job involvement see, for
Responsibilities, career fields and gender 299
JMP 23,3
300
example, Rottenberry and Moberg, 2007). For subjective career success, a positive link to work centrality can be assumed, too. From a dissonance theoretical point of view (Festinger, 1959) it is highly probable that the degree of investment of time and energy into work positively relates to subjective impressions of being successful. In order to avoid cognitive dissonance, individuals putting a lot of emphasis on work are under internal pressure to see their work involvement and outcomes more positively than persons where work plays a lesser role. Objective and subjective career success Various possibilities of influencing directions between objective and subjective career success have been formulated. Most frequently it is assumed that objective success has a positive influence on subjective success (e.g. Korman et al., 1981) since individuals interpret their subjective success on the basis of their objective accomplishments (Judge et al., 1995). In their review, Arthur et al. (2005) also identify studies considering a two-way interdependence between the two dimensions of career success. Individuals constantly interpret and reinterpret the work experience and career success they have had. They experience certain levels of objective success, create understandings about what constitutes career success for them and individually act on those understandings. Empirically unfolding as a moderate correlation between the two (Ng et al., 2005), their relationship stays a complex one (Nicholson and DeWaal-Andrews, 2005, p. 142). Sample, variables and methods Data collection and sample The data for this study were collected via questionnaire during the 2004 and 2005 follow-up surveys of a panel study started in 2000. The sample consists of 305 business school graduates (52 percent male) from a major Central European university who finished their studies around 2000. The data roughly cover the participants’ third and fourth career year – still the career entry stage, but not the immediate beginning. The mean sample age is 32.2 years (women: 31:7 ^ 2:8; men: 32:8 ^ 2:4). Concerning age and gender distribution, the sample is representative for the graduates of the whole university at the time, which supplies roughly half the national market for business school graduates. As to marital status and number of children, 33 percent/30 percent were single in the first and second survey round, respectively; 11 percent/7 percent lived in separate households, 29 percent/29 percent lived in a shared household, and 27 percent/32 percent were married. Regarding the number of children, there was virtually no difference between the survey rounds: 83 percent had no child, 11 percent had one child, and the remaining 6 percent had two or more children. Although this low percentage of parents seems quite normal for a sample of young business professionals, it makes for an extremely skewed distribution. Still, this should not distort the results, as the method employed (see below) is quite robust against skewed data (Cassel et al., 1999) and the chosen index for family responsibility (see “Measures” section) actually has very moderate skewness values (0.27 and 0.09 for the first and second survey rounds, respectively). The inclusion of other forms of family responsibilities such as elder care would have been beneficial. Unfortunately, no systematic data was available in this respect.
Methods We analyzed our data with a partial least squares (PLS) procedure (e.g. Chin, 1998; Lohmoeller, 1989; Wold, 1975), which allows structural modeling with latent variables but has less strict assumptions than “traditional” covariance-based structural equation modeling (LISREL approach) concerning sample size, level of measurement, and multinormality (Fornell and Bookstein, 1982, p. 440). As in LISREL, the complete model consists of two components: (1) the measurement (or outer) model where the observed variables form the latent variables; and (2) the structural (or inner) model, which deals with the relationships between the latent variables (e.g. Tenenhaus et al., 2005, p. 161). An evaluation of these two model components for this study is presented in the “Results” section. The PLS calculations were conducted with SmartPLS 2.0 (Ringle et al., 2005); all other calculations, including standardizing the variables before conducting the PLS analyses, were done with SPSS. Standardization was especially relevant for including the interaction term (Chin et al., 1996, p. 26), with changeability supposed to moderate the effect of family responsibility on work centrality. The significance of the path coefficients was determined using a bootstrap procedure (e.g. Chin, 1998) with 500 subsamples. Gender differences were tested for statistical significance by calculating t-values, using a formula proposed by Chin (2000) (see also Sanchez-Franco, 2006, p. 30; Keil et al., 2000, p. 315), i.e.: PCwomen 2 PCmen t ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffi ; 2 2 2 ðm21Þ ðn21Þ2 1 1 mþn ðmþn22Þ £ SE women þ ðmþn22Þ £ SEmen £ where m and n are the number of men and women in the sample, respectively, PC is the path coefficient and SE is the standard error of the path in the structural model. Although PLS allows formative as well as reflective measures (Sanchez-Franco, 2006, p. 26; for a detailed discussion of reflective/formative measures, see, for example, Jarvis et al., 2003) and especially family responsibility basically suggests a formative approach, our variables do not cover the range of relevant indicators for this variable, such as the partner’s professional status and/or career/family orientation, or available social support (see above). As sound formative measurement requests that all relevant indicators be included (Bollen and Lennox, 1991, p. 307), we chose reflective measurement models throughout. Measures The following variables from both survey rounds were entered into the model as indicators of the latent variables. For family responsibility, the available data were the number of children and marital status (single, with partner/separate households, with partner/shared household, married). Based on the reasoning that being single without children usually means less family responsibility than living in partnership and the assumption that family responsibilities increase when one has to take care of children, the following scheme was applied to quantify family responsibility: for participants without children, being single scored 0, a partner living in a separate household scored
Responsibilities, career fields and gender 301
JMP 23,3
302
1, a shared household or being married scored 2. For participants with children, the same marital statuses scored 4, 3, and 2, respectively, with the number of children being added to the score. For example, this leads to the following scores: single, no child: 0; married, no child: 2; married, one child: 3; married, two children: 4; single parent, one child: 5. Basically, the children’s age arguably plays an important role for the family responsibility, too. However, in the latter of the two follow-up surveys one year ago, the mean age of the children was 2.4 years (^ 2.1 years). As all children of our sample are still in pre-adolescence infanthood, we did not include the children’s age in our analyses. Furthermore, all children lived in the participants’ households, which is why there was no need to differentiate between children living in the household and children living elsewhere in the analyses. In a related vein, family responsibility may include duties like elder care, obligations towards “chosen” family members, or community responsibilities, which are not included in our study owing to omission of these data in the survey. Although this is admittedly a minor drawback for this topic, we assume that marital status and childcare are by far the most important determinants of family responsibility, especially for our sample. Career changeability was measured with two items: (1) instability of work content (11-point scale ranging from “very stable” to “ever-changing”); and (2) instability of professional relations (same scaling). Perceived job alternatives were measured by the question how easily another adequate job could be found should the need arise (11-point scale from “not at all” to “very easily”). Work centrality was measured by focusing on behavior/output related as well as attitudinal aspects. For the former, we used reported actual work hours per week; for the latter, we used the proportion of “life energy” invested in the job (11-point scale ranging from 0 percent to 100 percent). Subjective career success was measured by two ratings: (1) career satisfaction (11-point scale ranging from “extremely dissatisfied” to “extremely satisfied”); and (2) perceived career success (11-point scale ranging from “totally unsuccessful” to “extremely successful”). Objective career success was measured by the reported annual income and the perceived amount of managerial/leadership-related tasks in the job (11-point scale ranging from 0 percent to 100 percent). Despite a sometimes fierce discussion about the use of objective and subjective success measures in organizational research (for firm performance see, for example, Wall et al., 2004) and the frequent use of perceptional measures in career research, it certainly has its limits. Hence, some additional analyses were conducted in order to see whether single-source biases can be detected. Table I shows the correlation matrix for all variables. It indicates no irregularities pointing towards problems with the variables. The fact that all these variables were collected via a self-report questionnaire raises the issue of single-source bias (e.g. Podsakoff and Organ, 1986). Therefore, we conducted a Harman one-factor test as a preliminary analysis. Although more of a
Variable
1
2
Family responsibility 1. Score 1 2. Score 2
– 0.82
–
Job alternatives 3. Perceived job alternatives 1 4. Perceived job alternatives 2 Changeability 5. Instability of 6. Instability of 7. Instability of relations 1 8. Instability of relations 2
work content 1 work content 2 professional
2 0.04 0.03
0.03 0.08
3
4
– 0.65
–
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
2 0.05 2 0.02 2 0.10 2 0.11
0.01 0.05 0.11 2 0.02
– 0.47
–
2 0.04 2 0.00
0.01
0.00
0.56
0.39
–
2 0.10 2 0.05
0.15
0.02
0.35
0.63
0.56
–
Work centrality 9. Weekly work hours 1 10. Weekly work hours 2 11. Percentage of energy for job 1 12. Percentage of energy for job 2
2 0.17 2 0.14 2 0.12 2 0.10
0.08 2 0.01 0.27 0.25 0.08 0.02 0.20 0.25
0.17 0.17 0.21 0.08
0.14 0.19 0.09 0.08
0.23 0.22 0.17 0.15
0.17 0.23 0.16 0.11
Subjective career success 13. Career satisfaction 1 14. Career satisfaction 2 15. Perceived career success 1 16. Perceived career success 2
2 0.02 0.08 0.16 0.13 2 0.05 0.03 0.01 2 0.02
0.15 0.28 0.21 0.23
0.16 0.11 0.07 2 0.04 2 0.01 0.06 0.28 2 0.01 20.05 2 0.02 2 0.02 20.03 0.16 0.26 0.14 0.11 0.16 0.30 0.32 0.11 0.04 0.16 2 0.02 0.13
0.02 0.16 0.17 2 0.08 0.21 0.36 0.28 0.10
0.27 – 0.25 0.44 – 0.38 0.60 0.37 – 0.52 0.47 0.61 0.52 –
2 0.00 2 0.00
0.07 0.16
0.09 0.21
0.13 0.12
0.12 0.03
0.24 0.15
0.46 0.33
0.37 0.49
0.25 0.16
0.25 0.31
0.08 0.13
0.07 0.17
0.25 0.22
0.14 – 0.25 0.84 –
0.08
0.13 2 0.01
0.02
0.16
0.06 2 0.03 2 0.06
0.12
0.22
0.07
0.12
0.19
0.02
0.24
0.10
0.29
0.40 –
0.05
0.11
0.13
0.25
0.14
0.12
0.25
0.11
0.31
0.23
0.24
0.31
0.33
0.24
0.32
20
professional
Objective career success 17. Income 1 (EURO) 18. Income 2 (EURO) 19. Amount of managerial/ leadership tasks 1 20. Amount of managerial/ leadership tasks 2
2 0.18 2 0.08 2 0.04 2 0.09
0.05 0.06
0.14
0.15 0.10
0.05 2 0.04
– 0.47 – 0.65 0.29 0.46 0.54
– 0.51 –
0.68
–
Notes: Listwise n ¼ 130. All correlations $ 0.18 are significant at the 0.05 level; all correlations $ 0.23 are significant at the 0.01 level (two-tailed)
Responsibilities, career fields and gender
303
Table I. Correlation matrix
JMP 23,3
rough diagnostic tool than a remedy, its results at least suggest that single-source bias poses no problem for this study: there was no “general factor” uniting supposedly unrelated variables, and the principal components extraction resulted in seven components with eigenvalues . 1. Beyond that, the perceptional nature of the data has to be taken into account when interpreting the findings.
304
Results Descriptive results Table II presents means and standard deviations for all variables entered into the PLS model for both survey rounds (labeled 1 and 2), for the whole sample and separated by gender. The sample numbers reported represent the minimum count of valid cases and are smaller than the original sample size owing to missing values. Measurement model Pertinent literature (e.g. Goetz and Liehr-Gobbers, 2004; Hulland, 1999) suggests a three-step procedure for evaluating the measurement model: (1) individual item reliabilities (i.e. their loadings/correlations with the latent variable); (2) composite reliability of the measures; and (3) discriminant validity. Concerning item loadings, it is recommended that all items have loadings . 0.7, and that items with loadings , 0.4 be dropped from the analysis (Hulland, 1999; Carmines and Zeller, 1979). In the present study, item loadings range from 0.95 to 0.64 for the complete model, with about one third of the variables having loadings slightly below 0.7. For the models separated by gender these values deteriorate slightly. There is still a minimum loading value of 0.57. The model has 50 percent and 85 percent of the loadings above the 0.7 threshold for the female and male models, respectively. The bootstrapping procedure resulted in all loadings for all models being significant at the 0.001 level. Composite reliability can be assessed via Cronbach’s a and the internal consistency measure proposed by Fornell and Larcker (1981), with a recommended minimum value of 0.7 for both (Nunnally, 1978). For the present model, both values are almost identical and universally . 0.7, ranging from 0.77 (objective career success) to 0.91 (family responsibility) for the total model, with virtually no difference for the gender-specific models and a minimum value of 0.73 (job alternatives in the female sample). Discriminant validity is examined via the average variance shared between a latent variable and its indicators (average variance extracted, AVE). This value should be larger than 0.5 (Fornell and Larcker, 1981, p. 46) and its square root should be considerably larger than the correlations of the latent variable with the other latent variables (Hulland, 1999, p. 200). For the present study, merely the AVE of work centrality (0.49) and objective career success (0.46) fall slightly short of the 0.5 threshold (with the gender-specific models even performing marginally better here). By contrast, the square root of the AVE clearly exceeds the correlations with the other constructs for all latent variables and all three models. Table III shows the statistical characteristics of the measurement models. Despite the satisfactory results, there are some caveats. First, the number of indicators per
Total (n $ 204)
Mean (SD) Women (n $ 91)
Men (n $ 110)
Family responsibility Score 1 Score 2
1.57 (1.37) 1.64 (1.34)
1.54 (1.28) 1.80 (1.27)
1.60 (1.46) 1.52 (1.40)
Job alternatives Perceived job alternatives 1 Perceived job alternatives 2
7.02 (2.54) 7.50 (2.58)
6.91 (2.47) 7.33 (2.62)
7.16 (2.61) 7.66 (2.53)
Changeability Instability of work content 1 Instability of work content 2 Instability of professional relations 1 Instability of professional relations 2
5.67 5.48 4.60 4.42
5.75 5.44 4.69 4.08
5.59 5.50 4.51 4.66
Variable
Work centrality Weekly work hours 1 Weekly work hours 2 Percentage of energy for job 1 Percentage of energy for job 2 Subjective career success Career satisfaction 1 Career satisfaction 2 Perceived career success 1 Perceived career success 2 Objective career success Income 1 (e) Income 2 (e) Amount of managerial/leadership tasks 1 Amount of managerial/leadership tasks 2
46.0 47.0 65.9 65.0
(2.78) (2.84) (2.78) (2.60)
(9.2) (11.2) (16.1) (18.3)
7.81 8.06 8.43 8.63
(2.38) (2.41) (1.69) (1.74)
39,987 (16,503) 46,141 (22,460) 33.4 (25.3) 34.9 (27.3)
44.8 43.3 65.6 61.8
(2.80) (2.97) (2.84) (2.53)
(9.7) (11.7) (18.1) (21.5)
8.00 8.15 8.35 8.53
(2.41) (2.28) (1.81) (1.91)
33,935 (15,110) 38,594 (17,842) 30.8 (24.7) 29.1 (25.4)
47.1 50.1 66.2 67.7
(2.76) (2.74) (2.75) (2.56)
(8.6) (9.9) (14.0) (14.6)
7.63 7.96 8.51 8.72
t-test
* *** **
(2.34) (2.56) (1.58) (1.60)
45,456 (15,895) 52,161 (24,270) 35.7 (25.8) 39.2 (27.6)
*** *** ***
Notes: *p , 0:1; * *p , 0:05; * * *p , 0:01 (two-tailed)
Responsibilities, career fields and gender
305
Table II. Means and standard deviations of all observed variables
Score 1 Score 2 2. Job alternatives Perceived job alternatives 1 Perceived job alternatives 2 3. Changeability Instability of work content 1 Instability of work content 2
CR loadings (total/women/men)
1.
0.91/ 0.90/ 0.89 0.872/ 0.873/ 0.881 0.953/ 0.941/ 0.914 0.90/ 0.82/ 0.95 0.887/ 0.753/ 0.980 0.926/ 0.905/ 0.921 0.85/ 0.80/ 0.88 0.731/ 0.674/ 0.763 0.708/ 0.596/ 0.821
0.913/ 0.907/ 0.898
Square root (AVE) LV correlations 2. 3. 4. 5.
0.037/ 0.048/ 0.003
0.907/ 0.833/ 0.951
20.072/ 20.212/ 20.007
0.054/ 0.154/ 20.001
JMP 23,3
1. Family responsibility
306
Table III. Summary of the measurement models Latent variable (LV) Indicator variable
6.
0.769/ 0.713/ 0.809
(continued)
Latent variable (LV) Indicator variable Instability of professional relations 1 Instability of professional relations 2 4. Work centrality Weekly work hours 1 Weekly work hours 2 Percentage of energy for job 1 Percentage of energy for job 2 5. Subjective career success Career satisfaction 1
CR loadings (total/women/men) 0.821/ 0.844/ 0.724 0.811/ 0.715/ 0.915 0.79/ 0.80/ 0.81 0.801/ 0.848/ 0.760 0.684/ 0.739/ 0.593 0.635/ 0.569/ 0.793 0.666/ 0.659/ 0.711 0.82/ 0.79/ 0.84 0.704/ 0.656/ 0.750
1.
Square root (AVE) LV correlations 2. 3. 4. 5.
20.154/ 20.349/ 20.026
0.212/ 0.213/ 0.234
0.268/ 0.367/ 0.227
0.700/ 0.711/ 0.718
0.037/ 20.047/ 0.095
0.315/ 0.227/ 0.403
0.118/ 0.196/ 0.077
0.352/ 0.390/ 0.353
6.
0.733/ 0.698/ 0.754
(continued)
Responsibilities, career fields and gender
307
Table III.
Perceived career success 1 Perceived career success 2 6. Objective career success Income 1 (e) Income 2 (e) Amount of managerial/leadership tasks 1 Amount of managerial/leadership tasks 2
CR loadings (total/women/men) 0.693/ 0.605/ 0.718 0.651/ 0.589/ 0.729 0.866/ 0.898/ 0.816 0.77/ 0.79/ 0.77 0.736/ 0.733/ 0.766 0.653/ 0.824/ 0.566 0.664/ 0.606/ 0.665 0.665/ 0.610/ 0.712
Notes: CR, composite reliability; AVE, average variance extracted
1.
0.081/ 20.219/ 0.241
JMP 23,3
Career satisfaction 2
308
Table III. Latent variable (LV) Indicator variable
Square root (AVE) LV correlations 2. 3. 4. 5.
0.155/ 0.364/ 0.041
0.170/ 0.343/ 0.069
0.442/ 0.536/ 0.410
0.326/ 0.267/ 0.436
6.
0.680/ 0.699/ 0.681
construct is rather low, which may lead to slightly biased estimates (underestimation of the structural paths linking the latent variables, and overestimation of the loadings; e.g. Chin et al., 1996), despite the adequate sample size. Second, job alternatives is a single-item measure. This is an additional caveat, even though there are claims that for specific and homogeneous constructs the use of single-item measures is acceptable (Loo, 2002; Robins et al., 2001; Gardner et al., 1998).
Responsibilities, career fields and gender 309
Structural model Figure 2 shows the results for the structural model, for the whole sample and for women and men separately. For all path coefficients except Family responsibility ! Job alternatives and Changeability ! Work centrality, the differences between women and men are statistically significant at the 0.01 level (two-tailed). In line with the proposed relationships, for the total sample the structural model shows that work centrality is positively influenced by both career field variables, i.e. changeability and job alternatives, and negatively influenced by family responsibilities and the interaction term, i.e. changeability moderating the influence of family responsibility. In turn, work centrality as predicted positively influences both objective and subjective career success. Objective career success has a positive relationship with subjective career success, again in line with the proposed model. All relationships for the total sample are statistically significant at the 0.05 or 0.01 level, respectively. Deviating from the model is only the lack of relationship between family responsibility and job alternatives, suggesting that these are actually unrelated constructs.
Figure 2. Structural model: results of analyses
JMP 23,3
310
When looking at gender differences, all statistically significant relationships for the overall sample show significant differences between women and men, the only exception being the path between changeability and work centrality. Otherwise, negative effects of family responsibility on work centrality and positive effects of the latter on both forms of career success are stronger for women. For job alternatives, the positive effect on work centrality is stronger for men. Beyond that, three interesting results emerge, all of them modifying the results of the overall sample and the propositions from the model. First, for men no statistically significant relationship exists between family responsibility and work centrality. For the overall and the female sample the expected negative relationship exists. However, this is not the case for men. Second, while for the overall and male sample, career field effects the results conform to our predictions, this is not the case for the women. For the former, the interaction term with changeability moderating the influence of family responsibility shows a negative effect. An increase in changeability by one standard deviation would increase the negative effect of family responsibility on work centrality, e.g. for the total sample from 2 0.13 to 2 0.3. For the women in our sample, by contrast, increased changeability has a positive effect on work centrality, i.e. it alleviates the hampering effect of family responsibility. For the men, where at the outset and contrary to our predictions, family responsibility has virtually no effect on work centrality, an increase in changeability creates the negative effect initially only experienced by the female part. Regarding the predictive value of the chosen variables for work centrality, career context and family variables plus the moderator term explain a higher proportion of variance for women than for men. Third, there is a nonsignificant path from objective to subjective career success in the female sample. In contrast, for the overall sample as well as for men there is a comparatively strong positive relationship, indicating that subjective career success depends on objective success. This indicates that women do not depend on objective success in their evaluation of subjective career success, while men do. Table IV shows the structural model’s statistical characteristics. As with the measurement model, the structural model has some limitations. The R 2 values of the endogenous variables are rather modest in most instances, but the purpose of this study was not to identify the most important predictors, but rather to investigate the relationships between the latent variables and compare them with regard to gender. In a related vein, despite the commonly used term “causal model”, the survey design of our analyses (cross-sectional field study), together with the rather exploratory nature of the PLS method, do not warrant inferences about causality stricto sensu. Discussion While the results support most of the relations proposed by our model, they reveal some unexpected findings, too. As for the impact of family responsibilities on work centrality, the total coefficient shows the expected negative impact, which is strong for females but not significant for males. Obviously more duties and efforts for family affairs reduce working hours and energy left for the job. However, this does not apply to male young business professionals. Family structures affect managerial advancement of women and men in
Predicted Predictor Family responsibility Changeability Family responsibility £ changeability
Job alternatives Work centrality
Path coefficient (t-value): total/women/men Job alternatives Work centrality Objective career success (R 2 ¼ 0.001/ 0.002/ 0.000) 0.037 (0.65)/ 0.048 (0.54)/ 0.003 (0.04)
(R 2 ¼ 0.158/ 0.281/ 0.164) 20.133 (1.77)/ 2 0.236 (2.08)/ 2 0.041 (0.58) 0.257 (3.35)/ 0.290 (3.45)/ 0.239 (2.29) 20.166 (2.02)/ 0.202 (1.79)/ 2 0.243 (2.05) 0.175 (2.13)/ 0.157 (1.83)/ 0.273 (2.68)
Subjective career success
(R 2 ¼ 0.195/ 0.287/ 0.168)
(R 2 ¼ 0.160/ 0.157/ 0.226)
0.442 (6.43)/ 0.536 (5.52)/ 0.409 (5.21)
0.259 (2.53)/ 0.347 (1.88)/ 0.209 (1.96) 0.212 (2.45)/ 0.081 (0.64)/ 0.350 (3.43)
Objective career success
Responsibilities, career fields and gender
311
Table IV. Summary of the structural models
JMP 23,3
312
different ways. Whereas fathers’ careers benefit from their family role, mothers’ career advancement suffers from additional family responsibilities (e.g. Tharenou, 1999). Various reasons explain this disparity: traditional gender stereotypes and significant imparities in work compensation, career-orientation and work-addiction (Snir and Harpaz, 2006; Harpaz and Snir, 2003) lead women to assume the bigger share of family duties. Whereas successful male managers are often expected to be paterfamilias, ideal type female business professionals do not have family duties. Schoon et al. (2007) show that for women the effects of early childbearing bring more adverse outcomes regarding employment. These are possible causes for the small proportion in our sample having one child or more. Accordingly, married women work fewer hours per week than unmarried women, while married men work more hours per week than unmarried men (Snir and Harpaz, 2006) – often due to single breadwinner constellations. Contrary to the model, our results do not show any influence of family responsibilities on job alternatives. We assumed that a higher level of family responsibilities will reduce job alternatives for in most cases spatial mobility as well as temporal and income flexibility suffers from tight family bindings. Arguably, the characteristics of our sample are the major reason for the lack of support. Family and partnership obligations of young business professionals – if they exist at all – are quite recent. Since individuals’ assessments of mobility, flexibility and opportunities on labor markets are subject to time-consuming adaptive processes, time lags need to be integrated into our model: it takes some years of family experiences to realize that degrees of freedom and job alternatives diminish due to existing family responsibilities. On a side note, the impact of age on job alternatives might be even stronger than that of family status, and evidently both influences will interact although we did not control for age in our sample since there is little variance in this respect. Job alternatives positively affect work centrality. For males this effect appears significantly stronger than for females. Thus, our reasoning already presented can be further elaborated. Within the career fields of business professionals, at the very least, the assumption that tense labor markets and few job alternatives bring about longer working hours is not supported. Again, the specifics of our sample provide some explanations: young business professionals are hardly intimidated by unemployment and status loss because of their high education and academic titles which work as stable career capital independent from concrete performance. On the other hand, their professional identity is still under construction, a process accompanied by permanent search for external justification. The amount of job alternatives serves as justification, thus reinforcing work efforts. Gender differences show that women’s work centrality, which is generally lower not only in our sample (Mannheim, 1993; Mannheim and Dubin, 1986), is less supported by a favorable position on the labor market (though negative effects of unfulfilled job-expectations specifically strike women; Rindfuss et al., 1999). Though our model assumes a link from job alternatives to work centrality, causalities between these two constructs are somehow ambiguous. On the one hand a higher work centrality goes along with a higher level of commitment and attraction to the organization (Carlson and Kacmar, 2000), but on the other hand this might also lead to career vigilance and perception of a wider range of alternatives. If work is more important than family or leisure, actors will generally broaden their evoked set of job
alternatives. Organizations which do not focus on this kind of transition risk losing their employees. Changeability positively relates to work centrality and does so equally for men and women – there is no significant gender difference. The turbulence and dynamics of work content and professional relationships are career context factors strongly affecting employee attitudes and behavior, more specifically, their focus of attention (Gardner et al., 1989). From a different perspective this instrumentality is the backbone of HRM concepts (e.g. job rotation, job enrichment) and motivation theories (Fox and Feldman, 1988). As for the moderating effect of changeability on family responsibilities’ impact on work centrality, its strength, direction and gender-disparity is somehow counter-intuitive and surprising: . The negative coefficient for the total sample shows that an increase of changeability makes the impact of family responsibilities on work centrality even more negative. . The even stronger negative male coefficient indicates that for male business professionals who do not experience a negative impact of family responsibilities per se and under rather stable conditions, this negative effect will occur if changeability increases. . For women, it is the other way round: the moderating effect of changeability is positive. For them, conditions of higher changeability alleviate the negative effect of family responsibilities on work centrality. This is an interesting finding rarely analyzed hitherto: in more changeable situations family responsibilities obviously get more important for men, detracting energy from the work sphere; for women, changeability weakens the negative impact of family responsibility on work centrality. Thus one could argue that increasing dynamics of career fields contributes to gender matching as far as family responsibilities’ impact on work centrality is concerned. Various explanations might apply, depending on the focus on actors or structures. From an actor’s perspective, women seem to be better prepared for dealing with work changeability. Recent findings show that especially parental role commitment has direct positive effects on outcomes. Graves et al. (2007, p. 53) point out that the parental role may provide an even greater opportunity to develop skills that are transferable to managerial role. According to the enhancement theory of role accumulation, an increase of role expectations also provides sources of identity, self-esteem, rewards and resources available to cope with the multiple demands (Thoits, 1987). Women with multiple life roles (e.g. mother, wife, employee) are less depressed, have higher self-esteem and are more satisfied with private life and job compared to women and men who were not married, unemployed or childless (Baruch and Barnett, 1987; Crosby, 1991). Maybe post-organizational contexts thus offer more opportunities for women to find jobs with high temporal, social and task-flexibility as anchors for professional identification. From a structural perspective, growing turbulence in work environments detracts even women’s attention and energy from the family sphere. Under conditions of rapid change women cannot afford to mainly concentrate on family issues anymore without being in danger to loose touch with changing job and work requirements.
Responsibilities, career fields and gender 313
JMP 23,3
314
Work centrality is positively linked with both subjective and objective career success, as assumed in the model. Two aspects are interesting and somehow unexpected. First, the overall effect of work-centrality on objective is stronger than on subjective career success. Second, the effect is stronger for women than for men. For objective career success, prior research has revealed positive effects of time and energy devoted to work on income and hierarchical advancement (Mannheim et al., 1997; Rottenberry and Moberg, 2007). Compared with female, male work centrality is higher rewarded, i.e. better transformed into income and power, which once again points towards gender injustice in career fields for business graduates (Strunk et al., 2005). For subjective career success the positive impact of work centrality is based on dissonance theory (Festinger, 1959). Investments of time and energy into work contribute positively to career satisfaction and subjective impressions of being successful. The theory of self-justification (Aronson, 1992a, b) further helps to interpret gender differences, as the impact of work centrality on subjective career success is significantly stronger for women. As the female business professionals of our sample are discriminated in terms of objective success, they will stronger rely on internal justification of their investments into careers, i.e. on being satisfied. This theory also offers explanations for our last finding: objective career success positively contributes to subjective career success, i.e. satisfaction and subjective evaluation – but not for women. The assumed and obvious impact of income and hierarchical power on career satisfaction is true only for male business professionals. Obviously women get their career satisfaction from other sources then objective success – and one of them is work centrality, but keeping in mind that in our sample we are taking about younger women we should be aware of the fact that especially midlife women strongly desire to continue and also be perceived as valuable to their organization also in the term of objective career success indicators (Gordon and Whelan, 1998). Conclusions Overall, we find clear support for effects of family responsibilities on career success. Via work centrality, there is a negative relationship between family responsibilities and objective and subjective career success. This result is in line with previous research and supports current insight by demonstrating this link for an important sample, young business professionals. Two further main results enlarge previous findings. There is substantive support for the effect of contextual factors (here, career fields) for the relationship between family situations and career success. This points towards the importance of a multi-level perspective when analyzing work-family issues. Without taking contextual factors into account, analyses can fall short by missing important direct and indirect effects of context factors. Concepts such as career fields or career capitals help to bridge the different levels by providing conceptual tools for contextual effects. For example, career capital as the symbolic capital “valid” in career fields help to explain why the “same” portfolio of capital can have different effects under different contextual conditions. If the rules of the game temporarily or permanently change, e.g. by a different pace of change, available capitals can used in
different ways leading to more or less work centrality and, in turn, influencing career success. These kinds of insights on career bring along interesting perspectives for career guidance, policy and counseling. Finally, the results provide strong support for different effects of family responsibilities and career context on male and female careers. In prior research, three explanations for these differences have been offered and supported empirically (e.g. Kirchmeyer, 2006). (1) The choice explanation claims that early career decisions affect later success. Family roles constrain especially women’s careers opportunities by imposing geographical, temporal, or task related restrictions. (2) The performance explanation assumes that family roles and responsibilities differently affect the abilities of men and women to perform at work. (3) According to the signaling explanation, the career consequences of family stem neither from performance failures nor from choice restrictions, but from biases of others. The granting of promotions and financial rewards is not solely based on actual achievements, but also on signals of ability, future contributions, and involvement. For our sample, the choice explanation seems to be least likely as there is no relationship between family responsibilities and job alternatives. But we have to be aware that this is a sample which is primarily in early career stage, whereas mid- and late career women make different choices and have other models of success (Gordon and Whelan, 1998, p. 11). the performance explanation is supported by our data, since family responsibilities affect the work centrality of men and women differently due to changeability as part of the career context. Likewise, findings about the gendered relation between work centrality and the dimensions of career success strongly indicate the impact of signaling. The results of this paper also have a number of practical consequences. First, given the substantial efforts that some companies make in order to retain valued employees, the results underscore the important relationship between external factors such as family responsibilities and work related factors such as work centrality and career success. Hence, in order to use the existing work force most effectively, the work-family interface has to be focused in employee retention measures, too. Second, the results point towards the importance of tailored HR policies. For example, for men and women the effects of family responsibilities and the importance of this area clearly differ. Thus, HR measures not only have to take into account a gender perspective, but should also try to take into account different private settings of employees and the respective requirements for managing the work-private-interface. Third, the results also have consequences for the political sphere. They demonstrate the importance of the external environment – family as well as economy related – for individual career success in the objective and subjective sense. Political regulations governing the degree of compatibility between work and private life have a crucial impact for career related outcomes, too. On the whole, beyond the analysis of US-American academics (Kirchmeyer, 2006), the analysis shows gender specific family-influences on career outcomes for a further career field in the European context. The debate on the relevance of family responsibilities emerged from a focus on women, particularly women with dependent
Responsibilities, career fields and gender 315
JMP 23,3
316
children (Fleetwood, 2007), but our results show that the practices clearly affect men and researchers as well as practitioners need to be far more aware of the coherences in this field. Additionally, it reveals that the gender-specific impact of family responsibilities is significantly moderated by the dynamics of career fields. The results and interpretation emphasize the importance of qualitative in-depth studies to further clarify issues raised in this paper, for example the question whether women perceive the moderating effect of changeability positively – for example as an opportunity to enhance multiple role identities – or negatively – for example as structural pressure to invest more energy into work with the implicit threat of being hampered in their career. References Altonji, J.G. and Usui, E. (2007), “Work hours, wages, and vacation leave”, Industrial & Labor Relations Review, Vol. 60 No. 3, pp. 408-28. Aronson, E. (1992a), “The return of the repressed: dissonance theory makes a comeback”, Psychological Inquiry, Vol. 3 No. 4, pp. 303-11. Aronson, E. (1992b), The Social Animal, Freeman, New York, NY. Arthur, M.B. and Rousseau, D.M. (1996), “Introduction: the boundaryless career as a new employment principle”, in Arthur, M.B. and Rousseau, D.M. (Eds), The Boundaryless Career, Oxford University Press, New York, NY, pp. 3-20. Arthur, M.B., Khapova, S.N. and Wilderom, C.P.M. (2005), “Career success in a boundaryless career world”, Journal of Organizational Behavior, Vol. 26 No. 2, pp. 177-202. Baruch, Y. and Barnett, R.C. (1987), “Role quality and psychological well-being”, in Crosby, J.R. (Ed.), Spouse, Partner, Worker: On Gender and Multiple Roles, Yale University Press, New Haven, CT, pp. 63-73. Behson, S.J. (2002), “Which dominates? The relative importance of work-family organizational support and general organizational context on employee outcomes”, Journal of Vocational Behavior, Vol. 61, pp. 53-72. Bernhardt, A., Morris, M., Handcock, M.S. and Scott, M.A. (2001), Divergent Paths: Economic Mobility in the New American Labor Market, Russell Sage Foundation, New York, NY. Blossfeld, H.P. and Drobnic, S. (Eds) (2001), Career of Couples in Contemporary Society: From Male Breadwinner to Dual Earner Families, Oxford University Press, Oxford. Bo¨heim, R. and Taylor, M.P. (2004), “Actual and preferred working hours”, British Journal of Industrial Relations, Vol. 42 No. 1, pp. 149-66. Bollen, K. and Lennox, R. (1991), “Conventional wisdom on measurement: a structural equation perspective”, Psychological Bulletin, Vol. 110 No. 2, pp. 305-14. Bourdieu, P. (1977), Outline of a Theory of Practice, Cambridge University Press, Cambridge. Carlson, D.S. (1999), “Personality and role variables as predictors of three forms of work-family conflict”, Journal of Vocational Behavior, Vol. 55, pp. 236-53. Carlson, D.S. and Kacmar, K.M. (2000), “Work-family conflict in the organization: do life role values make a difference?”, Journal of Management, Vol. 26 No. 5, pp. 1031-54. Carlson, D.S. and Perrewe, P.L. (1999), “The role of social support in the stressor-strain relationship: an examination of work-family conflict”, Journal of Management, Vol. 25 No. 4, pp. 513-40.
Carmines, E.G. and Zeller, R.A. (1979), Reliability and Validity Assessment, Sage Publications, Beverly Hills, CA. Cassel, C., Hackl, P. and Westlund, A.H. (1999), “Robustness of partial least-squares method for estimating latent variable quality structures”, Journal of Applied Statistics, Vol. 26 No. 4, pp. 435-46. Chin, W.W. (1998), “The partial least squares approach to structural equation modelling”, in Marcoulides, G.A. (Ed.), Modern Methods for Business Research, Lawrence Erlbaum Associates, Mahwah, NJ, pp. 295-336. Chin, W.W. (2000), “Frequently asked questions – partial least squares & PLS-Graph”, available at: http://disc-nt.cba.uh.edu/chin/plsfaq.htm Chin, W.W., Marcolin, B.L. and Newsted, P.R. (1996), “A partial least squares latent variable modeling approach for measuring interaction effects: results from a Monte Carlo simulation study and voice mail emotion/adoption study”, paper presented at the 17th International Conference on Information Systems, Cleveland, OH. Cooper, C.L. and Williams, S. (1994), Creating Healthy Work Organizations, Wiley, Chichester. Crosby, J.R. (1991), Juggling: The Unexpected Advantages of Balancing Career and Home for Women and their Families, The Free Press, New York, NY. Day, A.L. and Chamberlain, T.C. (2006), “Committing to your work, spouse, and children: implications for work-family conflict”, Journal of Vocational Behavior, Vol. 68, pp. 116-30. Diefendorff, M.M., Brown, D.J., Kamin, A.M. and Lord, R.G. (2002), “Examining the roles of job involvement and work centrality in predicting organizational citizenship behaviors and job performance”, Journal of Organizational Behavior, Vol. 23 No. 1, pp. 93-108. Dif, M.H. (2004), “Vocational identities in change in telecommunications sector”, Career Development International, Vol. 9 No. 3, pp. 305-22. Doeringer, P.B. and Piore, M.J. (1971), Internal Labor Markets and Manpower Analysis, D.C. Heath and Company, London. Dubin, R., Champoux, J.E. and Porter, L.W. (1975), “Central life interests and organizational commitment of blue-collar and clerical workers”, Administrative Science Quarterly, Vol. 20, pp. 411-21. Eby, L., Casper, W., Lockwood, A., Bordeaux, C. and Brinley, A. (2005), “Work and family research in IO/OB: content analysis and review of the literature (1980-2002)”, Journal of Vocational Behavior, Vol. 66, pp. 124-97. Edwards, J.R. and Rothbard, N.P. (2000), “Mechanisms linking work and family: clarifying the relationship between work and family constructs”, Academy of Management Review, Vol. 25, pp. 178-99. Edwards, P. and Wajcman, J. (2005), The Politics of Working Life, Oxford University Press, Oxford. Edwards, R.C. (Ed.) (1975), Labor Market Segmentation, Lexington Books, Lexington, MA. Erler, M. (1996), Die Dynamik der modernen Familie. Empirische Untersuchung zum Wandel der Familienformen in Deutschland, Juventa, Weinheim. Esping-Andersen, G. (2000), Social Foundations of Postindustrial Economics, Oxford University Press, Oxford. European Foundation for the Improvement of Living and Working Conditions (2003), Work Time Preferences and Work-Life Balance in the EU: Some Policy Considerations for Enhancing for Quality of Life, European Foundation for the Improvement of Living and Working Conditions, Dublin.
Responsibilities, career fields and gender 317
JMP 23,3
318
Evans, J.A., Kunda, G. and Barley, S.R. (2004), “Beach time, bridge time, and billable hours: the temporal structure of technical contracting”, Administrative Science Quarterly, Vol. 49 No. 1, pp. 1-38. Festinger, L. (1959), A Theory of Cognitive Dissonance, Tavistock Publishers, London. Fleetwood, S. (2007), “Re-thinking work-life balance: editor’s introduction”, International Journal of Human Resource Management, Vol. 18 No. 3, pp. 351-9. Ford, M.T., Heinen, B.A. and Langkamer, K.T. (2007), “Work and family satisfaction and conflict: a meta-analysis of cross-domain relations”, Journal of Applied Psychology, Vol. 92 No. 1, pp. 57-80. Fornell, C. and Bookstein, F.L. (1982), “Two structural equation models: LISREL and PLS applied to consumer exit-voice theory”, Journal of Marketing Research, Vol. 19, pp. 440-52. Fornell, C. and Larcker, D.F. (1981), “Evaluating structural equation models with unobservable variables and measurement error”, Journal of Marketing Research, Vol. 18, pp. 39-50. Fox, M.L. and Dwyer, D.J. (1999), “An investigation of the effects of time and involvement in the relationship between stressors and work-family conflict”, Journal of Occupational Health Psychology, Vol. 4, pp. 164-74. Fox, S. and Feldman, G. (1988), “Attention state and critical psychological states as mediators between job dimensions and job outcomes”, Human Relations, Vol. 41 No. 3, pp. 229-45. Ganster, D.C. and Schaubroeck, J. (1991), “Work, stress and employee health”, Journal of Management, Vol. 17, pp. 235-71. Gardner, D.G., Cummings, L.L., Dunham, R.B. and Pierce, J.L. (1998), “Single-item versus multiple-item scales: an empirical comparison”, Educational and Psychological Measurement, Vol. 58 No. 6, pp. 898-915. Gardner, D.G., Dunham, R.B., Cummings, L.L. and Pierce, J.L. (1989), “Focus of attention at work: construct definition and empirical validation”, Journal of Occupational Psychology, Vol. 62 No. 1, pp. 61-77. Goetz, O. and Liehr-Gobbers, K. (2004), “Analyse von Strukturgleichungsmodellen mit Hilfe der Partial-Least-Squares (PLS)-Methode”, Die Betriebswirtschaft, Vol. 64 No. 6, pp. 714-38. Gordon, J.R. and Whelan, K.S. (1998), “Successful professional women in midlife: how organizations can more effectively understand and respond to the challenges”, Academy of Management Executive, Vol. 12 No. 1, pp. 8-24. Graves, L.M., Ohlott, P.J. and Ruderman, M.N. (2007), “Commitment to family roles: effects on managers’ attitudes and performance”, Journal of Applied Psychology, Vol. 92 No. 1, pp. 44-56. Greenhaus, J.H. (2003), “Career dynamics”, in Borman, W.C., Ilgen, D.R. and Klimoski, R.J. (Eds), Comprehensive Handbook of Psychology, Industrial and Organizational Psychology, Vol. 12, Wiley, New York, NY, pp. 519-40. Greenhaus, J.H. and Parasuraman, S. (1986), “A work-nonwork interactive perspective of stress and its consequences”, Journal of Organizational Behavior Management, Vol. 8, pp. 37-60. Greenhaus, J.H., Bedeian, A.G. and Moosholder, K.W. (1987), “Work experiences, job performance, and the feeling of personal and family well-being”, Journal of Vocational Behavior, Vol. 31, pp. 200-15. Greenhaus, J.H., Collins, K.M. and Shaw, J.D. (2003), “The relation between work-family balance and quality of life”, Journal of Vocational Behavior, Vol. 63, pp. 510-31. Greenhaus, J.H., Parasuraman, S. and Collins, K.M. (2001), “Career involvement and family involvement as moderators of relationships between work-family conflict and withdrawal from a profession”, Journal of Occupational Health Psychology, Vol. 6, pp. 91-100.
Grzywacz, J.G. and Marks, N.F. (2000), “Reconceptualizing the work-family interface: an ecological perspective on the correlates of positive and negative spillover between work and family”, Journal of Occupational Health Psychology, Vol. 5, pp. 111-26. Gunz, H.P. and Heslin, P.A. (2005), “Reconceptualizing career success”, Journal of Organizational Behavior, Vol. 26, pp. 105-11. Hakim, C. (2005), “Sex differences in work-life balance goals”, in Houston, D.M. (Ed.), Work-Life Balance in the 21st Century, Palgrave, New York, NY, pp. 55-79. Hall, D.T. (1990), “Promoting work/family balance: an organization-change approach”, Organization Dynamics, Vol. 18, pp. 5-18. Hall, D.T. (1996), “Protean careers of the 21st century”, Academy of Management Executive, Vol. 10, pp. 8-16. Harpaz, I. (1986), “The factorial structure of the meaning of working”, Human Relations, Vol. 29 No. 7, pp. 595-614. Harpaz, I. and Snir, R. (2003), “Workaholism: its definition and nature”, Human Relations, Vol. 56 No. 3, pp. 291-312. Harris, H., Brewster, C. and Erten, C. (2005), “Auslandseinsatz, aber wie? Klassisch oder alternative Formen: neueste empirische Erkenntnisse aus Europa und den USA”, in Stahl, G., Mayrhofer, W. and Ku¨hlmann, T.M. (Eds), Internationales Personalmanagement. Neue Aufgaben, neue Lo¨sungen, Hampp, Mu¨nchen, pp. 275-92. Herlth, A., Brunner, E., Tyrell, H. and Kriz, J. (1994), Abschied von der Normalfamilie, Partnerschaft contra Elternschaft, Berlin. Heslin, P.A. (2003), “Self- and other-referent criteria of career success”, Journal of Career Assessment, Vol. 11, pp. 262-82. Higgens, C., Duxbury, L. and Johnson, K.L. (2000), “Part-time work for women: does it really help balance work and family?”, Human Ressource Management, Vol. 39 No. 1, pp. 17-32. Holtzman, M. and Glass, J. (1999), “Explaining changes in mother’s job satisfaction following childbirth”, Work and Occupations, Vol. 26 No. 3, pp. 365-404. Hughes, E.C. (1937), “Institutional office and the person”, American Journal of Sociology, Vol. 43 No. 3, pp. 404-13. Hughes, E.C. (1951), “Ambition, mobility and sponsorship”, in Dubin, R. (Ed.), Human Relations in Administration, Prentice-Hall, New York, NY, pp. 240-3. Hulland, J. (1999), “Use of partial least squares (PLS) in strategic management research: a review of four recent studies”, Strategic Management Journal, Vol. 20, pp. 195-204. Iellatchitch, A., Mayrhofer, W. and Meyer, M. (2003), “Career fields: a small step towards a grand career theory?”, International Journal of Human Resource Management, Vol. 14 No. 5, pp. 728-50. Jacobs, J.A. and Gerson, K. (2001), “Overworked individuals or overworked families? Explaining trends in work, leisure, and family time”, Work and Occupations, Vol. 28 No. 1, pp. 40-63. Jarvis, C.B., MacKenzie, S.B. and Podsakoff, P.M. (2003), “A critical review of construct indicators and measurement model misspecification in marketing and consumer research”, Journal of Consumer Research, Vol. 30, pp. 199-218. Jennings, J.E. and McDougald, M.S. (2007), “Work-family interface experiences and coping strategies: implications for entrepreneurship research and practice”, Academy of Management Review, Vol. 32 No. 3, pp. 747-60. Judge, T.A., Cable, D.M., Boudreau, J.W. and Bretz, R.D.J. (1995), “An empirical investigation of the predictors of executive career success”, Personnel Psychology, Vol. 48 No. 3, pp. 485-507.
Responsibilities, career fields and gender 319
JMP 23,3
320
Kanungo, R.N. (1982), Work Alienation: An Integrative Approach, Praeger, New York, NY. Kaufmann, F.X. (1990), Zukunft der Familie: Stabilita¨t, Stabilita¨tsrisiken und Wandel der familialen Lebensformen sowie ihre gesellschaftlichen und politischen Bedingungen, Beck, Mu¨nchen. Keil, M., Tan, B.C.Y., Wei, K.-K., Saarinen, T., Tuunainen, V. and Wassenaar, A. (2000), “A cross-cultural study on escalation of commitment behavior in software projects”, MIS Quarterly, Vol. 24 No. 2, p. 299. King, L.A., Mattimore, L.K., King, D.W. and Adams, G.A. (1995), “Family Support Inventory for Workers: a new measure of perceived social support from family members”, Journal of Organizational Behavior, Vol. 16, pp. 235-58. Kirchmeyer, C. (2006), “The different effects of family on objective careers success across gender: a test of alternative explanations”, Journal of Vocational Behavior, Vol. 68, pp. 323-46. Korman, A.K., Wittig-Berman, U. and Lang, D. (1981), “Career success and personal failure: alienation in professionals and mangers”, Academy of Management Journal, Vol. 24 No. 2, pp. 342-60. Kossek, E. and Lambert, S. (2005), Work and Life Integration: Organizational, Cultural, and Individual Perspectives, Lawrence Erlbaum Associates, Mahwah, NJ. Leontaridi, R.M. (2002), “Career, experience and returns to human capital: is the dual labour market hypothesis relevant for the UK?”, Research in Economics, Vol. 56 No. 4, pp. 399-426. Lewis, S.N.C. and Cooper, C.L. (1987), “Stress in two-earner couples and stage in the life-cycle”, Journal of Occupational Psychology, Vol. 60, pp. 289-303. Lohmoeller, J.-B. (1989), Latent Variable Path Modelling with Partial Least Squares, Physica-Verlag, Heidelberg. Loo, R. (2002), “A caveat on using single-item versus multiple-item scales”, Journal of Managerial Psychology, Vol. 17 No. 1, pp. 68-75. Loscocco, K.A. (1997), “Work-family linkages among self-employed women and men”, Journal of Vocational Behavior, Vol. 50, pp. 204-26. McGovern, P., Stiles, P. and Hope, V. (1996), “The flexible psychological contract? Career management in an era of insecurity”, Management Research News, Vol. 19 No. 4, pp. 81-4. Mannheim, B. (1993), “Gender and the effects of demographics, status, and work values on work centrality”, Work and Occupations, Vol. 20 No. 1, pp. 3-22. Mannheim, B. and Dubin, R. (1986), “Work role centrality of industrial workers as related to organizational conditions, task autonomy, managerial orientations and personal characteristics”, Journal of Occupational Behavior, Vol. 7 No. 2, pp. 107-24. Mannheim, B., Baruch, Y. and Tal, J. (1997), “Alternative models for antecedents and outcomes of work centrality and job satisfaction of high-tech personnel”, Human Relations, Vol. 50 No. 12, pp. 1537-62. Marsden, D. (2004), “The ‘network economy’ and models of the employment contract”, British Journal of Industrial Relations, Vol. 42 No. 4, pp. 659-84. Mayrhofer, W., Meyer, M. and Steyrer, J. (2007), “Contextual issues in the study of careers”, in Gunz, H.P. and Peiperl, M.A. (Eds), Handbook of Career Studies, Sage Publications, London, pp. 215-39. Ng, T.W.H., Eby, L.T., Sorensen, K.L. and Feldman, D.C. (2005), “Predictors of objective and subjective career success: a meta-analysis”, Personnel Psychology, Vol. 58 No. 2, pp. 367-408.
Nicholson, N. and DeWaal-Andrews, W. (2005), “Playing to win: biological imperatives, self-regulation, and trade-offs in the game of career success”, Journal of Organizational Behavior, Vol. 26 No. 2, pp. 137-54. Nunnally, J.C. (1978), Psychometric Theory, 2nd ed., McGraw-Hill, New York, NY. O’Hara, K.B., Beehr, T.A. and Colarelli, S.M. (1994), “Organizational centrality: third dimension of intraorganizational career movement”, Journal of Applied Behavioral Science, Vol. 30 No. 2, pp. 198-216. Osherson, S. and Dill, P. (1983), “Varying work and family choices: their impact on men’s work satisfaction”, Journal of Marriage and the Family, Vol. 45, pp. 339-46. Osterman, P. (1995), “Work/family programs and the employment relationship”, Administrative Science Quarterly, Vol. 40, pp. 681-700. Parasuraman, S. and Simmers, C.A. (2001), “Type of employment, work-family conflict and well-being: a comparative study”, Journal of Organizational Behaviour, Vol. 22, pp. 551-68. Paullay, I.M., Alliger, G.M. and Stone-Romero, E.F. (1994), “Construct validation of two instruments designed to measure job involvement and work centrality”, Journal of Applied Psychology, Vol. 79 No. 2, pp. 224-8. Perry-Smith, J.E. and Blum, T.C. (2000), “Work-family human resource bundles and perceived organizational performance”, Academy of Management Journal, Vol. 43 No. 6, pp. 1107-17. Peterson, M.F. and Ruiz-Quintanilla, S.A. (2003), “Cultural socialization as a source of intrinsic work motivation”, Group & Organization Management, Vol. 28 No. 2, pp. 188-216. Peuckert, R. (1996), Familienformen im sozialen Wandel, Leske & Budrich, Opladen. Piore, M.J. (1975), “Notes for a theory of labor market stratification”, in Edwards, R.C., Reich, M. and Gordon, D.M. (Eds), Labor Market Segmentation, Heath, Lexington, MA, pp. 125-50. Pleck, J.H. (1985), Working Wives/Working Husbands, Sage Publications, Newbury Park, CA. Podsakoff, P.M. and Organ, D.W. (1986), “Self-reports in organizational research: problems and prospects”, Journal of Management, Vol. 12 No. 4, pp. 531-44. Poelmans, S. and Sahibzada, K. (2004), “A multi-level model for studying the context and impact of work-family politics and culture in organizations”, Human Resource Management Review, Vol. 14, pp. 409-31. Pryor, R.G.L. and Davies, R. (1989), “A comparison of conceptualizations of work centrality”, Journal of Organizational Behavior, Vol. 10 No. 3, pp. 283-9. Rindfuss, R.R., Cooksey, E.C. and Sutterlin, R.L. (1999), “Young adult occupational achievement: early expectations versus behavioral reality”, Work and Occupations, Vol. 26 No. 2, pp. 220-63. Ringle, C.M., Wende, S. and Will, A. (2005), SmartPLS 2.0 (beta), University of Hamburg, Hamburg. Robins, R.W., Hendin, H.M. and Trzesniewski, K.H. (2001), “Measuring global self-esteem: construct validation of a single-item measure and the Rosenberg Self-Esteem Scale”, Personality and Social Psychology Bulletin, Vol. 27 No. 2, pp. 151-61. Rottenberry, P.F. and Moberg, P.J. (2007), “Assessing the impact of job involvement on performance”, Management Research News, Vol. 30 No. 3, pp. 203-15. Sanchez-Franco, M.J. (2006), “Exploring the influence of gender on the web usage via partial least squares”, Behaviour & Information Technology, Vol. 25 No. 1, pp. 19-36. Schmidt, A. (2001), Familie und Organisation: Systeme als Widerpart?, Peter Lang, Frankfurt. Schneer, J.A. and Reitman, F. (1993), “Effects of alternative family structures on managerial careers”, Academy of Management Journal, Vol. 38 No. 4, pp. 830-43.
Responsibilities, career fields and gender 321
JMP 23,3
322
Schneer, J.A. and Reitman, F. (2002), “Managerial life without a wife: family structure and managerial career success”, Journal of Business Ethics, Vol. 37 No. 1, pp. 25-38. Schneider, N.F. (2005), “Leben an zwei Orten. Die Folgen beruflicher Mobilita¨t fu¨r Familie und Partnerschaft” (“Job mobility and living arrangements”), in Mischau, A. and Oechsle, M. (Eds), Arbeitszeit-Familienzeit-Lebenszeit: Verlieren wir die Balance?, VS Verlag fu¨r Sozialwissenschaften, Wiesbaden, pp. 110-26. Schoon, I., Martin, P. and Ross, A. (2007), “Career transistions in times of social change. His and her story” Journal of Vocational Behavior, Vol. 70, pp. 78-96. Senge, P.M. (1990), The Fifth Discipline, Currency Doubleday, New York, NY. Snir, R. and Harpaz, I. (2006), “The workaholism phenomenon: a cross-national perspective”, Career Development International, Vol. 11 No. 5, pp. 374-93. Stoner, C.R., Hartman, R.I. and Arora, R. (1990), “Work/family conflict: a study of women in management”, Journal of Applied Business Research, Vol. 7 No. 1, pp. 67-75. Stroh, L.K., Brett, J.M. and Reilly, A.H. (1996), “Family structure, glass ceiling, and traditional explanations for the differential rate of turnover of female and male managers”, Journal of Vocational Behavior, Vol. 49, pp. 99-118. Strunk, G., Hermann, A. and Praschak, S. (2005), “Eine Frau muss ein Mann sein, um Karriere zu machen”, in Mayrhofer, W., Meyer, M. and Steyrer, J. (Eds), Macht? Erfolg? Reich? Glu¨cklich? Einflussfaktoren auf Karrieren, Linde, Wien, pp. 211-42. Sturges, J. (1999), “What it means to succeed: personal conceptions of career success held by male and female managers at different ages”, British Journal of Management, Vol. 10, pp. 239-52. Tenenhaus, M., Vinzi, V.E., Chatelin, Y.-M. and Lauro, C. (2005), “PLS path modeling”, Computational Statistics & Data Analysis, Vol. 48, pp. 159-205. Tharenou, P. (1999), “Is there a link between family structures and women’s and men’s managerial career advancement?”, Journal of Organizational Behavior, Vol. 20, pp. 837-63. Tharenou, P., Latimer, S. and Conroy, D. (1994), “How do you make it to the top? An examination of influences on women’s and men’s managerial advancement”, Academy of Management Journal, Vol. 37 No. 4, pp. 899-931. Theodossiou, I. (1995), “Wage determination for career and non-career workers in the UK: is there labour market segmentation?”, Economica, Vol. 62 No. 246, pp. 195-211. Thoits, P.A. (1987), Negotiating Roles, Spouse, Partner, Worker: On Gender and Multiple Roles, Yale University Press, New Haven, CT, pp. 11-22. Tolbert, C.M.I. (1982), “Industrial segmentation and men’s career mobility”, American Sociological Review, Vol. 47, August, pp. 457-77. Tolbert, P.S. (1996) “Occupations, organizations, and boundaryless careers”, in Arthur, M.B. and Rousseau, D.B. (Eds), The Boundaryless Career: A New Employment Principle for a New Organizational Era, Oxford University Press, New York, NY, pp. 331-49. van Ours, J.C. (2004), “The locking-in effect of subsidized jobs”, Journal of Comparative Economics, Vol. 32 No. 1, pp. 37-55. Voydanoff, P. (1988), “Work-role characteristics, family structure demands, and work-family conflict”, Journal of Marriage and the Family, Vol. 50, pp. 749-61. Wall, T.D., Michie, J., Patterson, M., Wood, S.J., Sheehan, M., Clegg, C.W. and West, M. (2004), “On the validity of subjective measures of company performance”, Personnel Psychology, Vol. 57 No. 1, pp. 95-119.
Williams, K. and Alliger, G.M. (1994), “Role stressors, mood spillover, and perceptions of work-family conflict in employed parents”, Academy of Management Journal, Vol. 37, pp. 837-68. Wold, H. (1975), “Path models with latent variables: the NIPALS approach”, in Blalock, H.M., Aganbegian, A., Borodkin, F.M., Boudon, R. and Capecchi, V. (Eds), Quantitative Sociology: International Perspectives on Mathematical and Statistical Modeling, Academic Press, New York, NY, pp. 307-57. Corresponding author Wolfgang Mayrhofer can be contacted at:
[email protected]
To purchase reprints of this article please e-mail:
[email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints
Responsibilities, career fields and gender 323
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0268-3946.htm
JMP 23,3
The expatriate family: an international perspective
324
Webster University Vienna, Vienna, Austria, and Ashridge Business School, Berkhamsted, UK, and
Arno Haslberger Chris Brewster University of Reading Business School, Reading, UK Abstract Purpose – This paper seeks to review and explore the relatively neglected notion of the adjustment of expatriate families to living abroad with the aim of developing a new model that can be used for future research. Design/methodology/approach – The paper draws on the few studies of the topic that have been carried out, but widens the search to include evidence from the related adjustment and family stress literature to create a new model of the process. Using the ideas of stressors, strains and hassles, capabilities, and shared meanings, the paper examines the situation of the expatriate family and explores how families can adjust to life in another country. Findings – By adopting a salutogenic approach and incorporating insights from these other literatures, the paper shows that family adaptation is a complex and many-faceted process. It is a process that greater awareness on the part of the family and the organization can improve. Research limitations/implications – With the help of the model of family adjustment the paper points to systematic gaps in studies on expatriate families and outlines a consequent research agenda. Practical implications – Awareness is a crucial element in adjustment. The paper shows that awareness by the family can alleviate problems, and that organizations employing members of the family can assist in the adjustment process for the family. Originality/value – The contribution of the paper comes in its attempt to encompass what is known about expatriate family adaptation directly with a wider view of family adjustment. This provides both a practical framework for future research and some practical implications. Keywords Expatriates, Family, Cross-cultural studies, Sociology of work Paper type Research paper
Journal of Managerial Psychology Vol. 23 No. 3, 2008 pp. 324-346 q Emerald Group Publishing Limited 0268-3946 DOI 10.1108/02683940810861400
Introduction The number of expatriate assignments continues to rise. Recent studies by consultancies (GMAC, 2006; Mercer Human Resource Consulting, 2006) show substantial continuing growth in the numbers of people sent abroad by their organizations. The consultancies report that about 60 per cent of expatriates are accompanied by a spouse or partner, and about half are accompanied by children. Before the assignment 60 per cent of partners were employed, but only 21 per cent worked abroad. Historical averages going back to 1993 are of similar magnitude (GMAC, 2006). Academic research (Dickmann et al., 2006) confirms the findings, with three-quarters of a UK sample of expatriates being accompanied by at least one family member[1]. These numbers attest to the continuing importance of family issues and the work-family interface in expatriate assignments. In spite of this, research on expatriate
assignments has paid only comparatively little attention to expatriate families, focusing instead on the adjustment of individuals. This paper attempts to outline the theoretical elements of expatriate families’ adjustment to living abroad. In particular, we: . develop a model of expatriate family adjustment based on family stress theory (Patterson, 1988, 2002a) by focusing on demands, capabilities and shared meanings in families on assignment; . explore the notion of crossover among family members and discuss variables that influence the crossover process; and . show areas of systematic omission in expatriate family studies. The history of research on expatriate family adjustment Although the literature on expatriate families is scant, it contains the seeds of a theoretical approach to the topic. Cleveland et al. (1960) were among the first authors to include a family perspective in a book on living abroad. Nash (1967, 1969) included the “domestic side of a foreign existence” as part of his study of American expatriates in Spain in the early 1960s. Hays (1971, 1974) investigated factors playing a role in success and failure of expatriate employees. To the variables commonly studied at the time he added preparation and support by the company, language skills, and the adaptability and supportiveness of the family among others. Cohen’s (1977) extensive review of expatriate communities had a few words to say about the expatriate family. He pointed out that expatriate communities tend to be male-centered, that while the husband’s working life is “continuous” the wife bears the burden of transferring family life, and that those women who had worked at home but had to stop while abroad faced the most difficult situation. These authors did not provide fully-fledged analyses of expatriate families’ adjustment. Rather, their contribution was restricted to recognizing that the accompanying partner or family faced a situation that was distinctly different from that of the expatriate employee and that warrants separate attention. At the beginning of the 1980s, Tung published some influential articles (Tung, 1981, 1982, 1984), in which she identified the family as a critical success factor in expatriate assignments. One of her articles pointed to “the inability of the manager’s spouse to adjust” as the major problem for US and European multinational companies (Tung, 1982). To this day this statement plays a role in academic and professional publications on the topic. Harvey (1985) pointed out that, despite this, research on expatriate assignments did not pay much attention to families. He discussed issues in selection and adjustment abroad, including the partner’s and children’s perspectives. His and Tung’s contributions could be called “expatriate-centric”, because spillover from the family into the work domain is the reason for the family focus. Black and colleagues (Black and Stephens, 1989; Black and Gregersen, 1991a, b; Black et al., 1991; Stephens and Black, 1991) looked at partner adjustment as a separate area for research. In addition, they investigated the crossover between partner and expatriate adjustment. Forster (1992) offered a theoretical contribution in answer to Black et al. (1991) that gave the partner and family more prominence in the overall model. Yet the model itself remained rather unclear. Fukuda and Chu (1994) added a Japanese perspective to the research on family variables in expatriate employee adjustment. Their study was
The expatriate family
325
JMP 23,3
326
expatriate-centric, but was one of the few to include respondents outside of the English-speaking world. Caligiuri et al. (1998) referred to family systems theory when they analyzed spillover between family and work adjustment of expatriate employees. This was an innovation in expatriate studies. Partner and family variables continue to feature in expatriate research (Shaffer et al., 1999, 2000; Kraimer et al., 2001). Takeuchi et al. (2002) explicitly looked at both crossover and spillover effects among expatriate couples for the first time, and others followed (van der Zee et al., 2005). Recently, initial attempts have been made to integrate theoretical knowledge from family studies into the theory of expatriate adjustment (Westman et al., 2006). In order to lay the groundwork for the discussion of expatriate family adjustment it is necessary to introduce relevant concepts from both family and expatriate studies. Overview of relevant concepts Adjustment and adaptation Adjustment, adaptation, and acculturation are often used interchangeably in the literature on expatriate adjustment to mean the process and result of change induced in individuals by the move into an unfamiliar cultural environment (see, for example, Schu¨tz, 1944; Taft, 1977; Church, 1982; Grove and Torbio¨rn, 1985; Kim, 1988; Kim and Gudykunst, 1988; Black and Mendenhall, 1990, 1991; Aycan, 1997; Ward et al., 1998; Evans et al., 2002; Yamazaki and Kayes, 2004; Bhaskar-Shrinivas et al., 2005). The literature on family systems distinguishes between adjustment, resulting in minor changes to cope with new situations, and adaptation, which indicates large-scale change and major realignment following a serious crisis (Patterson, 1988, 2002a, b). Patterson (1988) mentions relocations as an example of a stressor that might lead to crisis. Our focus here is on “adjustment”, the process of coping with a new situation (expatriation), which is expected to be temporary, and we will use this term as a default throughout this article. Cross-cultural adjustment of individuals encompasses cognitions, emotions, and behaviors (Kim, 1988, 1991; Ward et al., 2001). Defined as a state (Berry et al., 1988), adjustment is the degree of fit between individual and environment regarding social processes and social structures (Gudykunst and Hammer, 1988). A good fit is defined as the adequacy of one’s behaviors, cognitions and emotions (Grove and Torbio¨rn, 1985). Defined as a process (Berry et al., 1988), adjustment is the acculturation of the newcomer, or the convergence (Barnett and Kincaid, 1983; Kincaid, 1988) over time of behaviors, values and norms, and underlying assumptions, of the individual with those prevailing in the environment (Schein, 1984; Black et al., 1992; Trompenaars, 1993). Successful individual adjustment is the integration of the elements of the individual’s original culture and the new one (Kim, 1988) while preserving the individual’s cultural identity overall. If there is a profound change in cultural identity, we speak of adaptation. Expatriate assignments of up to three years in duration generally lead to adjustment only. The longer an assignment, the more profound changes in cultural identity are likely to be. Adjustment outcomes are psychological well-being (Searle and Ward, 1990; Ward and Kennedy, 1999; Zimmermann et al., 2003) and a restructuring of a person’s mental frame of reference (Schu¨tz, 1944) through learning processes (Black and Mendenhall, 1991; Zimmermann et al., 2003). Successful family adjustment is the
preservation of its pre-assignment functionality and that of its members. When a major realignment of the family setup is required, we speak of adaptation. Domains of adjustment People play different roles in different domains. A high-flying expatriate executive at work may also be just one of a number of members of the parent-teacher organization while juggling other family responsibilities on top. Navas et al. (2005, 2007) distinguish six domains relevant in the acculturation of immigrants: (1) politics and government, i.e. systems of public order; (2) work; (3) economic, including consumption of goods and services; (4) family relations; (5) social relations; and (6) ideology, which includes two sub-divisions, i.e. ways of thinking, principles and values on the one side, and religious beliefs and customs on the other.
The expatriate family
327
The expatriate literature includes only some of the domains listed by Navas and colleagues. Following Black’s (1988) distinction, many studies have focused on the facets of work, interaction and adjustment to general living conditions (see Hechanova et al., 2003 and Bhaskar-Shrinivas et al., 2005 for lists of studies). Expatriates have to adjust to each of the domains separately. Of course the domains are not isolated or independent from each other. Therefore, spillover takes place from one domain into another. Work-family conflict is an intensively studied area in organizational behavior and has been shown to exist for expatriates as well (Bhaskar-Shrinivas et al., 2005). Besides spillover among domains, crossover may occur between individuals. This crossover can affect the same domain in both individuals, for example the partner’s adjustment to social relations may influence the expatriate employee’s adjustment to social relations. It can also cut across domains, for example the expatriate’s work adjustment impacting the partner’s adjustment to family life (Parasuraman et al., 1992; Westman, 2001). Any of the cells in Figure 1 might be connected by a one- or bi-directional arrow. A vertical arrow indicates spillover and a horizontal one crossover. Dyads and systems Expatriate adjustment studies have uncovered significant crossover effects among partners (Black and Stephens, 1989; Black and Gregersen, 1991c; Shaffer et al., 1999,
Figure 1. Crossover and spillover
JMP 23,3
328
2000; Kraimer et al., 2001; Takeuchi et al., 2002; van der Zee et al., 2005). They looked at individuals’ adjustment and the mutual influences of partners on each other, i.e. the dynamics of dyads. Since, as shown, half the expatriates who have family members with them have children as well as partners, the expatriate multi-person family as a system needs to be studied too. The integration of ideas from family systems theory (Patterson, 1988, 2002a) may bring additional insights into the dynamics of expatriate adjustment. The Family Adjustment and Adaptation Response (FAAR) model was built on the ABCX model, which was originally developed in the 1940s (Patterson, 1988). The latter states that a stressor (A) interacts with the family’s abilities to meet it (B) and with the family’s attribution of meaning to the stressor (C) to produce a crisis (X). Patterson and her colleagues then developed the ABCX model further, eventually resulting in the FAAR model (Patterson, 1988). The Family Adjustment and Adaptation Response model In a nutshell, the FAAR model looks at the outcomes of a balancing process between demands on the family and its capabilities to cope with the demands. The process is mediated and moderated by the meanings the family attaches to its current and general situation. If the family’s capabilities suffice to answer the demands, adjustment takes place. If they are not enough, demands pile up and the family eventually slides into a crisis. In the cross-cultural literature, elements of this crisis are referred to as culture shock (Oberg, 1960; Lundstedt, 1963; Gullahorn and Gullahorn, 1963; Smalley, 1963). Recovery from the crisis depends on the family’s adaptive capabilities (Patterson, 1988, 2002a, b). Demands. Stress theory distinguishes three main sources of demands (Ruffin, 1993; Patterson, 2002a): (1) stressors, i.e. discrete events such as a move abroad; (2) strains, i.e. ongoing unresolved tensions as a result of a stressor or of not meeting demands; and (3) daily hassles, i.e. troubles with neighbors, traffic problems, or weather experienced as unpleasant. The literature on expatriation identifies demands such as work demands or cultural differences, but tends not to classify them into stressors, strains and daily hassles. Capabilities: resources (Patterson, 1988; Feinberg, 2003). Demands trigger the application of resources and coping behaviors. Resources include those of individual family members such as knowledge, skills, personality traits, emotional and physical health, sense of mastery, and self esteem. A second source of resources is the family itself, i.e. its cohesion, adaptability, organization and structure, and its communication skills. The third set of resources is community-based, including social networks and the support from them. Support may be emotional, informational, or instrumental. Capabilities: coping behaviors. Families and individual members try to cope with demands by either reducing them or managing their effects in order to restore a balance. Patterson (1988) lists five ways families do this: (1) direct action to reduce demands in number and/or intensity; (2) direct action to preserve existing resources;
(3) direct action to get hold of additional resources; (4) families may also attempt to manage the tension that comes from demands, e.g. by playing or joking together; and (5) families may reappraise the situation to change its meaning. Meanings. The balancing of demands against resources and coping behaviors is influenced by the meanings the family attributes to its current situation, to itself as an entity, i.e. the family’s identity, and to the world at large, i.e. its environment and the other systems it interacts with (Patterson, 2002a). The meaning of the current situation stems from the interpretation of current demands and current capabilities. This is an immediate and shorter-term attribution. It is rooted in longer-term meanings, which include the commitment of members to a common purpose, goals and values, the affective climate or optimism of the family as a unit, the extent to which the family sees itself and its circumstances as changeable or fixed, whether it sees itself as isolated or integrated into a larger whole, and, finally, to what extent it is willing to share control and trust those outside the family. Families are continuously engaged in the balancing of demands against capabilities. Adjustment is a process; sometimes a pile-up of demands overwhelms capabilities, resulting in a crisis and a subsequent need to adapt. Failure to adapt would result in dissolution of the family. The FAAR model provides a salutogenic instead of a pathogenic approach (Antonovsky, 1987). Rather than focusing on deficits in dealing with challenges, it emphasizes capabilities that make families more resilient (Ho and Keiley, 2003). In this it is in agreement with a long-held view in expatriate studies that asks for a similar reorientation (Adler, 1975, 1976, 1987; Ellingsworth, 1988; Furnham, 1988; Kim and Ruben, 1988; Glanz et al., 2001). Elements of a family perspective in expatriation Having introduced some key concepts we now turn to discussing some of their implications as we focus on the elements of dyadic and family crossover in expatriate adjustment. Crossover among individuals may involve all types of demands. A straightforward crossover may be the stressors of one partner creating stressors for the other partner, or the strains experienced by one causing strain in the other (Bolger et al., 1989; van der Zee et al., 2005). The crossover may also be more complicated in that the stressors of one partner produce strain in the other, and vice versa, i.e. one partner’s strain results in a stressor for the other. Similar dynamics are likely to exist regarding daily hassles (Westman and Etzion, 1995, discuss these relationships, though using a different terminology). Crossover is not limited to demands. It includes capabilities, too (Westman, 2001). Resources may be shared and effective coping behaviors may be passed on among family members. The impact of positive coping behaviors of one member will be felt by the others as well. Crossover can have a positive or negative direction. Demands may reinforce each other, leading to greater demands as a result and vice versa. Alternatively, some demands may actually reduce others. One study found that cognitive demands at home and at work were negatively related to burnout and positively to work engagement (Bakker et al., 2005). While this is a spillover phenomenon, there is good reason to
The expatriate family
329
JMP 23,3
330
assume crossover as well since the cognitive demands of home are influenced to some extent by all family members. The challenges of learning a new culture and, possibly, a new language can be interpreted as cognitive demands for expatriates. As long as the capabilities of the individuals and the family are sufficient, an international move may have a positive effect. The so-called “honeymoon” phase (Oberg, 1960) in expatriation may be related to the positive effect of demands. Only as demands pile up over time and exceed the adjustment capabilities of the expatriate family does crisis result. The potential positive effect of demands is also recognized in stress theory via the recognition of qualitative and quantitative “underload” as work stressors (Greenberg and Baron, 2003). Expatriate partners who give up their jobs may eventually experience boredom and underload in addition to a loss of professional status. They may feel overload at first – regarding culture, language and setting up a new family life – followed by underload in terms of having too much time on their hands. Like demands, capabilities may also have a reinforcing or weakening effect. In most cases support and effective coping will further strengthen a family’s capabilities. Sometimes, however, this does not hold true. One study found that family support related to a partner’s work problems increased rather than decreased burnout (Westman and Etzion, 1995). The authors emphasized that support has to be relevant for the situation to create the desired results. This raises some interesting questions about support provided by companies to expatriates and their families. The wrong type of support may aggravate rather than alleviate problems. We will return to this issue later. Crossover may occur as a result of conscious processing of information, as Westman (2001) points out referring to social learning theory. Partners empathize by imagining how they would feel in the other’s place. Alternatively, crossover may be automatic: an involuntary synchronization of feelings called emotional contagion (Goleman, 1995; van der Zee et al., 2005). Goleman (1995) cites research showing that even strangers synchronize their moods while just sitting next to each other and waiting. Not everything that appears as such is actually crossover (Westman and Etzion, 1995; Westman, 2001). If a demand on one individual causes a reaction in another individual, we can speak of crossover. Alternatively, both individuals may be affected by a common demand causing independent reactions, which might otherwise appear to be crossover. One necessary condition to speak of crossover is a time lag between individuals’ reactions. But a lagged reaction is not sufficient to show crossover. Individuals may have realized the demand at different times or may show reactions at differing speeds. Demands on expatriate families A move abroad brings additional demands on families. The literature about expatriate assignments is replete with variables that constitute demands. Many influences have been studied, from food and weather through cultural values to determinants of the expatriate’s job. The FAAR model’s typology enables the exploration of these influences through a different lens. Stressors. Stressors are distinct events (Patterson, 1988). The underlying stressor is the expatriate’s assignment to a new job. It triggers the main event for the family, which is the move abroad. It coincides with a series of (possible) further stressors such as the partner giving up a job, children going to a new school, occupying a new residence, changing family routines, cultural differences between and differing gender
roles in home and host country, etc. One of the higher-rated items on most stress scales is a change in financial status. Every expatriation brings with it financial uncertainties. The expatriate’s income is nearly always higher while on assignment and for many expatriates many of their living expenses are covered by the company; on the other hand, the loss of a second income affects about 40 per cent of expatriate families (GMAC, 2006). These life events are common stressors for the family members (Westman, 2001). Using the Holmes and Rahe Social Readjustment Scale (Cherrington, 1994) an international move will almost certainly involve enough change to raise the probability of stress-induced health problems to about 50 per cent; an international transfer may even bring it up to 80 per cent[2]. An innovative longitudinal study of expatriate managers that included medical examinations and the measurement of stress hormones in addition to social and psychological variables showed that a quarter of participants suffered negative physiological responses (Anderze´n and Arnetz, 1997). Strains. Strains represent ongoing tensions (Patterson, 1988). They are related to a feeling that something needs to change. Strains may result from unresolved stressor events. A cross-cultural move brings with it numerous stressors, many of which take a while to deal with. Strains manifest themselves over time. When a family arrives in a new location, it encounters most of the stressors related to the move right from the start. Yet the tensions that they create may grow slowly. Expatriates, who experience the classic stages of culture shock (Oberg, 1960), start to feel the strains of the move only after a time lag. Expatriates, who have enjoyed preparatory training and have worked on anticipatory adjustment, on the other hand, may already feel some of the strains upon arrival. This would explain the “J-curve” pattern observed by some researchers (Black and Mendenhall, 1991; Ward et al., 1998). It might also explain why previous expatriate experience shows mixed influence on adjustment (Black and Stephens, 1989; Black and Gregersen, 1991b, c; Nicholson and Imaizumi, 1993; Taylor and Napier, 1996). As long as there is an unresolved tension from an earlier move, where adjustment or adaptation was not entirely successful, a strain may be carried over into the start of a new assignment. Another source of strain comes from not meeting expectations in a role (Patterson, 1988). An expatriate employee or his or her superiors may be unhappy with work performance. Family members may be disappointed with themselves or other family members about how they are doing in their various new roles abroad. Anxiety or depression may result from this. Research on families confirms that anxiety and depression impact other family members (Klever, 2001; Leinonen et al., 2003). There are significant differences in how anxiety and depression in fathers and mothers impact sons and daughters. This is mediated by the quality of partner interaction and parenting style, which are related to anxiety and depression themselves (Leinonen et al., 2003). These dynamics will affect expatriate families as well. Finally, strains may result as outcomes from an expatriate family’s unsuccessful attempts at adjustment or adaptation. Daily hassles. Daily hassles may be more important in predicting stress and other negative outcomes than major life events (DeLongis et al., 1988; Patterson, 1988; Reich et al., 1988; Braun, 1989; Ruffin, 1993; Wu and Lam, 1993). Potential daily hassles as measured by the Hassles and Uplifts Scale (DeLongis et al., 1988) include child(ren), partner, time spent with family, health or well-being of a family member, family-related obligations, fellow workers, supervisor or employer, the nature of the work, workload,
The expatriate family
331
JMP 23,3
332
meeting deadlines or goals on the job, money, medical care, the weather, environment (e.g. quality of air, noise level, greenery), and neighborhood (e.g. neighbors, physical setting), amount of free time, recreation and entertainment outside the home. A move abroad will change nearly all of these: this carries the risk that many of these potential hassles will become salient. The hassles cut across domains to include cultural and job items. Many expatriate studies have found that various work role variables such as role clarity and role conflict are associated with adjustment outcomes (Bhaskar-Shrinivas et al., 2005). These variables are likely to contribute to expatriate employees experiencing daily hassles. Family capabilities Resources. Many expatriate studies have included personal resources and found them to be related to adjustment outcomes (e.g. Black, 1990; Forster, 1992; Harrison et al., 1996; Anderze´n and Arnetz, 1997; Aycan, 1997; Shaffer et al., 1999; Caligiuri, 2000a, b; Glanz et al., 2001; van Vianen et al., 2004; Waxin, 2004; Yamazaki and Kayes, 2004; Holopainen and Bjo¨rkman, 2005). Personal resources of the expatriate feature prominently in the literature on selection criteria used in choosing individuals for assignments (even if these are often not reflected in the actual decision process; see Harris and Brewster, 1999). Besides technical and managerial skills and knowledge, companies sometimes use psychological instruments to assess characteristics known to be related to adjustment (Sparrow, 1999; Anderson, 2005). Many companies are hesitant to involve family members in the selection process (Schell and Solomon, 1997; Franke and Nicholson, 2002; Anderson, 2005). Therefore, the personal resources of family members and family resources are not sufficiently taken into account when making selection decisions. Preparatory and on-site training programs are another way companies can ensure that expatriates possess enough resources for dealing with the demands of the move. There are many different ways in which training is provided (Mendenhall and Oddou, 1986; Brewster and Harris, 1999). The many studies of preparatory training generally support its effectiveness (Earley, 1987; Kealey and Protheroe, 1996; Waxin and Panaccio, 2005). Still, only a quarter of companies offer cross-cultural training for all assignments, a fifth offer none at all. Those who do offer training include family members in most cases (GMAC, 2006). Expatriate studies pay little attention to the family itself as a resource, i.e. its cohesion, adaptability, organization and structure, and its communication skills, although there are some exceptions (Nash, 1969; Caligiuri et al., 1998; Shaffer et al., 2000). Community resources, the third group of family capabilities, tend to be geographically stationary. Therefore, families that move from one country to another must rebuild their support, find new networks and cope with fewer community resources in the short and long term. In the short-term, the type of community resource most easily replaced is instrumental support. Many company relocation policies include a relocation allowance and help with the move and with work as well as residency permits. Fewer companies provide assistance in finding a job for the accompanying partner, which is part of informational support. Expatriate networks and organizations in the host location often fill the gap resulting from leaving behind professional and private informational support. Emotional support can be found in such places as religious bodies, sports and social clubs and friendships. Some of these
can take a long time to develop into meaningful support. This is the element of community resources that is most difficult to replace in a planned fashion and may take a longer time to build back up. Social support variables feature prominently in many expatriate studies (Nash, 1967; David, 1972; Harvey, 1982; Tung, 1987; Berry et al., 1988; Black et al., 1991, 1992; Black and Gregersen, 1991a; Forster, 1992; Anderze´n and Arnetz, 1997; Aycan, 1997; Shaffer et al., 1999; Glanz et al., 2001; Kraimer et al., 2001; Bhaskar-Shrinivas et al., 2005). The internet is increasingly providing a new form of emotional support, allowing family members to retain extensive contacts with the wider family and other people from their home locale. Coping behaviors. Coping behaviors are the last item in family capabilities. They mediate the crossover process among family members (Westman, 2001). Westman (2001) distinguishes problem-focused strategies, which are positively related to well-being, and emotion-focused strategies such as outbursts, which are negatively related. Most expatriate studies do not directly and systematically address coping behaviors; Mendenhall and Oddou’s (1986) typology of expatriates, Feldman and Tompson’s (1993) and Shaffer et al.’s (2000) studies are some of the exceptions. The literatures on work role transitions (Nicholson, 1984; West et al., 1987; Nicholson and West, 1988) and on immigrant adjustment (Berry et al., 1988; Segall et al., 1990; Berry, 1990) discuss modes of adjustment in terms of changing oneself, or one’s environment in the case of work role transitions, and in terms of the value attached to maintaining relationships with host and home groups in the case of immigrants. These concepts have been integrated occasionally in expatriate research (Zimmermann et al., 2003). The only coping behaviors that have been studied extensively with regards to expatriates and repatriates are the ones that deal with a reduction in number and/or intensity of demands, i.e. premature return, psychological withdrawal, social marginalization, and insufficient work effort and performance (Hays, 1971, 1974; David, 1972; Misa and Fabricatore, 1979; Conway, 1984; Tung, 1987; Black and Stephens, 1989; Black and Gregersen, 1992; Naumann, 1992, 1993a, b; Banai, 1992; Banai and Reisel, 1993; Birdseye and Hill, 1995; Harzing, 1995; Bhuian and Islam, 1996; Gregersen and Black, 1996; Aycan, 1997; Martinko and Douglas, 1999; Liu, 2005; Bhaskar-Shrinivas et al., 2005). Company programs do not pay much attention to coping behaviors except, perhaps, for elements of stress training as part of preparation. Meanings In business studies there is a rich literature on sensemaking based on the work of Weick (1995). Meanings have only occasionally been studied in expatriate settings. Exceptions are Glanz and colleagues (Glanz et al., 2001; Glanz, 2003), who have applied Weick’s approach to expatriates. An attempt at changing the meanings expatriates attribute to living abroad comes from Osland (1995), who reframes the experience as a “hero’s journey” based on Campbell’s (1968) ideas. The types of meanings of most interest in research on expatriate families are the situational ones on the one hand and the ones related to the outside world on the other. The latter represent the sense the family makes of the new cultural environment and how it sees itself in it. This includes the reference schemes (Schu¨tz, 1944) or cognitive maps (Tolman, 1948) the family shares. Schu¨tz points out that expatriates arrive abroad with a cognitive representation of the host culture that worked for the purposes of the home culture, i.e. interpretation by disinterested observers (Schu¨tz, 1944). But
The expatriate family
333
JMP 23,3
334
their knowledge will not guide them equally well in interactions with members of the host culture. As expatriates move in and learn about the new culture they may encounter two types of situations: (1) old and new elements in their reference scheme may recommend incompatible behaviours leading to cognitive inconsistency; and (2) they may not possess any knowledge at all to guide them, resulting in cognitive ambiguity (Ball-Rokeach, 1973; Grove and Torbio¨rn, 1985). Over time expatriates change their reference scheme to reduce ambiguities and become more effective in navigating the new culture. The sensemaking literature calls this change a process of establishing a new “sense of coherence” (Antonovsky, 1987; Glanz et al., 2001). Individual members will contribute to the family’s creation of meanings. They all bring different experiences in the host culture to the table. Older children can play an important role. They sometimes act as socio-cultural brokers (Glanz et al., 2001). In immigrant families, for example, children sometimes literally interpret for their parents as they are normally quicker to pick up a new language and culture. Socio-cultural brokerage is also provided by community resources of the family, such as host contacts, fellow expatriates or cross-cultural coaches provided by the company. Two countervailing trends are present in expatriate families: on the one hand, the greater resources that a family possesses compared to single individuals facilitate an accelerated changing of the reference scheme. On the other hand, the family often constitutes a miniature environmental bubble of people with the same cultural background. For conversations within the family a reference frame that works for observation and interpretation from a distance may suffice. Therefore, the family as a unit may resist a change in reference frame and thus slow down adjustment. Whether one or the other prevails probably depends on the particular characteristics of each expatriate family and its members – such as their tolerance, flexibility, openness to experience and cohesiveness. The changing of reference schemes and establishment of meanings is a cognitive adjustment process. It is related to affective and behavioral adjustment. Accurate reference schemes develop in interactions with members of the host country and in turn allow expatriates to have more effective and rewarding interactions in the host culture. Improved interactions, reference schemes and confidence in reference schemes will make expatriates also feel more positively. Individual family members – differences and interactions The individual family members’ roles differ significantly. Literature on expatriates points to the rather more difficult situation of the “trailing” partner compared to the expatriate employee (Cohen, 1977; Tung, 1982; Harvey, 1985; Adler, 2002; Selmer and Leung, 2003b; GMAC, 2006). The partner faces an unstructured and minimally supported situation compared to the expatriate employee, who generally has an established office environment with all the structure and support that this entails. School-age children’s situation is also relatively structured. Details depend on whether the school is an international or local one, and whether the language of instruction is familiar or new to the child. Local school systems are very different in how they structure the learning experience, the type of child behavior that is rewarded and the design of the curriculum. Therefore, the demands children face – because of possible
differences in schooling from country to country – are potentially broader in range than those of either of the parents. But research on the adjustment of expatriate children hardly exists (De Leon and McPartlin, 1995; Miyamoto and Kuhlman, 2001). An unexplored area in expatriate studies is the influence that demands faced by children have on parents’ adjustment. Schooling demands may play an important role in parents’ willingness to stay or accept a posting abroad. Take, for example, Japanese parents’ understandable reluctance to spend too much of their children’s school time out of the country in the face of difficulties experienced by their children upon return (Kanno, 2000). Gender, too, is a rarely studied aspect in expatriate research. There are indications that the process of cross-cultural adjustment is not the same for female and male employees (Selmer and Leung, 2003a). Male trailing partners get less support from companies than female accompanying partners (Selmer and Leung, 2003b), which may influence the adjustment of families depending on which partner is the assignee. Family research shows that the crossover among partners varies depending on gender (Parasuraman et al., 1992; Westman, 2001; Leinonen et al., 2003). Parents’ influence on children depends on the gender of both parents and children (Leinonen et al., 2003). Therefore, the study of gender dynamics might fruitfully be applied to expatriate adjustment research. Conclusion Evidence from expatriate and from family research shows systematic omissions and points out questions about expatriate families that can guide future research in the area. The literature on expatriates and expatriate families covers only some of the demands and capabilities discussed. Many of the variables studied fall into the category of independent stressors. Strains, which result from ongoing tensions, do of necessity not feature in a field that has produced hardly any longitudinal data sets. It would be interesting to find out which stressors eventually lead to strains for expatriate families and the conditions under which this happens. Further, are stressors and strains equally likely to cross over among family members, and if not, which ones do so the most? These questions are important for effective planning and support of expatriates and their families by companies. Stressors that lead to strains as well as stressors and strains likely to cross over from one family member to another should get special attention in the design of effective support programs. While outside the scope of this paper, these questions are also relevant for spillover among different roles of expatriates. The third type of demands, i.e. daily hassles, also has not been studied systematically. It might be productive to find out the impact of the various items in the Hassles and Uplifts Scale (DeLongis et al., 1988) on the adjustment of expatriates and their families. Again, this would help in the planning of assignments and targeted support provided by companies to expatriates. Expatriate families face more demands than single expatriates. They also have capabilities that go above and beyond those of individuals. There is very little research on the resources and coping behaviors of families. We know about the role of personality characteristics in expatriate adjustment (Caligiuri, 2000a, b; Graf and Harland, 2005; Mol et al., 2005). We do not know if and how personality characteristics of family members impact the adjustment of others. The crossover and compensation
The expatriate family
335
JMP 23,3
336
of personality characteristics will differ between expatriate and partner; the expatriate has more opportunity to influence directly the adjustment in all domains of a non-working partner than vice versa; the partner can influence work adjustment of the expatriate only remotely. The family as a resource, its structure, organization, cohesion, and communication skills, etc., has received little attention. There is a need to raise theory and empirical research on expatriates from the level of the individual to the level of the family unit in order to gain additional insights into the process of cross-cultural adjustment overall. In addition, companies would benefit from recognizing the full impact of families on expatriate assignments and involving families more in selection. Awareness by the company of the special issues faced by families is a crucial element in adjustment and its support. Such awareness can alleviate many problems. From the company’s perspective, the degree of involvement of families should vary for optimal results depending on factors such as the overall difficulty of the posting and the criticality of the mission for the business. Future research based on the model advanced here will uncover specific contingencies for the better management and support of expatriate assignments that involve families. The drawing in Figure 2 incorporates and modifies Patterson’s (1988, 2002a) ideas to provide an overview of some important areas of study in the field of expatriate family adjustment. Figure 2 shows possible interactions among various demands and capabilities of a two-person family. The drawing focuses on adjustment and excludes the crisis-adaptation dynamic, which may follow a lack of adjustment with the resulting pile-up of demands. One person’s stressors may become stressors for the other; they may develop into strains for either or both. Stressors and daily hassles are, by definition, independent of each other. Clearly not all paths shown will turn out to be relevant. In order to design studies on expatriate families systematically researchers should consider all the paths to be aware of the scope of their approach. The ethics of sending families abroad Finally, we note that contemporary human resources management takes into account the interests of various stakeholders in the company. It tries to answer the question of the employing organization’s responsibilities for its employees and how it should act. This is a difficult question that is attracting the attention of HR scholars (Budd and Scoville, 2005). Expatriate assignments raise additional questions above and beyond
Figure 2. Interactions among demands, capabilities and meanings
those with which domestic HR management grapples. They include issues such as fairness in the way expatriates are treated in comparison to local employees (Toh and DeNisi, 2005) or whether and to what extent the company is responsible for the expatriate family. For international assignees the employer interferes with family life much more than for local employees. Therefore, it also bears additional responsibilities. Abroad, many companies pay for housing, schooling and home visits for the whole family. But how far should this go? Does the company share responsibility for preventable psychological hardships sometimes related to the adjustment and adaptation process? Should companies accept responsibility for partners who give up a job and to what extent[3]? What about the wellbeing of the children? What is the balance of responsibilities between family and company? Countless professional and academic publications give advice on what support to provide to expatriates and their families. Often utilitarian ethics is the basis for recommendations – in order to win the “war for talent.” The question of what is ethical management of expatriate assignments needs to be addressed by ethics scholars. Even if one accepts utilitarianism rather than subscribes to another ethical system to guide one’s decisions (Budd and Scoville, 2005; Scoville et al., 2005), the support granted by many companies may not be found to maximize utility. Given the prevailing approaches to expatriate management following utilitarian ethics will lead to better results for both families and companies. A study of repatriates and HR managers about a range of repatriation-related items supports this contention (Paik et al., 2002). The study identifies several discrepancies in the views of the two groups indicating sub-optimal repatriation management. A study about support to male accompanying partners noticed an overall low level. For a significant minority of respondents this level of support was just right; some even felt more then enough had been provided (Selmer and Leung, 2003b). In addition to individual differences in needs, the specifics of what support is relevant will also depend on the situation (Westman and Etzion, 1995). Therefore, it is important to carefully gauge what is essential in all cases, what has to be provided when needed, and what is not required at all. Notes 1. It should be noted that in all these cases the reference is to partners and children: it is very rare that other members of the more extended family are taken on expatriate assignment and this may in itself create a problem for expatriates and their nuclear family from some cultures. 2. A score of greater than 300 on the scale triggers an 80 percent chance of health problems; a score of between 150 and 300 results in a 50 percent probability. The following items are drawn from a US survey (points in parentheses; source: Cherrington, 1994): change in financial state (38), change in number of arguments with spouse (35), change in responsibilities at work (29), wife beginning or stopping work (26), beginning or ending school (26), change in living conditions (25), revision of personal habits (24), trouble with boss (23), change in work hours or conditions (20), change in residence (20), change in schools (20), change in recreation (19), change in social activities (18), change in number of family get-togethers (15), change in eating habits (15). The scale was updated and revised recently (Hobson et al., 1998). While there were changes in some details, the overall message remains the same. We use the older scale here because for the newer one does not include health-problem predictions. Both scales are based on US samples and may not apply to other nationalities.
The expatriate family
337
JMP 23,3
338
3. Take the example of a Scandinavian multinational corporation that, because of bad experiences in the past, introduced completely separate pension plans for partners who give up their own job, to ensure a level of financial independence in retirement (Personal communication, 1995). References Adler, N.J. (2002), International Dimensions of Organizational Behavior, South Western College Publishing, Cincinnati, OH. Adler, P.S. (1975), “The transitional experience: an alternative view of culture shock”, Journal of Humanistic Psychology, Vol. 15 No. 4, pp. 13-23. Adler, P.S. (1976), “Beyond cultural identity: reflections on cultural and multicultural man”, in Samovar, L.A. and Porter, R.E. (Eds), Intercultural Communication: A Reader, 2nd ed., Wadsworth, Belmont, CA. Adler, P.S. (1987), “Culture shock and the cross-cultural learning experience”, in Luce, L.F. and Smith, E.C. (Eds), Toward Internationalism: Readings in Cross-cultural Communication, 2nd ed., Newbury House, Cambridge, MA. Anderson, B.A. (2005), “Expatriate selection: good management or good luck?”, International Journal of Human Resource Management, Vol. 16, pp. 567-83. Anderze´n, I. and Arnetz, B.B. (1997), “Psychophysiological reactions during the first year of a foreign assignment: results of a controlled longitudinal study”, Work & Stress, Vol. 11, pp. 304-18. Antonovsky, A. (1987), “The salutogenic perspective: toward a new view of health and illness”, Advances – Journal of the Institute for the Advancement of Health, Vol. 4, pp. 47-55. Aycan, Z. (1997), “Expatriate adjustment as a multifaceted phenomenon: individual and organizational predictors”, The International Journal of Human Resource Management, Vol. 8 No. 4, pp. 434-56. Bakker, A.B., Demerouti, E. and Schaufeli, W.B. (2005), “The crossover of burnout and work engagement among working couples”, Human Relations, Vol. 58, pp. 661-89. Ball-Rokeach, S.J. (1973), “From pervasive ambiguity to a definition of the situation”, Sociometry, Vol. 36 No. 3, pp. 378-89. Banai, M. (1992), “The ethnocentric staffing policy in multinational corporations: a self-fulfilling prophecy”, The International Journal of Human Resource Management, December, pp. 451-72. Banai, M. and Reisel, W.D. (1993), “Expatriate managers’ loyalty to the MNC: myth or reality? An explorative study”, Journal of International Business Studies, Second Quarter, pp. 233-45. Barnett, G.A. and Kincaid, D.L. (1983), “Cultural convergence – a mathematical theory”, in Gudykunst, W.B. (Ed.), Intercultural Communication Theory – Current Perspectives, Sage Publications, Beverly Hills, CA. Berry, J.W. (1990), “Psychology of acculturation: understanding individuals moving between cultures”, in Brislin, R.W. (Ed.), Applied Cross-Cultural Psychology, Sage Publications, Newbury Park, CA. Berry, J.W., Kim, U. and Boski, P. (1988), “Psychological acculturation of immigrants”, in Kim, Y.Y. and Gudykunst, W.B. (Eds), Cross-Cultural Adaptation – Current Approaches, Sage Publications, Newbury Park, CA. Bhaskar-Shrinivas, P., Harrison, D.A., Shaffer, M.A. and Luk, D.M. (2005), “Input-based and time-based models of international adjustment: meta-analytic evidence and theoretical extensions”, Academy of Management Journal, Vol. 48, pp. 257-81.
Bhuian, S.N., Islam, M.S. (1996), Continuance commitment and extrinsic job satisfaction among a novel multicultural expatriate workforce, The Mid-Atlantic Journal of Business, March, pp. 35-46. Birdseye, M.G. and Hill, J.S. (1995), “Individual, organizational/work and environmental influences on expatriate turnover tendencies: an empirical study”, Journal of International Business Studies, Fourth Quarter, pp. 787-813. Black, J.S. (1988), “Work role transitions: a study of American expatriate managers in Japan”, Journal Of International Business Studies, Summer, pp. 277-94. Black, J.S. (1990), “Locus of control, social support, stress, and adjustment in international transfers”, Asia Pacific Journal of Management, Vol. 7 No. 1, pp. 1-29. Black, J.S. and Gregersen, H.B. (1991a), “When Yankee comes home: factors related to expatriate and spouse repatriation adjustment”, Journal of International Business Studies, Fourth Quarter, pp. 671-94. Black, J.S. and Gregersen, H.B. (1991b), “Antecedents to cross-cultural adjustment for expatriates in Pacific Rim assignments”, Human Relations, Vol. 44 No. 5, pp. 497-515. Black, J.S. and Gregersen, H.B. (1991c), “The other half of the picture: antecedents of spouse cross-cultural adjustment”, Journal of International Business Studies, Third Quarter, pp. 461-77. Black, J.S. and Gregersen, H.B. (1992), “Serving two masters: managing the dual allegiance of expatriate employees”, Sloan Management Review, Summer, pp. 61-71. Black, J.S. and Mendenhall, M. (1990), “Cross-cultural training effectiveness: a review and a theoretical framework for future research”, Academy of Management Review, Vol. 15 No. 1, pp. 113-36. Black, J.S. and Mendenhall, M. (1991), “The U-curve adjustment hypothesis revisited: a review and theoretical framework”, Journal of International Business Studies, Second Quarter, pp. 225-47. Black, J.S. and Stephens, G.K. (1989), “The influence of the spouse on American expatriate adjustment and intent to stay in Pacific Rim overseas assignments”, Journal of Management, Vol. 15 No. 4, pp. 529-44. Black, J.S., Gregersen, H.B. and Mendenhall, M.E. (1992), “Toward a theoretical framework of repatriation adjustment”, Journal of International Business Studies, Fourth Quarter, pp. 737-60. Black, J.S., Mendenhall, M. and Oddou, G. (1991), “Toward a comprehensive model of international adjustment: an integration of multiple theoretical perspectives”, Academy of Management Review, Vol. 16 No. 2, pp. 291-317. Bolger, N., DeLongis, A., Kessler, R.C. and Wethington, E. (1989), “The contagion of stress across multiple roles”, Journal of Marriage and the Family, Vol. 51, pp. 175-83. Braun, L.M. (1989), “Predicting adaptational outcomes from hassles and other measures”, Journal of Social Behavior and Personality, Vol. 4, pp. 363-76. Brewster, C. and Harris, H. (Eds) (1999), International HRM – Contemporary Issues in Europe, Routledge, London. Budd, J.W. and Scoville, J.G. (2005), “Moral philosophy, business ethics, and the employment relationship”, in Budd, J.W. and Scoville, J.G. (Eds), The Ethics of Human Resources and Industrial Relations, Labor and Employment Relations Association, Champaign, IL. Caligiuri, P.M. (2000a), “Selecting expatriates for personality characteristics: a moderating effect of personality on the relationship between host national contact and cross-cultural adjustment”, Management International Review, Vol. 40, pp. 61-80.
The expatriate family
339
JMP 23,3
Caligiuri, P.M. (2000b), “The Big Five personality characteristics as predictors of expatriate’s desire to terminate and supervisor-rated performance”, Personnel Psychology, Vol. 53, pp. 67-88. Caligiuri, P.M., Hyland, M.A.M. and Bross, A.S. (1998), “Testing a theoretical model for examining the relationship between family adjustment and expatriates’ work adjustment”, Journal of Applied Psychology, Vol. 83, pp. 598-614.
340
Campbell, J. (1968), The Hero with a Thousand Faces, Princeton University Press, Princeton, NJ. Cherrington, D.J. (1994), Organizational Behavior – The Management of Individual and Organizational Performance, Allyn and Bacon, Boston, MA. Church, A.T. (1982), “Sojourner adjustment”, Psychological Bulletin, Vol. 91 No. 3, pp. 540-72. Cleveland, H., Mangone, G.J. and Adams, J.C. (1960), The Overseas Americans, Arno Press, New York, NY. Cohen, E. (1977), “Expatriate communities”, Current Sociology/Sociologie Contemporaine, Vol. 24 No. 3, pp. 5-133. Conway, M.A. (1984), “Reducing expatriate failure rates”, Personnel Administrator, July, pp. 31-8. David, K.H. (1972), “Intercultural adjustment and applications of reinforcement theory to problems of ‘culture shock’”, Trends, Vol. 4 No. 3, pp. 1-64. De Leon, C.T. and McPartlin, D. (1995), “Adjustment of expatriate children”, in Selmer, J. (Ed.), Expatriate Management – New Ideas for International Business, Greenwood Press, Westport, CT. DeLongis, A., Folkman, S. and Lazarus, R.S. (1988), “The impact of daily stress on health and mood: psychological and social resources as mediators”, Journal of Personality and Social Psychology, Vol. 54, pp. 486-95. Dickmann, M., Doherty, N. and Brewster, C. (2006), “Why do they go? Individual and corporate perspectives on the factors influencing the decision to accept an international assignment”, Knowledge, Action and the Public Concern, The Academy of Management, Atlanta, GA. Earley, P.C. (1987), “Intercultural training for managers: a comparison of documentary and interpersonal methods”, Academy of Management Journal, Vol. 30 No. 4, pp. 685-98. Ellingsworth, H.W. (1988), “A theory of adaptation in intercultural dyads”, in Kim, Y.Y. and Gudykunst, W.B. (Eds), Theories in Intercultural Communication, Sage Publications, Thousand Oaks, CA. Evans, P., Pucik, V. and Barsoux, J.-L. (2002), The Global Challenge: Frameworks for International Human Resource Management, McGraw-Hill Irwin, Boston, MA. Feinberg, M.E. (2003), “The internal structure and ecological context of coparenting: a framework for research and intervention”, Parenting: Science and Practice, Vol. 3, pp. 95-131. Feldman, D.C. and Tompson, H.B. (1993), “Expatriation, repatriation, and domestic geographical relocation: an empirical investigation of adjustment to new job assignments”, Journal of International Business Studies, Third Quarter, pp. 507-29. Forster, N. (1992), “International managers and mobile families: the professional and personal dynamics of trans-national career pathing and job mobility in the 1990s”, The International Journal of Human Resource Management, Vol. 3 No. 3, pp. 605-23. Franke, J. and Nicholson, N. (2002), “Who shall we send? Cultural and other influences on the rating of selection criteria for expatriate assignments”, International Journal of Cross Cultural Management, Vol. 2, pp. 21-36.
Fukuda, K.J. and Chu, P. (1994), “Wrestling with expatriate family problems – Japanese experience in East Asia”, International Studies of Management and Organization, Vol. 24 No. 3, pp. 36-47.
The expatriate family
Furnham, A. (1988), “The adjustment of sojourners”, in Kim, Y.Y. and Gudykunst, W.B. (Eds), Cross-Cultural Adaptation – Current Approaches, Sage Publications, Newbury Park, CA. Glanz, L. (2003), “Expatriate stories: a vehicle of professional development abroad?”, Journal of Managerial Psychology, Vol. 18, pp. 259-74. Glanz, L., Williams, R. and Hoeksema, L. (2001), “Sensemaking in expatriation – a theoretical basis”, Thunderbird International Business Review, Vol. 43, pp. 101-19. GMAC (2006), Global Relocation Trends 2005 Survey Report, GMAC Global Relocation Services, Oak Brook, IL. Goleman, D. (1995), Emotional Intelligence, Bloomsbury, London. Graf, A. and Harland, L.K. (2005), “Expatriate selection: evaluating the discriminant, convergent, and predictive validity of five measures of interpersonal and intercultural competence”, Journal of Leadership & Organizational Studies, Vol. 11, pp. 46-62. Greenberg, J. and Baron, R.A. (2003), Behavior in Organizations: Understanding and Managing the Human Side of Work, Pearson Education Prentice-Hall, Upper Saddle River, NJ. Gregersen, H.B. and Black, J.S. (1996), “Multiple commitments upon repatriation: the Japanese experience”, Journal of Management, Vol. 22 No. 2, pp. 209-29. Grove, C.L. and Torbio¨rn, I. (1985), “A new conceptualization of intercultural adjustment and the goals of training”, International Journal of Intercultural Relations, Vol. 9, pp. 205-33. Gudykunst, W.B. and Hammer, M.R. (1988), “Strangers and hosts: an uncertainty reduction based theory of intercultural adaptation”, in Kim, Y.Y. and Gudykunst, W.B. (Eds), Cross-Cultural Adaptation: Current Approaches, Sage Publications, Newbury Park, CA. Gullahorn, J.T. and Gullahorn, J.E. (1963), “An extension of the U-curve hypothesis”, Journal of Social Issues, Vol. 19 No. 3, pp. 33-47. Harris, H. and Brewster, C. (1999), “The coffee-machine system: how international selection really works”, The International Journal of Human Resource Management, Vol. 10, pp. 488-500. Harrison, J.K., Chadwick, M. and Scales, M. (1996), “The relationship between cross-cultural adjustment and the personality variables of self-efficacy and self-monitoring”, International Journal of Intercultural Relations, Vol. 20 No. 2, pp. 167-88. Harvey, M.C. (1982), “The other side of foreign assignments: dealing with the repatriation dilemma”, Columbia Journal of World Business, Vol. 17, pp. 53-9. Harvey, M.G. (1985), “The executive family: an overlooked variable in international assignments”, Columbia Journal of World Business, Spring. Harzing, A.-W.K. (1995), “The persistent myth of high expatriate failure rates”, The International Journal of Human Resource Management, Vol. 6, pp. 457-74. Hays, R.D. (1971), “Ascribed behavioral determinants of success-failure among US expatriate managers”, Journal of International Business Studies, Vol. 2, pp. 40-6. Hays, R.D. (1974), “Expatriate selection: insuring success and avoiding failure”, Journal of International Business Studies, Vol. 5, pp. 25-37. Hechanova, R., Beehr, T.A. and Christiansen, N.D. (2003), “Antecedents and consequences of employees’ adjustment to overseas assignment: a meta-analytic review”, Applied Psychology, Vol. 52, pp. 213-36.
341
JMP 23,3
342
Ho, K.M. and Keiley, M.K. (2003), “Dealing with denial: a systems approach for family professionals working with parents of individuals with multiple disabilities”, The Family Journal: Counseling and Therapy for Couples and Families, Vol. 11, pp. 239-47. Hobson, C.J., Kamen, J., Szostek, J., Nethercut, C.M., Tiedmann, J.W. and Wojnarowicz, S. (1998), “Stressful life events: a revision and update of the Social Readjustment Rating Scale”, International Journal of Stress Management, Vol. 5, pp. 1-23. Holopainen, J. and Bjo¨rkman, I. (2005), “The personal characteristics of the successful expatriate”, Personnel Review, Vol. 34, pp. 37-50. Kanno, Y. (2000), “Kikokushijo as bicultural”, International Journal of Intercultural Relations, Vol. 24, pp. 361-82. Kealey, D.J. and Protheroe, D.R. (1996), “The effectiveness of cross-cultural training for expatriates: an assessment of the literature on the issue”, International Journal of Intercultural Relations, Vol. 20, pp. 141-65. Kim, Y.Y. (1988), Communication and Cross-Cultural Adaptation, Multilingual Matters, Clevedon/Philadelphia, PA. Kim, Y.Y. (1991), “Intercultural communication competence”, in Ting-Toomey, S. and Korzenny, F. (Eds), Cross-Cultural Interpersonal Communication, Sage Publications, Thousand Oaks, CA. Kim, Y.Y. and Gudykunst, W.B.E. (1988), Cross-Cultural Adaptation: Current Approaches, Sage Publications, Newbury Park, CA. Kim, Y.Y. and Ruben, B.D. (1988), “Intercultural transformation: a systems theory”, in Kim, Y.Y. and Gudykunst, W.B. (Eds), Theories in Intercultural Communication, Sage Publications, Thousand Oaks, CA. Kincaid, D.L. (1988), “The convergence theory and intercultural communication”, in Kim, Y.Y. and Gudykunst, W.B. (Eds), Theories in Intercultural Communication, Sage Publications, Thousand Oaks, CA. Klever, P. (2001), “The nuclear family functioning scale: initial development and preliminary validation”, Families, Systems & Health, Vol. 19, pp. 397-410. Kraimer, M.L., Wayne, S.J. and Jaworski, R.A. (2001), “Sources of support and expatriate performance: the mediating role of expatriate adjustment”, Personnel Psychology, Vol. 54, pp. 71-99. Leinonen, J.A., Solantaus, T.S. and Punama¨ki, R.-L. (2003), “Parental mental health and children’s adjustment: the quality of marital interaction and parenting as mediating factors”, Journal of Child Psychology and Psychiatry, Vol. 44, pp. 227-41. Liu, C.-H. (2005), “The effects of repatriates’ overseas assignment experiences on turnover intentions”, Journal of American Academy of Business, Vol. 7, pp. 124-30. Lundstedt, S. (1963), “An introduction to some evolving problems in cross-cultural research”, Journal of Social Issues, Vol. 19 No. 3, pp. 1-9. Martinko, M.J. and Douglas, S.C. (1999), “Culture and expatriate failure: an attributional explication”, International Journal of Organizational Analysis, Vol. 7, pp. 265-93. Mendenhall, M. and Oddou, G. (1986), “Acculturation profiles of expatriate managers: implications for cross-cultural training programs”, Columbia Journal of World Business, Vol. 21, pp. 73-9. Mercer Human Resource Consulting (2006), “International assignments increasing”, Mercer Human Resource Consulting, New York, NY. Misa, K.F. and Fabricatore, J.M. (1979), “Return on investment of overseas personnel”, Financial Executive, April.
Miyamoto, Y. and Kuhlman, N. (2001), “Ameliorating culture shock in Japanese expatriate children in the US”, International Journal of Intercultural Relations, Vol. 25, pp. 21-40. Mol, S.T., Born, M.P., Willemsen, M.E. and van der Molen, H.T. (2005), “Predicting expatriate job performance for selection purposes: a quantitative review”, Journal of Cross-Cultural Psychology, Vol. 36, pp. 590-620. Nash, D. (1967), “The fate of Americans in a Spanish setting: a study in adaptation”, Human Organization, Vol. 26 No. 3, pp. 157-63. Nash, D. (1969), “The domestic side of a foreign existence”, Journal of Marriage and the Family, Vol. 31, pp. 574-83. Naumann, E. (1992), “A conceptual model of expatriate turnover”, Journal of International Business Studies, Third Quarter, pp. 499-531. Naumann, E. (1993a), “Organizational predictors of expatriate job satisfaction”, Journal of International Business Studies, First Quarter, pp. 61-80. Naumann, E. (1993b), “Antecedents and consequences of satisfaction and commitment among expatriate managers”, Group & Organization Management, Vol. 18 No. 2, pp. 153-87. Navas, M., Rojas, A.J., Garcı´a, M. and Pumares, P. (2007), “Acculturation strategies and attitudes according to the Relative Acculturation Extended Model (RAEM): the perspectives of natives versus immigrants”, International Journal of Intercultural Relations, Vol. 31, pp. 67-86. Navas, M., Garcia, M.C., Sa´nchez, J., Rojas, A.J., Pumares, P. and Ferna´ndez, J.S. (2005), “Relative Acculturation Extended Model (RAEM): new contributions with regard to the study of acculturation”, International Journal of Intercultural Relations, Vol. 29, pp. 21-37. Nicholson, N. (1984), “A theory of work role transitions”, Administrative Science Quarterly, Vol. 29, June, pp. 172-91. Nicholson, N. and Imaizumi, A. (1993), “The adjustment of Japanese expatriates to living and working in Britain”, British Journal of Management, Vol. 4, pp. 119-34. Nicholson, N. and West, M. (1988), Managerial Job Change: Men and Women in Transition, Cambridge University Press, Cambridge. Oberg, K. (1960), “Cultural shock: adjustment to new cultural environments”, Practical Anthropology, July/August, pp. 177-82. Osland, J.S. (1995), The Adventure of Working Abroad: Hero Tales from the Global Frontier, Jossey-Bass, San Francisco, CA. Paik, Y., Segaud, B. and Malinowski, C. (2002), “How to improve repatriation management: are motivations and expectations congruent between the company and expatriates?”, International Journal of Manpower, Vol. 23, pp. 635-48. Parasuraman, S., Greenhaus, J.H. and Granrose, C.S. (1992), “Role stressors, social support, and well-being among two-career couples”, Journal of Organizational Behavior, Vol. 13, pp. 339-56. Patterson, J.M. (1988), “Families experiencing stress”, Family Systems Medicine, Vol. 6, pp. 202-37. Patterson, J.M. (2002a), “Integrating family resilience and family stress theory”, Journal of Marriage and the Family, Vol. 64, pp. 349-60. Patterson, J.M. (2002b), “Understanding family resilience”, Journal of Clinical Psychology, Vol. 58, pp. 233-46. Reich, W.P., Parrella, D.P. and Filstead, W.J. (1988), “Unconfounding the hassles scale: external sources versus internal responses to stress”, Journal of Behavioral Medicine, Vol. 11, pp. 239-49.
The expatriate family
343
JMP 23,3
344
Ruffin, C.L. (1993), “Stress and health: little hassles vs. major life events”, Australian Psychologist, Vol. 28, pp. 201-8. Schein, E.H. (1984), “Coming to a new awareness of organizational culture”, Sloan Management Review, Winter, pp. 3-16. Schell, M.S. and Solomon, C.M. (1997), Capitalizing on the Global Workforce: A Strategic Guide for Expatriate Management, Irwin, Chicago, IL. Schu¨tz, A. (1944), “The stranger: an essay in social psychology”, The American Journal of Sociology, May, pp. 499-507. Scoville, J.G., Lawler, J.J. and Yi, X. (2005), “Non-western ethical frameworks: implications for human resources and industrial relations”, in Budd, J.W. and Scoville, J.G. (Eds), The Ethics of Human Resources and Industrial Relations, Labor and Employment Relations Association, Champaign. IL. Searle, W. and Ward, C. (1990), “The prediction of psychological and sociocultural adjustment during cross-cultural transitions”, International Journal of Intercultural Relations, Vol. 14, pp. 449-64. Segall, M.H., Dasen, P.R., Berry, J.W. and Poortinga, Y.H. (1990), Human Behavior in Global Perspective: An Introduction to Cross-Cultural Psychology, Pergamon Press, New York, NY. Selmer, J. and Leung, A.S.M. (2003a), “International adjustment of female vs. male business expatriates”, The International Journal of Human Resource Management, Vol. 14, pp. 1117-31. Selmer, J. and Leung, A.S.M. (2003b), “Provision and adequacy of corporate support to male expatriate spouses: an exploratory study”, Personnel Review, Vol. 32, pp. 9-21. Shaffer, M.A., Harrison, D.A. and Gilley, K.M. (1999), “Dimensions, determinants, and differences in the expatriate adjustment process”, Journal of International Business Studies, Vol. 30, pp. 557-81. Shaffer, M., Harrison, D., Luk, D.M. and Gilley, M. (2000), “Spouse adjustment to international assignments: direct determinants and the moderating effects of coping strategies”, Management Research News, Vol. 23, pp. 29-31. Smalley, W.A. (1963), “Culture shock, language shock, and the shock of self-discovery”, Practical Anthropologist, March/April, pp. 49-56. Sparrow, P.R. (1999), “International recruitment, selection and assessment: whose route map will you follow?”, in Joynt, P. and Morton, B. (Eds), The Global HR Manager: Creating the Seamless Organisation, IPD, London. Stephens, G.K. and Black, S. (1991), “The impact of spouse‘s career-orientation on managers during international transfers”, Journal of Management, July, pp. 417-28. Taft, R. (1977), “Coping with unfamiliar cultures”, in Warren, N.E. (Ed.), Studies in Cross-Cultural Psychology, Academic Press, London/New York, NY. Takeuchi, R., Yun, S. and Tesluk, P.E. (2002), “An examination of crossover and spillover effects of spousal and expatriate cross-cultural adjustment on expatriate outcomes”, Journal of Applied Psychology, Vol. 87, pp. 655-66. Taylor, S. and Napier, N.K. (1996), “Successful women expatriates: the case of Japan”, Journal of International Management, Vol. 2 No. 1, pp. 51-78. Toh, S.M. and DeNisi, A.S. (2005), “A local perspective to expatriate success”, The Academy of Management Executive, Vol. 19, pp. 132-46. Tolman, E.C. (1948), “Cognitive maps in rats and man”, The Psychological Review, Vol. 55 No. 4, pp. 189-208.
Trompenaars, F. (1993), Riding the Waves of Culture: Understanding Cultural Diversity in Business, Nicholas Brealey Publishing, London. Tung, R.L. (1981), “Selection and training of personnel for overseas assignments”, Columbia Journal of World Business, pp. 68-78. Tung, R.L. (1982), “Selection and training procedures of US, European, and Japanese multinationals”, California Management Review, Vol. 25, pp. 57-71. Tung, R.L. (1984), “Strategic management of human resources in the multinational enterprise”, Human Resource Management, Vol. 23, pp. 129-43. Tung, R.L. (1987), “Expatriate assignments: enhancing success and minimizing failure”, Academy of Management Executive, Vol. 1, pp. 117-26. van der Zee, K.I., Ali, A.J. and Salome´, E. (2005), “Role interference and subjective well-being among expatriate families”, European Journal of Work and Organizational Psychology, Vol. 14, pp. 239-62. van Vianen, A.E.M., de Pater, I.E., Kristof-Brown, A.L. and Johnson, E.C. (2004), “Fitting in: surface- and deep-level cultural differences and expatriates’ adjustment”, Academy of Management Journal, Vol. 47, pp. 697-709. Ward, C. and Kennedy, A. (1999), “The measurement of sociocultural adaptation”, International Journal of Intercultural Relations, Vol. 23, pp. 659-77. Ward, C., Bochner, S. and Furnham, A. (2001), The Psychology of Culture Shock, Routledge, London. Ward, C., Okura, Y., Kennedy, A. and Kojima, T. (1998), “The U-curve on trial: a longitudinal study of psychological and sociocultural adjustment during cross-cultural transition”, International Journal of Intercultural Relations, Vol. 11 No. 22, pp. 277-91. Waxin, M.F. (2004), “Expatriates’ interaction adjustment: the direct and moderator effects of culture of origin”, International Journal of Intercultural Relations, Vol. 28, pp. 61-79. Waxin, M.-F. and Panaccio, A. (2005), “Cross-cultural training to facilitate expatriate adjustment: it works!”, Personnel Review, Vol. 34, pp. 51-67. Weick, K.E. (1995), Sensemaking in Organizations, Sage Publications, Thousand Oaks, CA. West, M.A., Nicholson, N. and Rees, A. (1987), “Transitions into newly created jobs”, Journal of Occupational Psychology, Vol. 60, pp. 97-113. Westman, M. (2001), “Stress and strain crossover”, Human Relations, Vol. 54, pp. 717-51. Westman, M. and Etzion, D. (1995), “Crossover of stress, strain and resources from one spouse to another”, Journal of Organizational Behavior, Vol. 16, pp. 169-81. Westman, M., Shaffer, M.A. and Lazarova, M. (2006), “Work-family interface among expatriates: an integrative framework”, paper presented at ACREW 2006: Socially Responsive, Socially Responsible Approaches to Employment and Work, Monash University, Prato Centre, Prato. Wu, K.K. and Lam, D.J. (1993), “The relationship between daily stress and health: replicating and extending previous findings”, Psychology and Health, Vol. 8, pp. 329-44. Yamazaki, Y. and Kayes, D.C. (2004), “An experiential approach to cross-cultural learning: a review and integration of competencies for successful expatriate adaptation”, Academy of Management Learning & Education, Vol. 3, pp. 362-79. Zimmermann, A., Holman, D. and Sparrow, P. (2003), “Unravelling adjustment mechanisms: adjustment of German expatriates to intercultural interactions, work, and living conditions in the People’s Republic of China”, International Journal of Cross Cultural Management, Vol. 3, pp. 45-66.
The expatriate family
345
JMP 23,3
346
About the authors Arno Haslberger is Research Professor at Webster University Vienna. His research focuses on cross-cultural adjustment and the management of expatriates. He teaches organizational behavior, human resource management and short courses for executives. He has lived and worked for several years each in the USA, the UK, Germany and Spain. After 18 years abroad he returned to his native Austria. Besides working in academia he has held several human resources generalist and specialist positions in one FTSE 100 and two Fortune 100 companies. He holds a doctorate in sociology and Master’s-level degrees in sociology and in business administration from Johannes Kepler Universita¨t, Linz, Austria, and a Master of Science in industrial relations from Loyola University, Chicago. Arno Haslberger is the corresponding author and can be contacted at:
[email protected] Chris Brewster is Professor of International Human Resource Management at Henley Management College and at the University of Reading, both in the UK, and at Nijmegen University in The Netherlands. He had substantial experience as a practitioner and gained his doctorate from the LSE before becoming an academic. He researches in the field of international and comparative HRM; and has published over 20 books and more than 100 articles. In 2002 he was awarded the Georges Petipas Prize by the practitioner body, the World Federation of Personnel Management Associations, in recognition of his work on IHRM. In 2005, a University of Chicago survey found he was one of the most published authors in international business journals. In 2006 Chris was awarded an Honorary Doctorate by the University of Vaasa, Finland.
To purchase reprints of this article please e-mail:
[email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints