This article was downloaded by: [University of Sydney] On: 06 September 2014, At: 09:33 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
The Journal of Psychology: Interdisciplinary and Applied Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/vjrl20
Work-Family Conflict and WellBeing in University Employees a
b
Helen R. Winefield , Carolyn Boyd & Anthony H. Winefield a
b
University of Adelaide
b
University of South Australia Published online: 14 Feb 2014.
To cite this article: Helen R. Winefield, Carolyn Boyd & Anthony H. Winefield (2014) Work-Family Conflict and Well-Being in University Employees, The Journal of Psychology: Interdisciplinary and Applied, 148:6, 683-697, DOI: 10.1080/00223980.2013.822343 To link to this article: http://dx.doi.org/10.1080/00223980.2013.822343
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.
Downloaded by [University of Sydney] at 09:33 06 September 2014
This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
The Journal of Psychology, 2014, 148(6), 683–697 C 2014 Taylor & Francis Group, LLC Copyright doi: 10.1080/00223980.2013.822343
Downloaded by [University of Sydney] at 09:33 06 September 2014
Work-Family Conflict and Well-Being in University Employees HELEN R. WINEFIELD University of Adelaide CAROLYN BOYD ANTHONY H. WINEFIELD University of South Australia
ABSTRACT. This is one of the first reported studies to have reviewed the role of workfamily conflict in university employees, both academic and nonacademic. The goal of this research was to examine the role of work-family conflict as a mediator of relationships between features of the work environment and worker well-being and organizational outcomes. A sample of 3,326 Australian university workers responded to an online survey. Work-family conflict added substantially to the explained variance in physical symptoms and psychological strain after taking account of job demands and control, and to a lesser extent to the variance in job performance. However, it had no extra impact on organizational commitment, which was most strongly predicted by job autonomy. Despite differing in workloads and work-family conflict, academic (“faculty”) and nonacademic staff demonstrated similar predictors of worker and organizational outcomes. Results suggest two pathways through which management policies may be effective in improving worker wellbeing and productivity: improving job autonomy has mainly direct effects, while reducing job demands is mediated by consequent reductions in work-family conflict. Keywords: work-family conflict, university workers, occupational stress
INCREASING WORKFORCE PARTICIPATION by women combined with longer work hours in industrialized nations has led to large social and lifestyle changes in the interactions between paid employment and other aspects of life such as care-giving work and recreational/leisure pursuits (Halpern, 2005; Pocock, 2003). There has been a rapid increase in research on what is variously known as work-home conflict, work-family interference, work-life balance, and related terms. Here we use the term work-family conflict (WFC) and the definition offered Address correspondence to Professor A. H. Winefield, School of Psychology, University of South Australia, Magill Campus, South Australia 5072; [email protected] (e-mail). 683
Downloaded by [University of Sydney] at 09:33 06 September 2014
684
The Journal of Psychology
by Taris et al. (2006): “the degree to which work demands clash with adequate and pleasurable performance in nonwork roles” (p. 140). Work-family conflict has been investigated sometimes as an antecedent of worker well-being and attitudes (i.e., as a job demand or stressor; e.g., Carlson, Grzywack, & Zivnuska, 2009); sometimes as an outcome of job stressors and other work conditions (Taris et al., 2006); sometimes as a consequence of exhaustion and impaired health (Bakker & Geurts, 2004); and sometimes as a mediator of the relationship between working conditions and worker health or commitment (Lourel, Ford, Gamassou, Gu´eguen, & Hartmann, 2009). Several meta-analyses have been published (see Michel, Kotrba, Mitchelson, Clark, & Baltes, 2011). Reciprocal relationships between job characteristics and WFC have also been investigated (Hall, Dollard, Tuckey, Winefield, & Thompson, 2010). In the present study we aimed first, to determine the degree to which WFC would contribute to worker outcomes, over and above the variance explained by work conditions. Second, we explored the role of WFC as a mediator of the relationship between work characteristics on the one hand and individual worker outcomes on the other. The outcomes of interest included worker health and well-being, and also two of interest to employers: organizational commitment and productivity. We investigated these questions using a large sample of university employees. Such an investigation is timely, because universities have become increasingly stressful work environments for both academic (“faculty”) and nonacademic employees. There are increased expectations of academic staff to attract research funding, publish in high quality journals, and teach increasing numbers of students; Jerejian, Reid, and Rees (2013) report an average of 48.8 email messages per day for a sample of Australian academics. There is also mounting pressure on nonacademic staff to carry out exacting administrative duties for both staff and students (Gillespie, Walsh, Winefield, Dua & Stough, 2001; Winefield, Gillespie, Stough, Dua, Hapuarachchi, & Boyd, 2003; Winefield, Boyd, Saebel, & Pignata, 2008). The investigation is one of the few to have studied WFC in university employees, and has particular significance because it compares academic and nonacademic staff. Carlson, Grzywack, and Kacmar (2010) suggested that flexible working arrangements (e.g., allowing employees to work from home) reduce WFC, and academic staff generally enjoy more flexible working arrangements than nonacademic staff, most of whom work fixed hours. However, a consequence of this is that academic staff are more likely to work weekends and during the evening and this may well increase WFC. Accordingly, we were not able to predict which group will experience greater WFC; however, it will be interesting to compare them, as well as to ascertain whether they show similar relationships between WFC and other outcome measures. Theoretical Framework and Hypotheses The job demands-resources (JD-R) model (Bakker & Demerouti, 2007) proposes that excessive or prolonged job demands entail physical and psychological
Downloaded by [University of Sydney] at 09:33 06 September 2014
Winefield et al.
685
costs that lead, via energy depletion, to impaired health and exhaustion. At the same time, job resources promote motivation and engagement, and reduce workrelated stress by offsetting the costs associated with psychological demands. We selected work pressure as our indicator of job demands, and job autonomy as our indicator of job resources. High levels of work pressure are well documented among university staff (Boyd et al., 2011;, and work pressure is a well-known predictor of negative health outcomes, including psychological strain and exhaustion (Sonnentag & Frese, 2003). Job autonomy: the amount of freedom, independence, and discretion that employees have over the scheduling of their work and the procedures used to carry it out, is also of particular significance to university staff especially given anecdotal reports of declining autonomy during recent decades. Job autonomy has been identified as a positive contributor to organizational commitment (Aub´e, Rousseau, & Morin, 2007) and job performance (Binnewies, Sonnentag, & Mojza, 2009) and is negatively related to strain (Van der Doef & Maes, 1999). By spilling over into nonwork domains, job demands curtail workers’ opportunities to detach from work and/or engage in pleasurable nonwork activities. Over time resources are eroded, leading to impaired health, increased strain, and exhaustion (Sonnentag, 2001; Taris et al., 2006), and by implication to reduced performance (Gilboa, Shirom, Fried, & Cooper, 2008). Thus, WFC may be expected to mediate the relationships between job demands (work pressure) to psychological strain, physical symptoms and job performance (self-rated productivity). Hypothesis 1: WFC will mediate the positive relationship between work pressure and (a) psychological strain and (b) physical symptoms. Hypothesis 2: WFC will mediate the negative relationship between work pressure and productivity.
The positive effects of job resources on worker health, attitudes, and performance are well documented (Bakker & Demerouti, 2007). Job autonomy is thought to reduce strain by offsetting the pressure associated with inflexible and demanding work schedules (Taris et al., 2006) and to increase performance by increasing individuals’ capacity to maximize the effectiveness with which they harness other job resources (Binnewies et al., 2009). Job autonomy also engenders positive perceptions of organizational support and fosters employees’ desire to reciprocate via increased commitment (Beauregard & Henry, 2009; Eisenberger, Rhoades, & Cameron, 1999). The extent to which relationships between job autonomy and worker outcomes are mediated by WFC is less clear cut than is the case for job demands (work pressure), partly because of the equivocal nature of the relationship between autonomy and WFC. Overall, the evidence suggests a significant, but relatively weak, negative effect of autonomy on WFC, net of job demands, (Andreassi &
686
The Journal of Psychology
Thompson, 2007; Taris et al., 2006). Therefore, we tested the following hypotheses. Hypothesis 3: WFC will mediate the negative relationship between job autonomy and (a) psychological strain and (b) physical symptoms.
Downloaded by [University of Sydney] at 09:33 06 September 2014
Hypothesis 4: WFC will mediate the positive relationship between job autonomy and productivity.
According to Carlson et al. (2009, 2010), effective role performance in both work and nonwork domains (i.e., work-life balance) fosters positive work attitudes, including high levels of organizational commitment. Conversely, because role interference (e.g., preoccupation with one role while trying to perform in another) undermines role effectiveness, WFC may be expected to reduce positive work attitudes. Carlson et al. (2010) reported that WFC mediated the relationship between schedule flexibility and outcomes such as job satisfaction, but we did not measure schedule flexibility, knowing that academic staff enjoy much greater schedule flexibility than nonacademic staff as noted earlier. Our final hypothesis was: Hypothesis 5: WFC will mediate the positive relationship between job autonomy and organizational commitment.
We analyzed results for academic and nonacademic staff separately as we expected the two groups to differ in work hours and autonomy, but did not formulate any specific hypotheses about how those differences might impact, if at all, on the relationships between the other variables investigated.
Method Participants Participants were 3,326 full-time Australian university workers who responded to an online survey about occupational stress in Australian universities in 2003/4. Participants included 1,342 men and 1,968 women (16 respondents, i.e., 0.5% of the sample, did not declare their sex). The average age of the sample was 42.82 years (SD = 10.66), and 1,308 were academic and 2,018 nonacademic staff. Academic staff, that is, “faculty” (52% male) held teaching and research roles while nonacademic staff (67% female) had professional and clerical/administrative responsibilities.
Winefield et al.
687
Downloaded by [University of Sydney] at 09:33 06 September 2014
Measures Work Pressure Three standard questions were used (Beehr, Walsh, & Taber, 1976). An example item is “I’m rushed in doing my job.” The items were rated on a 4-point scale, from 1 (definitely false) to 4 (definitely true); α = .81. Job Autonomy Six items from the autonomy sub-scale of the Moos Work Environment Scale (Moos & Insel, 1974) were used. The items in question capture diverse aspects of autonomy: the extent to which respondents experience freedom on the job; are encouraged to make their own decisions, use their initiative, and undertake training; are free to criticize management; and discuss their work-related goals with management. An example item is “Staff are encouraged to make their own decisions.” Each item was rated on a 5-point scale (1 = strongly disagree, 5 = strongly agree); α = .76. Work-Family Conflict This comprised five items drawn from a well-known scale (Frone, 2000). Example items are, “My family dislike how often I am preoccupied with my work while I am at home” and “After work, I come home too tired to do some of the things I’d like to do.” Each item was rated on a 5-point scale from 1 (never) to 5 (very frequently); α = .86. Psychological Strain Psychological strain was measured using the 12-item version of the General Health Questionnaire with Likert scoring (GHQ-12; Goldberg & Williams, 1988), which has been widely used as an indicator of psychological distress in both occupational studies and population studies (Andrews, Hall, Teeson, & Henderson, 1999); α = .92. Physical Symptoms Respondents were asked to rate the frequency with which they experienced each of nine stress-related physical symptoms, such as headaches, tiredness, muscle and back pain, dizziness, and shortness of breath (1 = hardly ever, 5 = almost always); α = .85. Organizational Commitment Organizational commitment was measured using four items from a wellknown scale (Porter, Steers, Mowday, & Boulian, 1974). An example item is, “I am willing to put in a great deal of effort beyond that normally expected in order
688
The Journal of Psychology
to help this university be successful” (1 = strongly disagree, 5 = strongly agree); α = .82.
Downloaded by [University of Sydney] at 09:33 06 September 2014
Job Performance A single item assessed respondents’ self-ratings of their productivity over the past 3 months on a 5-point scale (1 = very low, 5 = very high). Demographic Variables Measures of age in years and gender (0 = male, 1 = female) were obtained to enter into analyses as control variables. Procedure Responses were collected during surveys of occupational stress across 13 Australian universities; see Winefield (2002) and Winefield et al. (2008) for full details about the selection of participating tertiary institutions, response rates and findings. Respondents answered questions online anonymously. The second wave of data is used here because it contained a more comprehensive measure of workhome conflict than did the first. As the differences in results between academic and nonacademic staff were few we mainly describe findings from the whole sample here. Analytical Strategy First, we computed descriptive statistics and variable inter-correlations. Second, we conducted confirmatory factor analyses to examine the structure of our measures. Third, we tested the hypothesized model using structural equation modeling (SEM) with the maximum likelihood method of estimation. Results Descriptive Statistics Scores for the scale-based measures were computed by averaging responses for the respective items, except for psychological strain which was based on the sum of the Likert-scored GHQ-12 scores. The academic staff were older, reported more work pressure, more psychological strain, more work-family conflict, less job satisfaction, and less organizational commitment than the nonacademic staff, but the groups did not differ in terms of autonomy. The men compared with the women were older and reported more work pressure, more work-family conflict, less autonomy, fewer physical symptoms, less job satisfaction, and less organizational commitment (details of t test results available from authors on request). Table 1 shows the zero-order correlations among the study variables. Because of the large sample, even small correlations were statistically significant at p < .001. Substantial positive correlations occurred among job demands, WFC,
Winefield et al.
689
TABLE 1. Variable Inter-Correlations
Downloaded by [University of Sydney] at 09:33 06 September 2014
M 1. Age 2. Gender 3. Work pressure 4. Autonomy 5. Work-family conflict 6. Psychological strain 7. Physical symptoms 8. Organizational commitment 9. Self-rated productivity
42.82 — 3.44 2.86 3.29 14.47 2.60 3.55
SD
1
2
3
4
5
6
7
8
10.88 — −.16∗∗ 0.68 .15∗∗ −.07∗∗ 0.75 −.12∗∗ .05∗∗ −.22∗∗ 1.02 .17∗∗ −.08∗∗ .66∗∗ −.30∗∗ 6.54 .00 .02 .28∗∗ −.35∗∗ .38∗∗ 0.68 .01 .10∗∗ .28∗∗ −.31∗∗ .43∗∗ .45∗∗ 0.76 −.01 .07∗∗ −.07∗∗ .36∗∗ −.09∗∗ −.19∗∗ −.17∗∗
3.25 1.25
.01
.08∗∗ −.10∗∗
.23∗∗ −.15∗∗ −.33∗∗ −.23∗∗ .28∗∗
M = Mean; SD = Standard Deviation. ∗ p ≤ .05. ∗∗ p ≤ .01.
psychological strain and physical symptoms. Autonomy was negatively correlated with WFC and the strain outcomes, and positively with organizational commitment and job performance.
Analysis of the Structural Model To test for mediation, we employed an adaptation for SEM of Baron and Kenny’s (1986) approach (Frazier, Tix, & Barron, 2004). The relative fit of a partial mediation model (M2: containing paths from both the predictors and the mediator to the outcomes) was compared with those of an initial direct effects model (M1: containing paths from only the predictors to the outcomes) and a total mediation model (M3: containing paths from only the mediator to the outcomes). We also assessed the significance of indirect effects using the method recommended by Preacher and Hayes (2004). Bootstrapping is used to calculate indirect effects, their standard errors and the associated confidence intervals. In all of the predictive models (M1–3), we also included the core paths from demands and autonomy to WFC. The fit indices of the direct effects, partial mediation and full mediation models are shown in Table 2. Initially, the direct effects model (M1) provided a good fit to the data. Consistent with expectations, work pressure was positively related to WFC, psychological strain, and physical symptoms, while autonomy was negatively related to WFC, strain and symptoms, and positively related to both organizational commitment and productivity. The significance of the direct effects meant that it was appropriate to test for the predicted mediation effects (see Baron & Kenny, 1986). Work pressure, however, was unrelated to productivity and did not predict organizational commitment.
Direct effects
Partial mediation
Full mediation
Final
1
2
3
4
Work pressure → WFC, strain, physical symptoms, productivity Autonomy → WFC, strain, physical symptoms, commitment, productivity M1 + WFC → strain, physical symptoms, commitment, productivity Work pressure, autonomy → WFC; WFC → strain, physical symptoms, organizational commitment, productivity M1 + T1 resources → T2 strain, T1 resources → commitment
Modeled effects
2280.77
3046.45
2278.01
2479.74
χ2
172
177
170
174
df
13.26
17.21
13.40
14.25
χ 2/df
.94
.92
.94
.93
GFI
.95
.93
.95
.94
CFI
Note. All analyses control for the effects of age and gender. M1, M2 . . . M5 = Model 1, Model 2 . . . Model 4. aThe first three values for χ 2 refer to differences between successive nested models. bThe final value for χ 2 refers to the difference M2–M4. ∗ p < .05. ∗∗∗ p < .001.
Description
Model
TABLE 2. Results of SEM Analyses (Maximum Likelihood Estimates), N = 3,326
.93
.91
.93
.93
TLI
Downloaded by [University of Sydney] at 09:33 06 September 2014
.06
.07
.06
.06
RMSEA
1.24
765.68∗∗∗
201.73∗∗∗
χ 2 a
690 The Journal of Psychology
Downloaded by [University of Sydney] at 09:33 06 September 2014
Winefield et al.
691
In the partial mediation model (M2), the addition of paths from WFC to strain, symptoms, organizational commitment, and productivity substantially improved model fit over the direct effects model, χ 2(4) = 201.73, p < .001. WFC was positively related to psychological strain and physical symptoms and negatively related to productivity, but the path from WFC to organizational commitment was not significant. Bootstrap analyses confirmed the significance of the indirect effects of work pressure via WFC on strain and symptoms, as well as those of autonomy via WFC on strain, symptoms, and productivity (see Table 2). Total mediation was only supported for the path from work pressure to psychological strain, and the direct effect of work pressure became nonsignificant in the partial mediation model. The direct path from pressure to physical symptoms (β = .07, p < .05) remained significant, whereas those from autonomy to each of strain, symptoms, and productivity remained small to medium (range: β = .19 to β = −.25, p < .001). Removal of the direct effects of both predictors decreased model fit, showing that the partial mediation model was superior to the full mediation model: χ 2(2) = 8.31, p < .05, for the direct effects of demands; χ 2(2) = 339.34, p < .001, for the direct effects of autonomy. The path from demands to productivity, which had been nonsignificant in the direct effects model became significant and positive in the partial mediation model, β = .07, p < .05. The indirect effects are summarized in Table 3. In the final model, therefore, work pressure was both directly and indirectly related to psychological strain and symptoms, whereas autonomy was both directly and indirectly related to strain, symptoms, and productivity. Autonomy was directly, positively related to organizational commitment. The final model accounted for 23% of the variance in psychological strain, 30% in physical symptoms, and 28% in organizational commitment, but only 8% in productivity. Tests of the final
TABLE 3. Summary of Indirect Effects
Effect Work pressure-strain Work pressure-physical symptoms Work pressure-productivity Autonomy-strain Autonomy-physical symptoms Autonomy-productivity
β
SE
.14∗∗∗ .22∗∗∗ −.08∗∗ −.05∗∗∗ −.06∗∗∗ .01∗∗
.02 .02 .02 .01 .01 .02
CI Lower .12 .20 −.01 −.05 −.07 .01
Upper .18 .26 −.05 −.03 −.05 .02
β (beta; i.e., standardized regression coefficient); SE = standard error; C1 = confidence interval. ∗∗ p < .01. ∗∗∗ p < .001.
692
The Journal of Psychology
Work Pressure
.06*
Psychological Strain
.61*
-.07*
.26*
WFC .36* Physical Symptoms .39*
Downloaded by [University of Sydney] at 09:33 06 September 2014
-.16*
Org Commit
-.27*
.50* Autonomy
.25*
Productivity
FIGURE 1. Final model of work pressure, autonomy and WFC in relation to health and work outcomes. All effects significant at p < .05.
model separately for academic and nonacademic staff showed trivial differences related to sample size effects. The final model is shown in Figure 1. Discussion This study examined the effects of job demands, job control/autonomy, and WFC on well-being and organizational commitment in a large sample of Australian university staff. The results indicated that WFC largely explained the effects of job demands (notably work pressure) on psychological strain and physical symptoms, suggesting that WFC might have acted as the mechanism by which job demands damaged worker health. The results concerning strain and symptoms are broadly consistent with those of Peeters, Montgomery, Bakkers, and Schaufeli (2005) who found that WFC partially mediated the effects of job demands on burnout. Our findings are also consistent with the effort-recovery approach (Meijman & Mulder, 1998), suggesting that it was not high job demands per se that led to undesirable health outcomes, but rather the lack of sufficient opportunity to recover from the effort and energy expended on them, either during, or following official working hours. By interfering with positive nonwork experiences, high job demands impinged on recovery time and, hence, led to psychological strain and psychosomatic complaints. An interesting finding was the equivocal relationship between work pressure and job performance. The indirect effect, via WFC, was negative, while the direct effect, net of WFC, was positive. This suggests that work pressure has both hindrance and challenge effects (e.g., Gilboa et al., 2008), both “driving” performance
Downloaded by [University of Sydney] at 09:33 06 September 2014
Winefield et al.
693
directly, and limiting it because of its deleterious effects on WFC. Multivariate analyses showed that both demands and WFC were largely unrelated to organizational commitment, which was instead predicted mainly by autonomy. This finding is consistent with the JD-R model’s claim that job resources, rather than job demands or stressors, are key predictors of positive work attitudes (Demerouti, Bakker, Nachreiner, & Schaufeli, 2001). It is important to note the substantial effect of autonomy on the outcomes, independently of WFC. Not only did autonomy directly reduce psychological strain and physical symptoms, and directly increase commitment and performance, but it also reduced the detrimental effects of work pressure on these variables. While this is perhaps not surprising, given the negative correlation between autonomy and work pressure, it underscores the importance of autonomy in protecting worker health and promoting worker commitment and productivity. Longitudinal research may establish causal directions and indicate whether the time periods differ over which job conditions and WFC exert their effects on the outcomes. An important finding with implications for managers in and outside the higher education sector arises from the comparisons between academic and nonacademic staff. These two groups of employees of the same organizations differed markedly in work pressure and work-family conflict. The academic staff were generally more senior than the nonacademic staff so this finding is in agreement with DiRenzo, Greenhaus, and Weer (2011) who studied both work interference with family and family interference with work. They reported that “higher-level workers experience greater conflict in both directions than lower-level workers” (p. 305). Nevertheless the relationships between the study variables are essentially the same in the two groups. This is a novel and important finding. One of the few other studies to have looked at work-family conflict in university staff was published recently by Calvo-Salguero, Salinas Martinez-de-Lecea, and Carrasco-Gonzalez (2011). They examined the possible mediating roles of intrinsic and extrinsic job satisfaction in the relationship between FIW (familywork conflict) and WIF (work-family conflict) and general job Satisfaction, in 303 workers from a Spanish public university (both academic and nonacademic staff). They found a negative relationship between FIW and general job satisfaction although intrinsic job satisfaction was a significant mediator. They suggest that some intrinsic aspects of the job such as lack of autonomy may hinder familywork balance, an interpretation that is consistent with the findings reported in the current study. Limitations Study limitations include the cross-sectional design and the reliance on selfreport data. Also we did not include separate measures of time- versus strainbased WFC, and so are unable to determine whether the effects of working hours and perceived (e.g., work pressure) demands differ according to these separate dimensions of WFC (Taris et al., 2006). The mediation effects that we observed
694
The Journal of Psychology
Downloaded by [University of Sydney] at 09:33 06 September 2014
would have been more plausibly interpreted as causal mechanisms had we been able to report longitudinal data. Study Implications WFC was most closely related to job demands in bivariate analyses, consistently with other research reports (e.g., Ilies, Schwind, Wagner, & Johnson, 2007). As workload and the accompanying experience of WFC contribute so strongly to physical symptoms, further research should investigate whether sickness absences increase with high workload, perhaps reducing the productivity benefits to employers that workers’ long hours might be expected to achieve. Perceptions of autonomy at work seem to have a decisive influence on commitment and performance, therefore possibly on retention outcomes. This finding is consistent with reports that some workers who work extremely long hours, but with a sense of having chosen to do so, enjoy their work very much (Western, Gray, Qu, & Stanton, 2004). It is also consistent with Karasek’s (1979) definition of active jobs – those combining high demands with high decision latitude. The significant but modest relationship between WFC and perceived autonomy found here, suggests that workers see a connection between the two variables, but differentiate them in a manner that explains their different relationships to the outcome variables. As well as being consistent with an effort-recovery approach, the results of this study are compatible with a “loss spiral” interpretation such that WFC causes strain and physical symptoms which render the worker less fit to cope with job demands, in turn increasing perceived WFC and further reducing sense of health and well-being. This model of WFC consequences is supported by the longitudinal analyses of Steinmetz, Frese, and Schmidt (2008). The desirability of employers developing “flexible” and “family-friendly” workplace policies to promote work-life balance is supported by our results. Yet, such policies are not useful if they are seen as only applicable to workers with low levels of ambition and commitment to their employer. In such a situation workers hesitate to make use of policies which might in fact enable them to maintain employment especially during peak age periods of informal care-giving responsibilities (McKenna, 1997). Workforce policy interventions which increase job satisfaction by increasing worker control without reducing productivity need to be developed and evaluated, taking account of the fact that these sometimes have unintended consequences. For example, Breaugh and Frye (2008) found that flexible hours and reporting to a family-supportive supervisor both reduced WFC in a sample of fulltime employed parents, but being able to take work home increased it. They proposed that the latter unexpected result might be due to blurring of the work-home boundary due to feeling obliged to take work home in order to finish it. Our findings carry important implications for workplace policies as well as for understanding of WFC. First, reducing WFC by reducing workload may improve worker psychological and physical well-being and possibly reduce sickness
Winefield et al.
695
absences. Second, increased work autonomy seems to have a directly beneficial effect on organizational commitment and productivity, hopefully improving retention. These results suggest testable interventions to increase the health and productivity of the higher education and perhaps other professional workforces.
Downloaded by [University of Sydney] at 09:33 06 September 2014
AUTHOR NOTES Helen R. Winefield is a professor of psychology at the University of Adelaide. Her current research interests are health psychology, job stress, and emotional labor. Carolyn Boyd is a research associate at the University of South Australia. Her current research interests are job stress and burnout. Anthony H. Winefield is a professor of psychology at the University of South Australia. His current research interests are the psychology of unemployment, job stress, and burnout. FUNDING The research reported in this article was supported by grants from the Australian Research Council. REFERENCES Andreassi, J. K., & Thompson, C. A. (2007). Dispositional and situational sources of control: relative impact on work-family conflict and positive spillover. Journal of Managerial Psychology, 22, 722–740. Andrews, G., Hall, W., Teeson, M., & Henderson, S. (1999). The mental health of Australians. Canberra, Australian Central Territory: Mental Health Branch, Commonwealth Department of Health and Aged Care. Aub´e, C., Rousseau, V., & Morin, E. M. (2007). Perceived organizational support and organizational commitment: The moderating effect of locus of control and work autonomy. Journal of Managerial Psychology, 22, 479–495. Bakker, A. B., & Demerouti, E. (2007). The Job Demands-Resources model: state of the art. Journal of Managerial Psychology, 22, 309–328. Bakker, A. B., & Geurts, S. A. E. (2004). Toward a dual-process model of work-home interference. Work and Occupations, 31, 345–366. Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182. doi:10.1037/0022–3514.51.6.1173 Beauregard, T. A. & Henry, L. C. (2009). Making the link between work-life balance practices and organizational performance. Human Resource Management Review, 19, 9–22. Beehr, T. A., Walsh, J. T., & Taber, T. D. (1976). Relationship of stress to individually and organizationally valued states: Higher order needs as a moderator. Journal of Applied Psychology, 61, 41–47. Binnewies, C., Sonnentag, S., & Mojza, E. J. (2009). Daily performance at work: feeling recovered in the morning as a predictor of day-level job performance. Journal of Organizational Behavior, 30, 67–93.
Downloaded by [University of Sydney] at 09:33 06 September 2014
696
The Journal of Psychology
Boyd, C. M., Bakker, A. B., Pignata, S., Winefield, A. H., Gillespie, N., & Stough, C. (2011). A longitudinal test of the job demands-resources model among Australian university academics. Applied Psychology: An International Review. 60, 112–140. Breaugh, J. A., & Frye, N. K. (2008). Work-family conflict: the importance of familyfriendly employment practices and family-supportive supervisors. Journal of Business Psychology, 22, 345–353. Calvo-Salguero, A., Salinas Martinez-de-Lecea, J.-M., & Carrasco-Gonzalez, A.-M. (2011). Work-family and family-work conflict: Does intrinsic-extrinsic satisfaction mediate the prediction of general job satisfaction? Journal of Psychology, 145, 435– 461. Carlson, D. S., Grzywack, J. G., & Kacmar, K. M. (2010). The relationship of schedule flexibility and outcomes via the work-family interface. Journal of Managerial Psychology, 25, 330–355. Carlson, D. S., Grzywack, J. G., & Zivnuska, S. (2009). Is work-family balance more than conflict and enrichment? Human Relations, 62, 1459–1487. Demerouti, E., Bakker, A. B., Nachreiner, F., & Schaufeli, W. B. (2001).The job demandsresources model of burnout. Journal of Applied Psychology, 86, 499–512. DiRenzo, M. S., Greenhaus, J. H., & Weer, C. H. (2011). Job level, demands, and resources as antecedents of work-family conflict. Journal of Vocational Behavior, 78, 305–314. Eisenberger, R. U., Rhoades, L., & Cameron, J. (1999). Does pay for performance incrase or decrease perceived self-determination and intrinsic motivation. Journal of Personality and Social Psychology, 77, 1026–1040. Frazier, P. A., Tix, A. P., & Barron, K. E. (2004). Testing moderator and mediator effects in counseling psychology research. Journal of Counseling Psychology, 51, 115–134. doi:10.1037/0022–0167.51.1.115 Frone, M. R. (2000). Work-family conflict and employee psychiatric disorders: The national comorbidity survey. Journal of Applied Psychology, 85, 888–895. Gilboa, S., Shirom, A., Fried, Y., & Cooper, C. (2008). A meta-analysis of work demand stressors, and job performance: Examining main and moderating effects. Personnel Psychology, 61, 227–271. Gillespie, N. A., Walsh, M., Winefield, A. H., Dua, J., & Stough, C. (2001). Occupational stress in universities: Staff perceptions of the determinants, consequences and moderators of stress. Work and Stress, 15, 53–72. Goldberg, D. P., & Williams, P. (1988). A user’s guide to the GHQ. London: NFER, Nelson. Hall, G. B., Dollard, M. F., Tuckey, M. R., Winefield, A. H., & Thompson, B. M. (2010). Job demands, work–family conflict, and emotional exhaustion in police officers: A longitudinal test of competing theories. Journal of Occupational and Organizational Psychology, 83, 237–250. doi:10.1348/096317908×401723 Halpern, D. F. (2005). Psychology at the intersection of work and family. American Psychologist, 60, 397–409. Ilies, R., Schwind, K. M., Wagner, D. T., & Johnson, M. D. (2007). When can employees have a family life? The effects of daily workload and affect on work-family conflict and social behaviors at home. Journal of Applied Psychology, 92, 1368–1379. Jerejian, A. C. M., Reid C., & Rees, C. S. (2013). The contribution of email volume, email management strategies and propensity to worry in predicting email stress among academics. Computers in Human Behavior, 29, 991–996. Karasek, R. A. (1979). Job demands, job decision latitude, and mental strain: Implications for job redesign. Administrative Science Quarterly, 24, 285–308. Lourel, M., Ford, M. T., Gamassou, C. E., Gu´eguen, N., & Hartmann, A. (2009). Negative and positive spillover between work and home: Relationship to perceived stress and job satisfaction. Journal of Managerial Psychology, 24, 438–449.
Downloaded by [University of Sydney] at 09:33 06 September 2014
Winefield et al.
697
McKenna, E. P. (1997). When work doesn’t work anymore: Women, work and identity. Rydalmere, Australia: Hodder & Stoughton. Meijman, T. F., & Mulder, G. (1998). Psychological aspects of workload. In P. J. Drenth, H. Thierry, & C. J. de Wolff (Eds.), Handbood of work and organizational psychology (2nd ed., pp. 5–33). Hove, UK: Psychology Press. Michel, J. S., Kotrba, L. M., Mitchelson, J. K., Clark, M. A., & Baltes, B. B. (2011). Antecedents of work-family conflict: A meta-analytic review. Journal of Organizational Behavior, 32, 689–725. Moos, R. H., & Insel, P. N. (1974). Work Environment Scale. Palo Alto, CA: Consulting Psychologists Press. Peeters, M. C. W., Montgomery, A. J., Bakker, A. B., & Schaufeli, W. B. (2005). Balancing work and home: How job and home demands are related to burnout. International Journal of Stress Management, 12, 43–61. Pocock, B. (2003). The work/life collision. Adelaide: Federation Press. Porter, L. W., Steers, R. M., Mowday, R. T., & Boulian, P. V. (1974). Organizational commitment, job satisfaction, and turnover among psychiatric technicians. Journal of Applied Psychology, 59, 603–609. Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS procedures for estimating indirect effect in simple mediation models. Behavioral Research Methods, Instruments and Computers, 36, 717–731. Sonnentag, S. (2001). Work, recovery activities, and individual well-being: A diary study. Journal of Occupational Health Psychology, 6, 196–210. doi:10.1037/ 1076–8998.6.3.196. Sonnentag, S., & Frese, M. (2003). Stress in organizations. In W. C. Borman, D. R. Ilgen, & R. J. Klimoski (Eds.), Handbook of psychology: Vol. 12. Industrial and organizational psychology (pp. 453–593). Hoboken, NJ: John Wiley and Sons, Inc. Steinmetz, H., Frese, M., & Schmidt, P. (2008). A longitudinal panel study on antecedents and outcomes of work-home interference. Journal of Vocational Behavior, 73, 231–241. Taris, T. W., Beckers, D. G. J., Verhoeven, L. C., Geurts, S. A. E., Kompier, M. A. J., & van der Linden, D. (2006). Recovery opportunities, work-home interference, and wellbeing among managers. European Journal of Work & Organizational Psychology, 15, 139–157. Van der Doef, M., & Maes, S. (1999). The job demand-control (-support) model and psychological well being: A review of 20 years of empirical research. Work and Stress, 13, 87–114. Western, R., Gray, M., Qu, L., & Stanton, D. (2004). Long work hours and the wellbeing of fathers and their families, Melbourne: Australian Institute of Family Studies Research Paper no. 35. Winefield, A. H. (2002). Unemployment, underemployment, occupational stress and psychological well-being. Australian Journal of Management, 27, 137–148. Winefield, A. H., Boyd, C. Saebel, J., & Pignata, S. (2008). Job stress in university staff: An Australian research study. Bowen Hills, Queensland: Australian Academic Press. Winefield, A. H., Gillespie, N. A., Stough, C., Dua, J., Hapuarachchi, J., & Boyd, C. (2003). Occupational stress in Australian university staff: Results from a national survey. International Journal of Stress Management, 10, 51–63.
Original manuscript received April 9, 2013 Final version accepted July 1, 2013
The Journal of Psychology: Interdisciplinary and Applied Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/vjrl20
Work-Family Conflict and WellBeing in University Employees a
b
Helen R. Winefield , Carolyn Boyd & Anthony H. Winefield a
b
University of Adelaide
b
University of South Australia Published online: 14 Feb 2014.
To cite this article: Helen R. Winefield, Carolyn Boyd & Anthony H. Winefield (2014) Work-Family Conflict and Well-Being in University Employees, The Journal of Psychology: Interdisciplinary and Applied, 148:6, 683-697, DOI: 10.1080/00223980.2013.822343 To link to this article: http://dx.doi.org/10.1080/00223980.2013.822343
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.
Downloaded by [University of Sydney] at 09:33 06 September 2014
This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
The Journal of Psychology, 2014, 148(6), 683–697 C 2014 Taylor & Francis Group, LLC Copyright doi: 10.1080/00223980.2013.822343
Downloaded by [University of Sydney] at 09:33 06 September 2014
Work-Family Conflict and Well-Being in University Employees HELEN R. WINEFIELD University of Adelaide CAROLYN BOYD ANTHONY H. WINEFIELD University of South Australia
ABSTRACT. This is one of the first reported studies to have reviewed the role of workfamily conflict in university employees, both academic and nonacademic. The goal of this research was to examine the role of work-family conflict as a mediator of relationships between features of the work environment and worker well-being and organizational outcomes. A sample of 3,326 Australian university workers responded to an online survey. Work-family conflict added substantially to the explained variance in physical symptoms and psychological strain after taking account of job demands and control, and to a lesser extent to the variance in job performance. However, it had no extra impact on organizational commitment, which was most strongly predicted by job autonomy. Despite differing in workloads and work-family conflict, academic (“faculty”) and nonacademic staff demonstrated similar predictors of worker and organizational outcomes. Results suggest two pathways through which management policies may be effective in improving worker wellbeing and productivity: improving job autonomy has mainly direct effects, while reducing job demands is mediated by consequent reductions in work-family conflict. Keywords: work-family conflict, university workers, occupational stress
INCREASING WORKFORCE PARTICIPATION by women combined with longer work hours in industrialized nations has led to large social and lifestyle changes in the interactions between paid employment and other aspects of life such as care-giving work and recreational/leisure pursuits (Halpern, 2005; Pocock, 2003). There has been a rapid increase in research on what is variously known as work-home conflict, work-family interference, work-life balance, and related terms. Here we use the term work-family conflict (WFC) and the definition offered Address correspondence to Professor A. H. Winefield, School of Psychology, University of South Australia, Magill Campus, South Australia 5072; [email protected] (e-mail). 683
Downloaded by [University of Sydney] at 09:33 06 September 2014
684
The Journal of Psychology
by Taris et al. (2006): “the degree to which work demands clash with adequate and pleasurable performance in nonwork roles” (p. 140). Work-family conflict has been investigated sometimes as an antecedent of worker well-being and attitudes (i.e., as a job demand or stressor; e.g., Carlson, Grzywack, & Zivnuska, 2009); sometimes as an outcome of job stressors and other work conditions (Taris et al., 2006); sometimes as a consequence of exhaustion and impaired health (Bakker & Geurts, 2004); and sometimes as a mediator of the relationship between working conditions and worker health or commitment (Lourel, Ford, Gamassou, Gu´eguen, & Hartmann, 2009). Several meta-analyses have been published (see Michel, Kotrba, Mitchelson, Clark, & Baltes, 2011). Reciprocal relationships between job characteristics and WFC have also been investigated (Hall, Dollard, Tuckey, Winefield, & Thompson, 2010). In the present study we aimed first, to determine the degree to which WFC would contribute to worker outcomes, over and above the variance explained by work conditions. Second, we explored the role of WFC as a mediator of the relationship between work characteristics on the one hand and individual worker outcomes on the other. The outcomes of interest included worker health and well-being, and also two of interest to employers: organizational commitment and productivity. We investigated these questions using a large sample of university employees. Such an investigation is timely, because universities have become increasingly stressful work environments for both academic (“faculty”) and nonacademic employees. There are increased expectations of academic staff to attract research funding, publish in high quality journals, and teach increasing numbers of students; Jerejian, Reid, and Rees (2013) report an average of 48.8 email messages per day for a sample of Australian academics. There is also mounting pressure on nonacademic staff to carry out exacting administrative duties for both staff and students (Gillespie, Walsh, Winefield, Dua & Stough, 2001; Winefield, Gillespie, Stough, Dua, Hapuarachchi, & Boyd, 2003; Winefield, Boyd, Saebel, & Pignata, 2008). The investigation is one of the few to have studied WFC in university employees, and has particular significance because it compares academic and nonacademic staff. Carlson, Grzywack, and Kacmar (2010) suggested that flexible working arrangements (e.g., allowing employees to work from home) reduce WFC, and academic staff generally enjoy more flexible working arrangements than nonacademic staff, most of whom work fixed hours. However, a consequence of this is that academic staff are more likely to work weekends and during the evening and this may well increase WFC. Accordingly, we were not able to predict which group will experience greater WFC; however, it will be interesting to compare them, as well as to ascertain whether they show similar relationships between WFC and other outcome measures. Theoretical Framework and Hypotheses The job demands-resources (JD-R) model (Bakker & Demerouti, 2007) proposes that excessive or prolonged job demands entail physical and psychological
Downloaded by [University of Sydney] at 09:33 06 September 2014
Winefield et al.
685
costs that lead, via energy depletion, to impaired health and exhaustion. At the same time, job resources promote motivation and engagement, and reduce workrelated stress by offsetting the costs associated with psychological demands. We selected work pressure as our indicator of job demands, and job autonomy as our indicator of job resources. High levels of work pressure are well documented among university staff (Boyd et al., 2011;, and work pressure is a well-known predictor of negative health outcomes, including psychological strain and exhaustion (Sonnentag & Frese, 2003). Job autonomy: the amount of freedom, independence, and discretion that employees have over the scheduling of their work and the procedures used to carry it out, is also of particular significance to university staff especially given anecdotal reports of declining autonomy during recent decades. Job autonomy has been identified as a positive contributor to organizational commitment (Aub´e, Rousseau, & Morin, 2007) and job performance (Binnewies, Sonnentag, & Mojza, 2009) and is negatively related to strain (Van der Doef & Maes, 1999). By spilling over into nonwork domains, job demands curtail workers’ opportunities to detach from work and/or engage in pleasurable nonwork activities. Over time resources are eroded, leading to impaired health, increased strain, and exhaustion (Sonnentag, 2001; Taris et al., 2006), and by implication to reduced performance (Gilboa, Shirom, Fried, & Cooper, 2008). Thus, WFC may be expected to mediate the relationships between job demands (work pressure) to psychological strain, physical symptoms and job performance (self-rated productivity). Hypothesis 1: WFC will mediate the positive relationship between work pressure and (a) psychological strain and (b) physical symptoms. Hypothesis 2: WFC will mediate the negative relationship between work pressure and productivity.
The positive effects of job resources on worker health, attitudes, and performance are well documented (Bakker & Demerouti, 2007). Job autonomy is thought to reduce strain by offsetting the pressure associated with inflexible and demanding work schedules (Taris et al., 2006) and to increase performance by increasing individuals’ capacity to maximize the effectiveness with which they harness other job resources (Binnewies et al., 2009). Job autonomy also engenders positive perceptions of organizational support and fosters employees’ desire to reciprocate via increased commitment (Beauregard & Henry, 2009; Eisenberger, Rhoades, & Cameron, 1999). The extent to which relationships between job autonomy and worker outcomes are mediated by WFC is less clear cut than is the case for job demands (work pressure), partly because of the equivocal nature of the relationship between autonomy and WFC. Overall, the evidence suggests a significant, but relatively weak, negative effect of autonomy on WFC, net of job demands, (Andreassi &
686
The Journal of Psychology
Thompson, 2007; Taris et al., 2006). Therefore, we tested the following hypotheses. Hypothesis 3: WFC will mediate the negative relationship between job autonomy and (a) psychological strain and (b) physical symptoms.
Downloaded by [University of Sydney] at 09:33 06 September 2014
Hypothesis 4: WFC will mediate the positive relationship between job autonomy and productivity.
According to Carlson et al. (2009, 2010), effective role performance in both work and nonwork domains (i.e., work-life balance) fosters positive work attitudes, including high levels of organizational commitment. Conversely, because role interference (e.g., preoccupation with one role while trying to perform in another) undermines role effectiveness, WFC may be expected to reduce positive work attitudes. Carlson et al. (2010) reported that WFC mediated the relationship between schedule flexibility and outcomes such as job satisfaction, but we did not measure schedule flexibility, knowing that academic staff enjoy much greater schedule flexibility than nonacademic staff as noted earlier. Our final hypothesis was: Hypothesis 5: WFC will mediate the positive relationship between job autonomy and organizational commitment.
We analyzed results for academic and nonacademic staff separately as we expected the two groups to differ in work hours and autonomy, but did not formulate any specific hypotheses about how those differences might impact, if at all, on the relationships between the other variables investigated.
Method Participants Participants were 3,326 full-time Australian university workers who responded to an online survey about occupational stress in Australian universities in 2003/4. Participants included 1,342 men and 1,968 women (16 respondents, i.e., 0.5% of the sample, did not declare their sex). The average age of the sample was 42.82 years (SD = 10.66), and 1,308 were academic and 2,018 nonacademic staff. Academic staff, that is, “faculty” (52% male) held teaching and research roles while nonacademic staff (67% female) had professional and clerical/administrative responsibilities.
Winefield et al.
687
Downloaded by [University of Sydney] at 09:33 06 September 2014
Measures Work Pressure Three standard questions were used (Beehr, Walsh, & Taber, 1976). An example item is “I’m rushed in doing my job.” The items were rated on a 4-point scale, from 1 (definitely false) to 4 (definitely true); α = .81. Job Autonomy Six items from the autonomy sub-scale of the Moos Work Environment Scale (Moos & Insel, 1974) were used. The items in question capture diverse aspects of autonomy: the extent to which respondents experience freedom on the job; are encouraged to make their own decisions, use their initiative, and undertake training; are free to criticize management; and discuss their work-related goals with management. An example item is “Staff are encouraged to make their own decisions.” Each item was rated on a 5-point scale (1 = strongly disagree, 5 = strongly agree); α = .76. Work-Family Conflict This comprised five items drawn from a well-known scale (Frone, 2000). Example items are, “My family dislike how often I am preoccupied with my work while I am at home” and “After work, I come home too tired to do some of the things I’d like to do.” Each item was rated on a 5-point scale from 1 (never) to 5 (very frequently); α = .86. Psychological Strain Psychological strain was measured using the 12-item version of the General Health Questionnaire with Likert scoring (GHQ-12; Goldberg & Williams, 1988), which has been widely used as an indicator of psychological distress in both occupational studies and population studies (Andrews, Hall, Teeson, & Henderson, 1999); α = .92. Physical Symptoms Respondents were asked to rate the frequency with which they experienced each of nine stress-related physical symptoms, such as headaches, tiredness, muscle and back pain, dizziness, and shortness of breath (1 = hardly ever, 5 = almost always); α = .85. Organizational Commitment Organizational commitment was measured using four items from a wellknown scale (Porter, Steers, Mowday, & Boulian, 1974). An example item is, “I am willing to put in a great deal of effort beyond that normally expected in order
688
The Journal of Psychology
to help this university be successful” (1 = strongly disagree, 5 = strongly agree); α = .82.
Downloaded by [University of Sydney] at 09:33 06 September 2014
Job Performance A single item assessed respondents’ self-ratings of their productivity over the past 3 months on a 5-point scale (1 = very low, 5 = very high). Demographic Variables Measures of age in years and gender (0 = male, 1 = female) were obtained to enter into analyses as control variables. Procedure Responses were collected during surveys of occupational stress across 13 Australian universities; see Winefield (2002) and Winefield et al. (2008) for full details about the selection of participating tertiary institutions, response rates and findings. Respondents answered questions online anonymously. The second wave of data is used here because it contained a more comprehensive measure of workhome conflict than did the first. As the differences in results between academic and nonacademic staff were few we mainly describe findings from the whole sample here. Analytical Strategy First, we computed descriptive statistics and variable inter-correlations. Second, we conducted confirmatory factor analyses to examine the structure of our measures. Third, we tested the hypothesized model using structural equation modeling (SEM) with the maximum likelihood method of estimation. Results Descriptive Statistics Scores for the scale-based measures were computed by averaging responses for the respective items, except for psychological strain which was based on the sum of the Likert-scored GHQ-12 scores. The academic staff were older, reported more work pressure, more psychological strain, more work-family conflict, less job satisfaction, and less organizational commitment than the nonacademic staff, but the groups did not differ in terms of autonomy. The men compared with the women were older and reported more work pressure, more work-family conflict, less autonomy, fewer physical symptoms, less job satisfaction, and less organizational commitment (details of t test results available from authors on request). Table 1 shows the zero-order correlations among the study variables. Because of the large sample, even small correlations were statistically significant at p < .001. Substantial positive correlations occurred among job demands, WFC,
Winefield et al.
689
TABLE 1. Variable Inter-Correlations
Downloaded by [University of Sydney] at 09:33 06 September 2014
M 1. Age 2. Gender 3. Work pressure 4. Autonomy 5. Work-family conflict 6. Psychological strain 7. Physical symptoms 8. Organizational commitment 9. Self-rated productivity
42.82 — 3.44 2.86 3.29 14.47 2.60 3.55
SD
1
2
3
4
5
6
7
8
10.88 — −.16∗∗ 0.68 .15∗∗ −.07∗∗ 0.75 −.12∗∗ .05∗∗ −.22∗∗ 1.02 .17∗∗ −.08∗∗ .66∗∗ −.30∗∗ 6.54 .00 .02 .28∗∗ −.35∗∗ .38∗∗ 0.68 .01 .10∗∗ .28∗∗ −.31∗∗ .43∗∗ .45∗∗ 0.76 −.01 .07∗∗ −.07∗∗ .36∗∗ −.09∗∗ −.19∗∗ −.17∗∗
3.25 1.25
.01
.08∗∗ −.10∗∗
.23∗∗ −.15∗∗ −.33∗∗ −.23∗∗ .28∗∗
M = Mean; SD = Standard Deviation. ∗ p ≤ .05. ∗∗ p ≤ .01.
psychological strain and physical symptoms. Autonomy was negatively correlated with WFC and the strain outcomes, and positively with organizational commitment and job performance.
Analysis of the Structural Model To test for mediation, we employed an adaptation for SEM of Baron and Kenny’s (1986) approach (Frazier, Tix, & Barron, 2004). The relative fit of a partial mediation model (M2: containing paths from both the predictors and the mediator to the outcomes) was compared with those of an initial direct effects model (M1: containing paths from only the predictors to the outcomes) and a total mediation model (M3: containing paths from only the mediator to the outcomes). We also assessed the significance of indirect effects using the method recommended by Preacher and Hayes (2004). Bootstrapping is used to calculate indirect effects, their standard errors and the associated confidence intervals. In all of the predictive models (M1–3), we also included the core paths from demands and autonomy to WFC. The fit indices of the direct effects, partial mediation and full mediation models are shown in Table 2. Initially, the direct effects model (M1) provided a good fit to the data. Consistent with expectations, work pressure was positively related to WFC, psychological strain, and physical symptoms, while autonomy was negatively related to WFC, strain and symptoms, and positively related to both organizational commitment and productivity. The significance of the direct effects meant that it was appropriate to test for the predicted mediation effects (see Baron & Kenny, 1986). Work pressure, however, was unrelated to productivity and did not predict organizational commitment.
Direct effects
Partial mediation
Full mediation
Final
1
2
3
4
Work pressure → WFC, strain, physical symptoms, productivity Autonomy → WFC, strain, physical symptoms, commitment, productivity M1 + WFC → strain, physical symptoms, commitment, productivity Work pressure, autonomy → WFC; WFC → strain, physical symptoms, organizational commitment, productivity M1 + T1 resources → T2 strain, T1 resources → commitment
Modeled effects
2280.77
3046.45
2278.01
2479.74
χ2
172
177
170
174
df
13.26
17.21
13.40
14.25
χ 2/df
.94
.92
.94
.93
GFI
.95
.93
.95
.94
CFI
Note. All analyses control for the effects of age and gender. M1, M2 . . . M5 = Model 1, Model 2 . . . Model 4. aThe first three values for χ 2 refer to differences between successive nested models. bThe final value for χ 2 refers to the difference M2–M4. ∗ p < .05. ∗∗∗ p < .001.
Description
Model
TABLE 2. Results of SEM Analyses (Maximum Likelihood Estimates), N = 3,326
.93
.91
.93
.93
TLI
Downloaded by [University of Sydney] at 09:33 06 September 2014
.06
.07
.06
.06
RMSEA
1.24
765.68∗∗∗
201.73∗∗∗
χ 2 a
690 The Journal of Psychology
Downloaded by [University of Sydney] at 09:33 06 September 2014
Winefield et al.
691
In the partial mediation model (M2), the addition of paths from WFC to strain, symptoms, organizational commitment, and productivity substantially improved model fit over the direct effects model, χ 2(4) = 201.73, p < .001. WFC was positively related to psychological strain and physical symptoms and negatively related to productivity, but the path from WFC to organizational commitment was not significant. Bootstrap analyses confirmed the significance of the indirect effects of work pressure via WFC on strain and symptoms, as well as those of autonomy via WFC on strain, symptoms, and productivity (see Table 2). Total mediation was only supported for the path from work pressure to psychological strain, and the direct effect of work pressure became nonsignificant in the partial mediation model. The direct path from pressure to physical symptoms (β = .07, p < .05) remained significant, whereas those from autonomy to each of strain, symptoms, and productivity remained small to medium (range: β = .19 to β = −.25, p < .001). Removal of the direct effects of both predictors decreased model fit, showing that the partial mediation model was superior to the full mediation model: χ 2(2) = 8.31, p < .05, for the direct effects of demands; χ 2(2) = 339.34, p < .001, for the direct effects of autonomy. The path from demands to productivity, which had been nonsignificant in the direct effects model became significant and positive in the partial mediation model, β = .07, p < .05. The indirect effects are summarized in Table 3. In the final model, therefore, work pressure was both directly and indirectly related to psychological strain and symptoms, whereas autonomy was both directly and indirectly related to strain, symptoms, and productivity. Autonomy was directly, positively related to organizational commitment. The final model accounted for 23% of the variance in psychological strain, 30% in physical symptoms, and 28% in organizational commitment, but only 8% in productivity. Tests of the final
TABLE 3. Summary of Indirect Effects
Effect Work pressure-strain Work pressure-physical symptoms Work pressure-productivity Autonomy-strain Autonomy-physical symptoms Autonomy-productivity
β
SE
.14∗∗∗ .22∗∗∗ −.08∗∗ −.05∗∗∗ −.06∗∗∗ .01∗∗
.02 .02 .02 .01 .01 .02
CI Lower .12 .20 −.01 −.05 −.07 .01
Upper .18 .26 −.05 −.03 −.05 .02
β (beta; i.e., standardized regression coefficient); SE = standard error; C1 = confidence interval. ∗∗ p < .01. ∗∗∗ p < .001.
692
The Journal of Psychology
Work Pressure
.06*
Psychological Strain
.61*
-.07*
.26*
WFC .36* Physical Symptoms .39*
Downloaded by [University of Sydney] at 09:33 06 September 2014
-.16*
Org Commit
-.27*
.50* Autonomy
.25*
Productivity
FIGURE 1. Final model of work pressure, autonomy and WFC in relation to health and work outcomes. All effects significant at p < .05.
model separately for academic and nonacademic staff showed trivial differences related to sample size effects. The final model is shown in Figure 1. Discussion This study examined the effects of job demands, job control/autonomy, and WFC on well-being and organizational commitment in a large sample of Australian university staff. The results indicated that WFC largely explained the effects of job demands (notably work pressure) on psychological strain and physical symptoms, suggesting that WFC might have acted as the mechanism by which job demands damaged worker health. The results concerning strain and symptoms are broadly consistent with those of Peeters, Montgomery, Bakkers, and Schaufeli (2005) who found that WFC partially mediated the effects of job demands on burnout. Our findings are also consistent with the effort-recovery approach (Meijman & Mulder, 1998), suggesting that it was not high job demands per se that led to undesirable health outcomes, but rather the lack of sufficient opportunity to recover from the effort and energy expended on them, either during, or following official working hours. By interfering with positive nonwork experiences, high job demands impinged on recovery time and, hence, led to psychological strain and psychosomatic complaints. An interesting finding was the equivocal relationship between work pressure and job performance. The indirect effect, via WFC, was negative, while the direct effect, net of WFC, was positive. This suggests that work pressure has both hindrance and challenge effects (e.g., Gilboa et al., 2008), both “driving” performance
Downloaded by [University of Sydney] at 09:33 06 September 2014
Winefield et al.
693
directly, and limiting it because of its deleterious effects on WFC. Multivariate analyses showed that both demands and WFC were largely unrelated to organizational commitment, which was instead predicted mainly by autonomy. This finding is consistent with the JD-R model’s claim that job resources, rather than job demands or stressors, are key predictors of positive work attitudes (Demerouti, Bakker, Nachreiner, & Schaufeli, 2001). It is important to note the substantial effect of autonomy on the outcomes, independently of WFC. Not only did autonomy directly reduce psychological strain and physical symptoms, and directly increase commitment and performance, but it also reduced the detrimental effects of work pressure on these variables. While this is perhaps not surprising, given the negative correlation between autonomy and work pressure, it underscores the importance of autonomy in protecting worker health and promoting worker commitment and productivity. Longitudinal research may establish causal directions and indicate whether the time periods differ over which job conditions and WFC exert their effects on the outcomes. An important finding with implications for managers in and outside the higher education sector arises from the comparisons between academic and nonacademic staff. These two groups of employees of the same organizations differed markedly in work pressure and work-family conflict. The academic staff were generally more senior than the nonacademic staff so this finding is in agreement with DiRenzo, Greenhaus, and Weer (2011) who studied both work interference with family and family interference with work. They reported that “higher-level workers experience greater conflict in both directions than lower-level workers” (p. 305). Nevertheless the relationships between the study variables are essentially the same in the two groups. This is a novel and important finding. One of the few other studies to have looked at work-family conflict in university staff was published recently by Calvo-Salguero, Salinas Martinez-de-Lecea, and Carrasco-Gonzalez (2011). They examined the possible mediating roles of intrinsic and extrinsic job satisfaction in the relationship between FIW (familywork conflict) and WIF (work-family conflict) and general job Satisfaction, in 303 workers from a Spanish public university (both academic and nonacademic staff). They found a negative relationship between FIW and general job satisfaction although intrinsic job satisfaction was a significant mediator. They suggest that some intrinsic aspects of the job such as lack of autonomy may hinder familywork balance, an interpretation that is consistent with the findings reported in the current study. Limitations Study limitations include the cross-sectional design and the reliance on selfreport data. Also we did not include separate measures of time- versus strainbased WFC, and so are unable to determine whether the effects of working hours and perceived (e.g., work pressure) demands differ according to these separate dimensions of WFC (Taris et al., 2006). The mediation effects that we observed
694
The Journal of Psychology
Downloaded by [University of Sydney] at 09:33 06 September 2014
would have been more plausibly interpreted as causal mechanisms had we been able to report longitudinal data. Study Implications WFC was most closely related to job demands in bivariate analyses, consistently with other research reports (e.g., Ilies, Schwind, Wagner, & Johnson, 2007). As workload and the accompanying experience of WFC contribute so strongly to physical symptoms, further research should investigate whether sickness absences increase with high workload, perhaps reducing the productivity benefits to employers that workers’ long hours might be expected to achieve. Perceptions of autonomy at work seem to have a decisive influence on commitment and performance, therefore possibly on retention outcomes. This finding is consistent with reports that some workers who work extremely long hours, but with a sense of having chosen to do so, enjoy their work very much (Western, Gray, Qu, & Stanton, 2004). It is also consistent with Karasek’s (1979) definition of active jobs – those combining high demands with high decision latitude. The significant but modest relationship between WFC and perceived autonomy found here, suggests that workers see a connection between the two variables, but differentiate them in a manner that explains their different relationships to the outcome variables. As well as being consistent with an effort-recovery approach, the results of this study are compatible with a “loss spiral” interpretation such that WFC causes strain and physical symptoms which render the worker less fit to cope with job demands, in turn increasing perceived WFC and further reducing sense of health and well-being. This model of WFC consequences is supported by the longitudinal analyses of Steinmetz, Frese, and Schmidt (2008). The desirability of employers developing “flexible” and “family-friendly” workplace policies to promote work-life balance is supported by our results. Yet, such policies are not useful if they are seen as only applicable to workers with low levels of ambition and commitment to their employer. In such a situation workers hesitate to make use of policies which might in fact enable them to maintain employment especially during peak age periods of informal care-giving responsibilities (McKenna, 1997). Workforce policy interventions which increase job satisfaction by increasing worker control without reducing productivity need to be developed and evaluated, taking account of the fact that these sometimes have unintended consequences. For example, Breaugh and Frye (2008) found that flexible hours and reporting to a family-supportive supervisor both reduced WFC in a sample of fulltime employed parents, but being able to take work home increased it. They proposed that the latter unexpected result might be due to blurring of the work-home boundary due to feeling obliged to take work home in order to finish it. Our findings carry important implications for workplace policies as well as for understanding of WFC. First, reducing WFC by reducing workload may improve worker psychological and physical well-being and possibly reduce sickness
Winefield et al.
695
absences. Second, increased work autonomy seems to have a directly beneficial effect on organizational commitment and productivity, hopefully improving retention. These results suggest testable interventions to increase the health and productivity of the higher education and perhaps other professional workforces.
Downloaded by [University of Sydney] at 09:33 06 September 2014
AUTHOR NOTES Helen R. Winefield is a professor of psychology at the University of Adelaide. Her current research interests are health psychology, job stress, and emotional labor. Carolyn Boyd is a research associate at the University of South Australia. Her current research interests are job stress and burnout. Anthony H. Winefield is a professor of psychology at the University of South Australia. His current research interests are the psychology of unemployment, job stress, and burnout. FUNDING The research reported in this article was supported by grants from the Australian Research Council. REFERENCES Andreassi, J. K., & Thompson, C. A. (2007). Dispositional and situational sources of control: relative impact on work-family conflict and positive spillover. Journal of Managerial Psychology, 22, 722–740. Andrews, G., Hall, W., Teeson, M., & Henderson, S. (1999). The mental health of Australians. Canberra, Australian Central Territory: Mental Health Branch, Commonwealth Department of Health and Aged Care. Aub´e, C., Rousseau, V., & Morin, E. M. (2007). Perceived organizational support and organizational commitment: The moderating effect of locus of control and work autonomy. Journal of Managerial Psychology, 22, 479–495. Bakker, A. B., & Demerouti, E. (2007). The Job Demands-Resources model: state of the art. Journal of Managerial Psychology, 22, 309–328. Bakker, A. B., & Geurts, S. A. E. (2004). Toward a dual-process model of work-home interference. Work and Occupations, 31, 345–366. Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182. doi:10.1037/0022–3514.51.6.1173 Beauregard, T. A. & Henry, L. C. (2009). Making the link between work-life balance practices and organizational performance. Human Resource Management Review, 19, 9–22. Beehr, T. A., Walsh, J. T., & Taber, T. D. (1976). Relationship of stress to individually and organizationally valued states: Higher order needs as a moderator. Journal of Applied Psychology, 61, 41–47. Binnewies, C., Sonnentag, S., & Mojza, E. J. (2009). Daily performance at work: feeling recovered in the morning as a predictor of day-level job performance. Journal of Organizational Behavior, 30, 67–93.
Downloaded by [University of Sydney] at 09:33 06 September 2014
696
The Journal of Psychology
Boyd, C. M., Bakker, A. B., Pignata, S., Winefield, A. H., Gillespie, N., & Stough, C. (2011). A longitudinal test of the job demands-resources model among Australian university academics. Applied Psychology: An International Review. 60, 112–140. Breaugh, J. A., & Frye, N. K. (2008). Work-family conflict: the importance of familyfriendly employment practices and family-supportive supervisors. Journal of Business Psychology, 22, 345–353. Calvo-Salguero, A., Salinas Martinez-de-Lecea, J.-M., & Carrasco-Gonzalez, A.-M. (2011). Work-family and family-work conflict: Does intrinsic-extrinsic satisfaction mediate the prediction of general job satisfaction? Journal of Psychology, 145, 435– 461. Carlson, D. S., Grzywack, J. G., & Kacmar, K. M. (2010). The relationship of schedule flexibility and outcomes via the work-family interface. Journal of Managerial Psychology, 25, 330–355. Carlson, D. S., Grzywack, J. G., & Zivnuska, S. (2009). Is work-family balance more than conflict and enrichment? Human Relations, 62, 1459–1487. Demerouti, E., Bakker, A. B., Nachreiner, F., & Schaufeli, W. B. (2001).The job demandsresources model of burnout. Journal of Applied Psychology, 86, 499–512. DiRenzo, M. S., Greenhaus, J. H., & Weer, C. H. (2011). Job level, demands, and resources as antecedents of work-family conflict. Journal of Vocational Behavior, 78, 305–314. Eisenberger, R. U., Rhoades, L., & Cameron, J. (1999). Does pay for performance incrase or decrease perceived self-determination and intrinsic motivation. Journal of Personality and Social Psychology, 77, 1026–1040. Frazier, P. A., Tix, A. P., & Barron, K. E. (2004). Testing moderator and mediator effects in counseling psychology research. Journal of Counseling Psychology, 51, 115–134. doi:10.1037/0022–0167.51.1.115 Frone, M. R. (2000). Work-family conflict and employee psychiatric disorders: The national comorbidity survey. Journal of Applied Psychology, 85, 888–895. Gilboa, S., Shirom, A., Fried, Y., & Cooper, C. (2008). A meta-analysis of work demand stressors, and job performance: Examining main and moderating effects. Personnel Psychology, 61, 227–271. Gillespie, N. A., Walsh, M., Winefield, A. H., Dua, J., & Stough, C. (2001). Occupational stress in universities: Staff perceptions of the determinants, consequences and moderators of stress. Work and Stress, 15, 53–72. Goldberg, D. P., & Williams, P. (1988). A user’s guide to the GHQ. London: NFER, Nelson. Hall, G. B., Dollard, M. F., Tuckey, M. R., Winefield, A. H., & Thompson, B. M. (2010). Job demands, work–family conflict, and emotional exhaustion in police officers: A longitudinal test of competing theories. Journal of Occupational and Organizational Psychology, 83, 237–250. doi:10.1348/096317908×401723 Halpern, D. F. (2005). Psychology at the intersection of work and family. American Psychologist, 60, 397–409. Ilies, R., Schwind, K. M., Wagner, D. T., & Johnson, M. D. (2007). When can employees have a family life? The effects of daily workload and affect on work-family conflict and social behaviors at home. Journal of Applied Psychology, 92, 1368–1379. Jerejian, A. C. M., Reid C., & Rees, C. S. (2013). The contribution of email volume, email management strategies and propensity to worry in predicting email stress among academics. Computers in Human Behavior, 29, 991–996. Karasek, R. A. (1979). Job demands, job decision latitude, and mental strain: Implications for job redesign. Administrative Science Quarterly, 24, 285–308. Lourel, M., Ford, M. T., Gamassou, C. E., Gu´eguen, N., & Hartmann, A. (2009). Negative and positive spillover between work and home: Relationship to perceived stress and job satisfaction. Journal of Managerial Psychology, 24, 438–449.
Downloaded by [University of Sydney] at 09:33 06 September 2014
Winefield et al.
697
McKenna, E. P. (1997). When work doesn’t work anymore: Women, work and identity. Rydalmere, Australia: Hodder & Stoughton. Meijman, T. F., & Mulder, G. (1998). Psychological aspects of workload. In P. J. Drenth, H. Thierry, & C. J. de Wolff (Eds.), Handbood of work and organizational psychology (2nd ed., pp. 5–33). Hove, UK: Psychology Press. Michel, J. S., Kotrba, L. M., Mitchelson, J. K., Clark, M. A., & Baltes, B. B. (2011). Antecedents of work-family conflict: A meta-analytic review. Journal of Organizational Behavior, 32, 689–725. Moos, R. H., & Insel, P. N. (1974). Work Environment Scale. Palo Alto, CA: Consulting Psychologists Press. Peeters, M. C. W., Montgomery, A. J., Bakker, A. B., & Schaufeli, W. B. (2005). Balancing work and home: How job and home demands are related to burnout. International Journal of Stress Management, 12, 43–61. Pocock, B. (2003). The work/life collision. Adelaide: Federation Press. Porter, L. W., Steers, R. M., Mowday, R. T., & Boulian, P. V. (1974). Organizational commitment, job satisfaction, and turnover among psychiatric technicians. Journal of Applied Psychology, 59, 603–609. Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS procedures for estimating indirect effect in simple mediation models. Behavioral Research Methods, Instruments and Computers, 36, 717–731. Sonnentag, S. (2001). Work, recovery activities, and individual well-being: A diary study. Journal of Occupational Health Psychology, 6, 196–210. doi:10.1037/ 1076–8998.6.3.196. Sonnentag, S., & Frese, M. (2003). Stress in organizations. In W. C. Borman, D. R. Ilgen, & R. J. Klimoski (Eds.), Handbook of psychology: Vol. 12. Industrial and organizational psychology (pp. 453–593). Hoboken, NJ: John Wiley and Sons, Inc. Steinmetz, H., Frese, M., & Schmidt, P. (2008). A longitudinal panel study on antecedents and outcomes of work-home interference. Journal of Vocational Behavior, 73, 231–241. Taris, T. W., Beckers, D. G. J., Verhoeven, L. C., Geurts, S. A. E., Kompier, M. A. J., & van der Linden, D. (2006). Recovery opportunities, work-home interference, and wellbeing among managers. European Journal of Work & Organizational Psychology, 15, 139–157. Van der Doef, M., & Maes, S. (1999). The job demand-control (-support) model and psychological well being: A review of 20 years of empirical research. Work and Stress, 13, 87–114. Western, R., Gray, M., Qu, L., & Stanton, D. (2004). Long work hours and the wellbeing of fathers and their families, Melbourne: Australian Institute of Family Studies Research Paper no. 35. Winefield, A. H. (2002). Unemployment, underemployment, occupational stress and psychological well-being. Australian Journal of Management, 27, 137–148. Winefield, A. H., Boyd, C. Saebel, J., & Pignata, S. (2008). Job stress in university staff: An Australian research study. Bowen Hills, Queensland: Australian Academic Press. Winefield, A. H., Gillespie, N. A., Stough, C., Dua, J., Hapuarachchi, J., & Boyd, C. (2003). Occupational stress in Australian university staff: Results from a national survey. International Journal of Stress Management, 10, 51–63.
Original manuscript received April 9, 2013 Final version accepted July 1, 2013
