Work Schedules: Shift Work and Long Work Hours
Extended Abstracts from Conference:
Rosalind Chait Barnett
Although it has been assumed that working long work hours is related to poor health and quality-of-life outcomes, the empirical results have been inconsistent (for a review, see Barnett, 1998). Some studies report that long work hours are related to such negative outcomes as high experienced job demands (Karasek & Theorell, 1990), emotional exhaustion (Landsbergis, 1988), marital tension (Hughes & Galinsky, 1994), and work-family conflict (Grzywacz & Marks, 2000). Other studies find a relationship between long work hours and such positive outcomes as high role balance (Marks & MacDermid, 1996), good physical health (Bird & Freemont, 1991; Herold & Waldron, 1985), low psychological distress (Barnett & Shen, 1997; Hughes & Galinsky, 1994), and low anxiety (Kohn & Schooler, 1982). Still others report no significant relationship between long work hours and such outcomes as life satisfaction (Barnett & Gareis, 2000a), marital-role quality (Brennan, Barnett, & Gareis, 2001), job-role quality (Barnett & Gareis, 2000b), and intention to turn over(Barnett & Gareis, 2000a). However, most of the studies have been cross-sectional and, in studies of two-earner couples, have relied on data from only one partner.
Building on a new couple-level conceptualization of objective work hours, we revisited the assumption linking work hours to negative outcomes. Jacobs and Gerson (2001) noted a large increase between 1970 and 1997 in total family work hours (TFWH) in dual-earner couples. In 1970, dual-earner couples clocked an average of 78.0 hours per week between the two of them, whereas by 1997 that figure had risen to 81.3 hours per week. More recent data confirm the increase, with dual-earner couples working a combined 82 hours per week in 2002 (Bond, Thompson, Galinsky, & Prottas, 2003). That number jumps to 91 hours per week among dual-earner couples with children (Bond, et al, 2003). Jacobs and Gerson set out the intriguing hypothesis that within couples, TFWH, not the individual’s work hours, would be related to work-family conflict, especially for parents.
We tested this hypothesis in a longitudinal study of a random sample of 211 dual-earner couples (N = 422 individuals) who were parents or became parents over the course of the study and who lived in one of two Boston suburbs. Each partner was interviewed extensively three times over a two-year period. We expanded on the Jacobs and Gerson conceptualization of couple-level work hours by creating variables representing the time-invariant (i.e., averaged over time) and time-varying components of (1) each partner’s own work hours; (2) the sum of both partners’ work hours, or TFWH; (3) each partner’s proportion of TFWH; and (4) the difference between the two partners’ work hours. In addition, building on the work of others (e.g., Frone, Russell, & Cooper, 1992), we included as outcomes several quality-of-life variables in addition to work-family conflict.
Specifically, we estimated a series of three hierarchical linear models in which we assessed the relationship between TFWH and four outcomes: work-family conflict, job-role quality, marital-role quality, and parent-role quality. In the first model we included the individual work hours for each partner; in the second model we included TFWH and the proportion of the hours worked by each partner; and in the third model we included TFWH and the difference between the two partners’ work hours. For each model, the work hours variables were divided into two components, one time-invariant and the other time-varying. The time-invariant component represents hours averaged over the three time points, while the time-varying component represents differences from the average as observed in each wave.
Two additional time-varying covariates were incorporated into the models: changes in household income per capita and in number of children are included as time-varying covariates in all models. In addition to the work hour variables averaged over the three time points, each model contained as time-invariant predictors the age of the participant and the educational attainment for each member of the couple at the first wave of data collection and the negative affectivity score for each participant and the per capita household income (logged) averaged over the three time points.
Although not explicitly stated as part of the Jacobs and Gerson hypothesis, we estimated the within-couple crossover effects of individual work hours on each outcome. For example, does change over time in the husband’s work hours impact change over time in the wife’s marital-role quality? We found no evidence of crossover effects in which one partner’s work hours affected the other partner’s outcomes, whether in the time-invariant or the time-varying analyses.
The rest of the results provide only limited support for the TFWH hypothesis. With respect to the time-invariant analyses, the only significant relationship between TFWH and an outcome, husband ratings of his work interfering with family, turned out to be driven by the absolute number of hours the husband worked. Similarly, in the time-varying analyses, relationships linking increases in TFWH to increases in his parent-role also turned out to be driven by increases in his work hours alone.
Some support for the Jacobs and Gerson hypothesis is found in two trends linking changes in TFWH to changes in outcomes. For her, as TFWH increased, her marital-role quality showed a trend to decrease. This finding was not driven by changes in her individual work hours or by a crossover effect of changes in his individual work hours. For him, as TFWH increased, his ratings of work interfering with family showed a trend to increase; likewise, this finding was not driven by changes in his individual work hours or by a crossover effect of changes in hers. However, both findings were only marginal; TFWH did not show statistically significant relationships with any study outcomes in either the time-invariant or time-varying analyses in any of the three models.
There were two other findings of note in the time-varying analyses, both involving the difference between the two partners’ work hours. For women, as her hours relative to his increased, her ratings of work interfering with family also increased. In 78.2% of the couples, the husbands worked more hours than did their wives, so in most cases this finding may be understood as wives’ experiences of interference increasing as her hours more closely approached his, or as she had fewer and fewer hours during which her husband was at work and she was not. It is important to note that changes in wives’ work hours alone were not related to her ratings of work interfering with family; instead, it is only in relation to her husband’s hours that this effect occurs.
For men, as his hours relative to hers increased, his marital-role quality decreased. By the reasoning described above, in most cases this finding may be understood as men’s marital-role quality declining as his hours surpassed hers ever further, or as he spent more and more hours at work while his wife was not at work herself. Again, changes in husbands’ work hours alone were not related to his marital-role quality; instead, this effect depended on changes in the relationship between his and his wife’s work hours over time.
The two findings described above may be interpreted as a form of support for the Jacobs and Gerson (2001) hypothesis in that the work hours of both partners need to be taken into account in order to understand the effects of work hours on health and quality-of-life outcomes. That is, knowledge of one partner’s work hours alone is not sufficient for outcome prediction. However, there is only weak evidence for the most direct interpretation of the hypothesis, that it is the simple sum of husbands’ and wives’ work hours that is the critical predictor of outcomes.
Overall, there are relatively few effects of any form of objective work hours, whether those are operationalized as individual work hours or as some form of composite couple work hours measure (TFWH, proportion, or difference). This suggests that objective number of hours worked is not a very strong predictor of health and quality-of-life outcomes. Instead, prior literature and our own previous work suggest that subjective indicators of the meaning of working long hours are generally better predictors of such outcomes.
Indeed, in a series of regression analyses with a different sample, we included objective work hours and one of three subjective indicators of the meaning of work hours as predictors of a host of quality-of-life and mental-health indicators as outcomes (Barnett & Gareis, 2000a #2462, Barnett & Gareis, 2000b #2467, Barnett, Gareis, & Brennan, 1999 #1631). In no instance was objective work hours significantly related to any outcome, and, in every case, a subjective indicator was significantly related to the outcomes we tested.
Barnett, R. C. (1998). Toward a review and reconceptualization of the work/family literature. Genetic, Social and General Psychology Monographs, 124(2), 125-182.
Barnett, R. C., & Gareis, K. C. (2000a). Reduced-hours employment: The relationship between difficulty of trade-offs and quality of life. Work and Occupations, 27(2), 168-187.
Barnett, R. C., & Gareis, K. C. (2000b). Reduced-hours job-role quality and life satisfaction among married women physicians with children. Psychology of Women Quarterly, 24, 358-364.
Barnett, R. C., & Shen, Y.-C. (1997). Gender, high- and low-schedule-control housework tasks, and psychological distress. Journal of Family Issues, 18(4), 403-428.
Barnett, R. C., Gareis, K. C., & Brennan, R. T. (1999). Fit as a mediator of the relationship between work hours and burnout. Journal of Occupational Health Psychology, 4(4), 307-317.
Bird, C. E., & Fremont, A. M. (1991). Gender, time use, and health. Journal of Health and Social Behavior, 32, 114-129.
Bond, J. T., Thompson, C., Galinsky, E., & Prottas, D. (2003). Highlights of the 2002 national study of the changing workforce (No. 3). New York : Families and Work Institute.
Brennan, R. T., Barnett, R. C., & Gareis, K. C. (2001). When she earns more than he does: A longitudinal study of dual-earner couples. Journal of Marriage and Family, 63, 168-182.
Frone, M. R., Russell, M., & Cooper, L. M. (1992). Prevalence of work-family conflict: Are work and family boundaries asymmetrically permeable? Journal of Organizational Behavior, 13, 723-729.
Grzywacz, J. G., & 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, 5(1), 111-126.
Herold, J., & Waldron, I. (1985). Part time employment and women's health. Journal of Occupational Medicine, 27, 405-412.
Hughes, D., & Galinsky, E. (1994). Gender, job and family conditions and psychological symptoms. Psychology of Women Quarterly, 18, 251-270.
Jacobs, J. A., & Gerson, K. (2001). Overworked individuals or overworked families? Explaining trends in work, leisure, and family time. Work and Occupations, 28(1), 40-63.
Karasek, R., & Theorell, T. (1990). Healthy work: Stress, productivity, and the reconstruction of working life. New York : Basic Books, Inc.
Kohn, M. L., & Schooler, C. (1982). Job conditions and personality: A longitudinal assessment of their reciprocal effects. American Journal of Sociology, 87, 1257-1286.
Landsbergis, P. A. (1988). Occupational stress among health care workers: A test of the job demands-control model. Journal of Organizational Behavior, 9(3), 217-239.
Marks, S. R., & MacDermid, S. M. (1996). Multiple roles and the
self: A theory of role balance. Journal of Marriage and the Family,