Consequences Associated With Work-to-Family Conflict: A Review and Agenda for Future Research Tammy D. Allen, David E. L. Herst, Carly S. Bruck, and Martha Sutton University of South Florida
A comprehensive review of the outcomes associated with work-to-family conflict was conducted and effect sizes were estimated. A typology was presented that grouped outcomes into 3 categories: work related, nonwork related, and stress related. Issues concerning the measurement of workfamily conflict were also discussed. The results demonstrate the widespread and serious consequences associated with work-to-family conflict. On the basis of the results of the review, an agenda for future research was provided.
Striking changes in the nature of families and the workforce, such as more dual-career couples and rising numbers of working mothers with young children, have increased the likelihood that employees of both genders have substantial household responsibilities in addition to their work responsibilities (Bond, Galinsky, & Swanberg, 1998; Gilbert, Hallett, & Eldridge, 1994). These radical changes have prompted considerable research related to work and family issues. The topic of work-family conflict has been of particular conflict interest. Recent research indicates that 40% of employed parents experience work-family at least some of the time 1993).
Moreover, (Galinsky, Bond, & Friedman, Kahn, Wolfe, Quinn, Snoek, and Rosenthal (1964) suggested that work-family conflict is a type of interrole conflict in which role pressures from the work and family domains are mutually incompatible to some degree. That is, work-family conflict occurs when demands associated with one domain are Kopelman, incompatible with demands associated with the other domain (Greenhaus & Buetell, 1985; Greenhaus, & Connolly, 1983). Although early research treated work-family conflict primarily as a unidimensional construct, recent research (Frone, Russell, & Cooper, 1992) suggests that it is reciprocal in nature, in that work can interfere with family (work-to-family conflict; WFC) and family can interfere with work (family-to-work conflict; FWC).
WFC and FWC are generally considered distinct but related constructs. Research to date has primarily investigated how work interferes or conflicts with family. Outcomes associated with excessive work interference with family include job dissatisfaction, job burnout, turnover, depression, life dissatisfaction, and marital dissatisfaction (e.g., Adams, King, & King, 19%; R. J. Burke, 1988; Frone et al., 1992; Greenhaus & Beutell, 1985; Netemeyer, Boles, & McMurrain, 1996; Thomas & Ganster, 1995).
Despite the rapidly growing body of literature examining WFC, few efforts have been made to review empirical findings. Over a decade ago, Greenhaus and Beutell (1985) reviewed the studies that had investigated sources or antecedents of WFC. More recently, Kossek and Ozeki (1998) conducted a meta-analysis examining the relationship between WFC and two specific outcomes: job satisfaction and life satisfaction. Kossek and Ozeki’s work was much needed and an informative addition to the literature. However, there are many additional outcome variables that have been empirically related to WFC that were not included in Kossek and Ozeki’s study.
The Galinsky et al. reported that workers who started a new job within the past 2 years stated that the effect of the job on family life was second in importance to open communications when formulating their decision to accept the job. Likewise, Galinsky, Johnson, and Friedman (1993) cited a study conducted by the New York Times indicating that 83% of working mothers and 72% of working fathers reported experiencing conflict between their job demands and their desire to spend more time with their families. These findings underscore the importance of the topic of work-family conflict to both organizations and employees.
Tammy D. Allen, David E. L. Herst, Carly S. Bruck, and Martha Sutton, Department of Psychology, University of South Florida. Aprevious version of this article was presented at the 14th Annual Meeting of the Society for Industrial and Organizational Psychology, Atlanta, Georgia. We thank Mark L. Poteet, Lillian T. Eby, and Paul E. Specter for their helpful comments regarding various aspects of this article. Correspondence concerning this article should he addressed to Tammy D. Allen, University of South Florida, Department of Psychology, 4202 East Fowler Avenue, BEH 339, Tampa, Florida 33620-8200. Electronic mail may he sent to [email protected] of the present article is to fill this void in the literature. This review provides a comprehensive summary and evaluation of empirical research of the outcomes associated with WFC, including an organizing framework and suggestions for future research. An extensive review is needed for several reasons, One area of concern is the limited amount of integration in the field.
The work and family research arena is fractionated because of diverse types of individuals working in it. For example, individuals working in disciplines such as psychology, sociology, business, and social work have contributed to work and family research. This tends to lead to an emphasis on different issues (e.g., sociologists are more concerned with family-related outcomes, whereas organizational psychologists are more concerned with work-related outcomes) without an examination of similar work in other disciplines. As noted by Russell (1991), implications associated with fractionation and isolation are that progress in research and practice is not systematic or integrated.
Separate, disjointed theories may develop across fields as a result. This limits the progress that could be made by taking a broader, more integrative perspective that builds on previous research. By providing a summary of existing research organized under one framework, we hope that researchers from various disciplines will become more familiar with one another’s work, facilitating the integration of findings from various subfields and subsequent theory building. Additionally, a comprehensive review of the area should help clarify and underscore the widespread negative effects of WFC. A better understanding of these effects might aid in efforts to manage the work and family interface. Moreover, highlighting the dysfunctional and socially costly effects associated with WFC may help convince policymakers of the need to provide interventions that can help mitigate WFC.
For clarity and parsimony, our article is restricted to a review of the outcomes associated with work-tofamily conflict (WFC). In some cases, results were reported in which researchers combined WFC with FWC or asked about work and family conflict in general. Those cases are noted in the review. This review is divided into four major sections. First, the criteria used to identify articles for the review and to conduct statistical analyses are briefly described. Second, we examine issues concerning the measurement of WFC. Third, we present the results of our review for the three categories of outcomes followed by a summary and suggestions for additional research for each. Finally, a general discussion of findings and future research are provided.
Relevant articles were identified through manual and computer searches. Computerized searches were conducted through PsycLJT and OVID information bases using the key words “work and family conflict.11 A manual search was conducted of all articles published from 1977 through 1998 in Journal of Applied Psychology, Journal of Organizational Behavior, Journal of Marriage and the Family, Journal of Vocational Behavior, Organizational Behavior and Human Decision-Making Processes, and Academy of Management Journal. Additionally, the reference list of each identified article was manually cross-checked for other potential articles. Attempts were also made to locate articles that used slightly different terms such as “interrole conflict” and “multiple role stress’1 to refer to WFC. Our criteria for inclusion in the review were twofold. One, WFC had to be a quantitatively measured variable in the study. Thus, we eliminated articles that were not empirical. Two, the study had to measure the relationship between WFC and one or more variables that could theoretically be considered outcomes of WFC.
This eliminated articles that focused solely on sources or antecedents of WFC. It should be noted at this point that the majority of studies in the work and family arena have been cross-sectional in nature, precluding firm causal inferences regarding the direction of the relationships studied. For the purposes of the present review, we included variables that seemed more reasonable or plausible as outcomes of WFC rather than as causes. It is not our intention to infer that reverse causality is not feasible. A total of 67 articles were located that fit these criteria. Statements regarding significance are based on the zeroorder bivariate correlation between WFC and the outcome variable reported in each study. Except where noted otherwise, relationships cited in text are in the expected direction (e.g., greater WFC was associated with less job satisfaction).
Figure 1 provides a framework of the variables included in the study. To provide an estimate of the effect size associated with each of the relationships reviewed, we followed meta-analytic procedures described in Rosenthal (1991). Both unweighted and weighted by sample size average correlations were computed. Only studies that included a zero-order bivariate correlation between WFC and another variable were included in these analyses. In circumstances in which a study involved a sample that was a subset of the same sample used in another study, the study with the largest sample was included in die analysis. An exception was made if sample selection criteria were clearly different.
If a study assessed several specific indices of WFC (conflict between parent and worker and conflict between spouse and worker) these were combined to form a general assessment of WFC. A similar approach was used in analyzing several outcome variables. For example, if a .study examined overall mental health and psychological distress, the correlations were combined for the statistical analyses. In studies in which separate correlations were reported for different subgroups (e.g., male vs. female; single-earner vs. dual-earner), the correlation for each subgroup was weighted by sample size and combined. For consistency purposes, we reversed the sign of the correlation in cases such as when a high score on the WFC measure.