Skip to content

This report was published by the former Department of Families, Community Services and Indigenous Affairs (FaCSIA).
Contents | Downloads

8. Employment and wellbeing



Paid employment delivers a range of benefits for families. As well as income, it can provide the opportunity for social interaction and support, and can be associated with increased self-esteem (London et al. 2004). On the other hand, unsatisfactory employment can be damaging to an individual and their family. For parents, paid employment generally needs to be combined with caring for their children. Thus, the extent to which work provides benefits for families is likely to also depend on working hours, whether the other parent is working (if a couple-parent family), and work conditions.

This section looks at the associations between work arrangements and individual and family wellbeing. Wellbeing has been defined broadly, in terms of parents' health, relationship quality and conflicts, work–family strains and gains, and time pressure. Work–family strains and gains are discussed and analysed in Section 7, and in this section a more detailed analysis is presented. Wellbeing also includes economic prosperity (or financial wellbeing), so references to those findings reported in Section 6 on income, financial hardship and perceived prosperity are made where relevant. This section focuses on the following questions:

As in previous sections, the results are based on cross-sectional analyses of LSAC. The analyses therefore, identify co-relationships in the data, that is they show where indicators of wellbeing vary or are constant across different combinations of work arrangements or conditions. The findings should not be interpreted as causal relationships. Further waves of LSAC will provide opportunities to examine the nature of these relationships more fully.

Understanding the connections between employment and wellbeing is timely. Since the early 1970s, there has been a steady increase in Australian mothers' labour force participation, especially into part-time jobs (Organisation for Economic Co-operation and Development 2002–2005; Gray et al. 2003). Dual-earner families have now become the most common family form (Renda 2003). Debate on parents' employment and wellbeing is shifting, with a focus on the sorts of jobs parents have, as well as the hours they work and their conditions of work. Further, there is evidence that parents' work hours and conditions are also changing. These changes include increases in part-time work (for women), jobs with very long hours, unpaid overtime, and casual employment. There has also been a debate about whether employment has become less secure (Campbell 2004; Australian Centre for Industrial Relations Research and Training 1999; Gray et al. 2003; Wooden 2002; Hancock 2002). The extent to which there is a relationship between characteristics of jobs and family wellbeing has important policy implications.

In this section several sets of analyses are conducted. A broad overview of the relations between mothers' and fathers' wellbeing and their employment status (not employed, part-time employed, and full-time employed) is presented. A comparison between single and couple mothers is also made.

This section also includes an examination of the way parental wellbeing varies with different work conditions, including flexible work hours, work on weekends or evenings/nights, job security and job autonomy or control. Alexander and Baxter (2005) estimated the associations between these variables and work-to-family spillover, also using LSAC. Both longitudinal and cross-sectional research has found that job insecurity and low autonomy are associated with a range of negative health outcomes, especially depression and anxiety (D'Souza et al. 2003; Ferrie 2001; Marmot et al. 1999; Stansfeld et al. 1995). Job insecurity has also been associated with more work-to-family conflict, impaired marital and family functioning including poorer communication and problem solving, and poorer mental health in spouses (Jansen et al. 2003; Larson, Wilson & Beley 1994; Westman, Etzion & Danon 2001).

In contrast, other features of work, including its complexity and the degree of challenge and employee skills involved, have been associated with more responsive and developmentally sound parenting, better language and educational stimulation in the home, and improved emotional wellbeing in children (Cooksey, Menaghan & Jekielek 1997; Greenberger, O'Neil & Nagel 1994; Menaghan 1991). Factors such as length of work hours, flexibility of work, autonomy and job security may be related to parents' wellbeing and how they manage family routines and care. For single parents, there may be no other person to help shoulder care and domestic work and often only one income to support the family and so the impact of work arrangements could play a crucial role in parental wellbeing. For dual-earner families, the employment status of both parents is relevant to family life and the changes outlined above could be compounded in families where both parents are employed (Jacobs & Gerson 2001), affecting their wellbeing. To examine this possibility, couple data were used to analyse variations in wellbeing across partners' working hours.

back to top

8.1 Data, measures and methods

Data and measures

Initial analyses of the data found that there were no substantive differences between the cohorts in associations between work arrangements and wellbeing. Thus both LSAC cohorts were combined for the analyses reported in this section. Later in this section, multivariate techniques are used to simultaneously control for possible differences between the cohorts by including the variables age of youngest child and number of children.

The wellbeing indicators are summarised in Box 8.1.

For health, fair or poor health was assessed with a single question shown to be predictive of later morbidity and mortality (Gandek et al. 1998; Idler & Benyamini 1997; McCallum, Shadbolt & Wang et al. 1994), and dichotomised to reflect good versus poor health (good = excellent, very good or good; poor = fair or poor).

Three measures of psychological wellbeing were used. Single items measured parents' rating of how difficult their life was currently (some to very many difficulties) and how well they were coping (problems coping).63 Psychological distress was assessed with the six-item Kessler distress scale (K6). This scale has been used in the Australian National Survey of Mental Health and Wellbeing and assesses the presence and extent of anxious or depressive symptoms. The K6 shows strong discriminatory power for mood and anxiety disorders but is not a diagnostic instrument (Furukawa et al. 2003; Kessler et al. 2002). High scores represent symptoms of non-specific psychological distress.

For partnered parents, relationship wellbeing was assessed using two measures. A subscale of the Quality of Co-parental Interaction Scale was used to assess the frequency of arguments and conflict in the relationships (arguments with partner). This scale has been validated against clinician ratings of relationship quality, and assesses conflict between spouses. Relationship quality is assessed using six items from the Relationship Assessment Scale (RAS) developed by Hendrick (1988). Scale scores correlate with measures of love, sexual attitudes, self-disclosure, commitment, and investment in a relationship. The scale is also strongly correlated with a much longer, well-validated measure, the Dyadic Adjustment Scale (Spanier 1976).

Work and family spillover measures were based on the work–family gains and work–family strains scores discussed in Section 7. As discussed in that section, these measures are based on work of Marshall and Barnett (1993). Another dimension of work–family spillover is measured using a rating of time pressure. This one item measure is widely used in social science surveys and was also used in the Household, Income, and Labour Dynamics in Australia (HILDA) survey and the ABS Living Standard Survey. Time pressure is often highest when there are young children in the family; employed mothers tend to report more time pressure compared to employed fathers; and time pressure is a risk factor for depression (Roxburgh 2004).

Box 8.1: Wellbeing measures
Name Description
Health:  
Fair or poor health Derived from perceived health ('In general, would you say your own health is ...'. This indicator is set to one for those who rate their health as fair or poor. Aggregate statistics show the percentage with fair or poor health, and analysis of individual-level data is done using logistic regression.
Psychological wellbeing:  
Some to very many difficulties Derived from the question, 'How difficult do you feel your life is at present?'. This indicator is set to one for those who report they have some, many or very many problems or stresses, and to zero for those who report they have no or few problems or stresses. Aggregate statistics show the percentage with some, many or very many problems or stresses, and analysis of individual-level data is done using logistic regression.
Problems coping Derived from the question, 'How well do you think you are coping?' This indicator is set to one for those who report they are not coping at all, or coping a little or fairly well, and to zero for those who report they are coping very or extremely well. Aggregate statistics show the percentage who are not coping very well, and analysis of individual-level data is done using logistic regression.
Psychological distress Derived from six items, which include the degree to which respondents felt nervous, hopeless or restless. Scored from one to five, a higher score equates to more distress. Aggregate statistics represent the mean of this score, and analysis of individual-level data is done using ordinary least squares (OLS) regression.
Relationship with partner: (only applicable for partnered parents):
Arguments with partner Derived from four items, which measure how often there are arguments, hostility or violence in the relationship. Scored from one to five, with a higher score meaning more arguing. Aggregate statistics represent the mean of this score, and analysis of individual-level data is done using OLS regression.
Relationship quality Derived from six items, which measure the respondent's contentment with the relationship with their partner. Scores range from one to five, a higher score indicating more positive view of the relationship. Aggregate statistics represent the mean of this score, and analysis of individual-level data is done using OLS regression.
Work–family spillover: (only applicable for employed parents)
Work–family gains Derived from six items, relating to the positive effects of work on parenting and on self. The scale ranges from one to five, where the higher score equates to more 'gains'. Aggregate statistics represent a mean of this score, and analysis of individual-level data is done using OLS regression.
Work–family strains Derived from four items, relating to the negative effects of work on family and the negative effects of family on work. The scale ranges from one to five, where a higher score equates to more 'strains'. Aggregate statistics represent the mean of this score, and analysis of individual-level data is done using OLS regression.
Time pressure Derived from the question, 'How often do you feel rushed or pressed for time?' This indicator is set to one for those 'often' or 'always' rushed. Aggregate statistics show the percentage often or always rushed, and analysis of individual-level data is done using logistic regression.

Methods

Relationships between wellbeing and family and job characteristics were investigated using multivariate techniques. These analyses are reported later in this section. This involved estimating a statistical model for each of the wellbeing measures. Separate models were estimated for mothers and fathers. The models included a wide range of explanatory variables. Where the dependent variable (the wellbeing measure) was a scale, it was treated as continuous, and estimated using ordinary least squares (OLS). Where the dependent variable was binary it was estimated using logistic regression.

The following control variables were included in all models:

Parent characteristics

Family characteristics

Other variables related to employment were also included. Four sets of models were estimated, to show the different effect of employment variables. These models are outlined in Box 8.2.

Box 8.2: Wellbeing model sets
Model Set 1 These models include all parents, with a single measure of parental employment status indicating not employed, employed part-time or employed full-time. This model set is used to discuss the effect of employment status on wellbeing and also to discuss associations between wellbeing and other family characteristics (the control variables).
Model Set 2 These models are restricted to employed parents and additional variables are added to show the associations with job characteristics, including a more detailed breakdown of hours worked, job type (permanent/casual/self-employed), whether works weekends or evenings/nights, the flexibility of start and finish times, job security and job autonomy.64
Model Set 3 These models are estimated only for couples and are used to test for the effect of partner's working hours on an individual's wellbeing. As in the first set of models, they include everyone, employed or not, and include the full set of family controls. Unlike the models estimated as part of Model Set 1, they include a more detailed classification of hours (as in Model Set 2), and also include hours worked by the partner.
Model Set 4 These models are restricted to employed parents. They contain the same variables as Model Set 2 but include one additional variable, preferred hours worked.

All estimation results (coefficients and significance levels) are provided in Appendix D, with only a selection of results presented in this section. Given the technical nature of many of the results, to make them more accessible, predicted values were calculated, allowing only the variable of interest (usually hours worked) to vary and holding the values of the other variables at the mean values for mothers and fathers. This shows the effect of one variable for the 'average mother' and the 'average father'. This is the approach used in Gray et al. (2004). Although all multivariate analyses are adjusted for variables that could affect associations between employment and wellbeing, results should be interpreted cautiously. There are likely to be variables not measured in LSAC that could influence the associations observed, or our interpretation of them.

back to top

8.2 Wellbeing of all parents

This section considers the wellbeing of all parents (that is, includes parents who are employed and those who are not). The focus is on differences between mothers and fathers, and also between couple and single mothers, on the measures of individual wellbeing. The measures of wellbeing are then considered in the context of their association with employment status (broadly defined) and other family characteristics.

The initial analysis is restricted to bivariate relationships to provide an overview of the wellbeing measures and differences across sex, family type and employment status. The results from the multivariate analysis are introduced throughout the section, to examine whether the bivariate relationships remain significant once other personal and family characteristics are controlled. These results are based on Model Set 1. The full regression results are provided in Tables D1 to D4.

Wellbeing by family type

Table 8.1 shows the mean wellbeing scores (or the percentage reporting the outcome for binary measures) for all mothers and fathers. Mothers have also been classified by family type, whether partnered or single. The binary wellbeing measures are reported in the top half of the table and the wellbeing measures (scales) are reported in the bottom half. By way of example, 9.1 per cent of mothers reported having fair or poor health compared to 11.4 per cent of fathers. Also, both mothers and fathers had high levels of relationship quality (on a scale from one to five), although for mothers the average score was slightly lower (4.33) than for fathers (4.41).

Overall, mothers reported higher distress, more difficulties, poorer relationship quality and were more often rushed for time compared to fathers. On the other hand, fathers were more likely to report poorer health and had more work–family strains. There was not a statistically significant difference between mothers' and fathers' reports of arguments with their partner, how well they were coping, or in their assessment of the positive effects of work on the family (work–family gains).

Mothers' wellbeing varied by their relationship status. Significantly worse outcomes65 on all wellbeing indicators were observed for single mothers compared to couple mothers, with two exceptions: the work–family gains did not vary significantly by relationship status; and partnered mothers were more likely to report being rushed. Single mothers reported poorer self-rated health, the highest levels of distress, more problems coping and more difficulties, as well as more strains from combining work with parenting.

Table 8.1: Wellbeing of mothers and fathers, mean scores
Wellbeing measures(a)

Mothers

 Fathers
Couple Single Total
 

Per cent

Fair or poor health (-) 8.2 16.1 9.1 11.4
Some to very many difficulties (-) 41.9 60.6 44.0 45.3
Problems coping (-) 45.5 53.6 46.4 46.0
Time pressure (-) 44.9 40.4 44.4 43.1
 

Scale (1–5)

Psychological distress (-) 1.62 1.91 1.65 1.57
Arguments with partner (-) 2.20 n/a 2.20 2.17
Relationship quality (+) 4.33 n/a 4.33 4.41
Work–family gains (+) 3.66 3.71 3.67 3.67
Work–family strains (-) 2.57 2.75 2.59 2.78
Source: LSAC 2004, Wave 1.
Note: For dichotomous variables, the mean is given as a percentage.
(a) Wellbeing measures as detailed in Box 8.1. Measures marked (+) indicate that a higher score means 'better' wellbeing. For those measures marked (-), a higher score means 'worse' wellbeing.

Differences between the sexes in the wellbeing measures were not tested through multivariate analysis, as separate regression models were run for mothers and fathers. For mothers and fathers, the effect of family type was analysed in each model through an indicator variable that identified single-parent families. Given there were very few single fathers in LSAC, the effect of this variable was only considered for mothers. Based on Model Set 1 (see Box 8.2), the multivariate analysis confirmed several of the findings reported above. Even after controlling for other family and personal characteristics, single mothers had significantly higher levels of distress, more difficulties and more strains from combining work with parenting compared to couple mothers. The other differences noted above (that single mothers had worse self-rated health and more problems coping) were also evident, but not significant at the 0.05 level (Tables D1 and D3).

Wellbeing by employment status

Table 8.2 considers the association between wellbeing and being employed part-time, full-time and not being employed. As these are bivariate results, the mean wellbeing scores and percentages shown in Table 8.2 are not adjusted for other factors, such as income or education. Family type, however, is taken into account by reporting separate results for single and couple mothers.

Better wellbeing was often associated with being employed, but there were important exceptions, especially for mothers. First, for couple mothers, those who were not currently employed reported fewer difficulties and less time pressure compared to mothers who worked full-time. Mothers who were not employed also showed similar relationship wellbeing (fewer arguments and better relationship quality) to part-time employed mothers, and both fared better than full-time employed mothers on these indicators. Across the other indicators, part-time work was associated with higher wellbeing in couple mothers compared to not being employed or to working full-time hours. On the other hand, couple mothers who were employed full-time fared worse on all indicators of wellbeing. This was in contrast to financial wellbeing, which was found to be highest (in couple-parent families with both parents working) when primary carers worked full-time, as measured by parental income, and found not to differ by primary carer's hours when measured by hardship or perceived prosperity (see Section 6).

For single mothers, part-time employment was consistently associated with better wellbeing, compared to either full-time employment, or no employment (except for being time pressured; mothers who were not employed reported the lowest time pressure). Note though, financial wellbeing was higher among families in which the mother was working full-time (Section 6). Part-time employed mothers were less likely to report poor mental or physical health, report fewer difficulties, were less likely to report problems coping, and report fewer work–family strains.

For fathers, the associations between wellbeing and employment were relatively clear-cut and consistent. On most indicators, full-time employed fathers were better off in terms of health, psychological wellbeing and relationships. Financial wellbeing was also highest for families with full-time employed fathers (Section 6). Full-time employed fathers did, however, report being more rushed for time, and experienced less work–family gains and more work–family strains, compared to fathers who were employed part-time or not at all.

Table 8.2 reveals that there are different relationships between employment and wellbeing for mothers and fathers. For fathers, full-time employment was fairly consistently associated with better wellbeing. Fathers did, however, report that full-time work created strains for themselves and their parenting. For mothers, both single and couple, full-time employment was not associated with better wellbeing, except financially (see Section 6). Mothers fared better on most indicators when they worked part-time.

The simple bivariate analysis suggests that full-time employment is associated with significant differences in wellbeing for mothers and fathers. However, these associations could be due to other personal or family variables.66 Multivariate analyses were conducted to address these possible influences (see Box 8.2 and results in Tables D1 to D4). The patterns of association between wellbeing and employment status remained virtually unchanged in the multivariate analyses. Adjusting for family and personal characteristics, the pattern of associations between the sexes remained. Mothers employed part-time showed significantly better wellbeing on all indicators compared to being employed full-time. Similarly, mothers who were not employed showed better wellbeing on most measures (not poor health or distress), compared to mothers who were employed full-time.

Optimal wellbeing was significantly associated with full-time employment for fathers. The only exceptions were in relation to work–family strains and being rushed, for which full-time employed fathers reported poorer wellbeing compared to part-time employed fathers.

Table 8.2: Wellbeing by work status: mean scores for mothers, by family type, and fathers
Wellbeing measures(a) Not employed Part-time work Full-time work Total
 

Couple mothers

 

Per cent

Fair or poor health (-) 9.6 5.9 8.9 8.2
More than a few difficulties (-) 40.2 42.1 49.7 41.9
Problems coping (-) 45.8 43.4 51.2 45.5
Time pressure (-) 37.8 50.0 61.7 44.9
 

Scale (1–5)

Psychological distress (-) 1.65 1.57 1.65 1.62
Arguments with partner (-) 2.20 2.19 2.28 2.20
Relationship quality (+) 4.33 4.34 4.26 4.33
Work–family gains (+) n/a 3.68 3.61 3.66
Work–family strains (-) n/a 2.48 2.91 2.57
 

Single mothers

 

Per cent

Fair or poor health (-) 18.9 9.5 12.8 16.1
More than a few difficulties (-) 61.3 57.5 65.8 60.6
Problems coping (-) 55.8 48.5 51.5 53.6
Time pressure (-) 33.9 47.0 78.3 40.4
 

Scale (1–5)

Psychological distress (-) 1.98 1.76 1.80 1.91
Work–family gains (+) n/a 3.71 3.72 3.71
Work–family strains (-) n/a 2.65 3.08 2.75
 

All fathers

 

Per cent

Fair or poor health (-) 24.2 14.5 10.1 11.4
More than a few difficulties (-) 48.8 46.5 44.8 45.3
Problems coping (-) 54.8 48.5 45.1 46.0
Time pressure (-) 23.1 34.5 45.3 43.1
 

Scale (1–5)

Psychological distress (-) 1.71 1.66 1.55 1.57
Arguments with partner (-) 2.16 2.23 2.17 2.17
Relationship quality (+) 4.31 4.34 4.42 4.41
Work–family gains (+) n/a 3.72 3.67 3.67
Work–family strains (-) n/a 2.57 2.79 2.78
Source: LSAC 2004, Wave 1.
Note: For dichotomous variables, the mean is given as a percentage.
(a) Wellbeing measures as detailed in Box 8.1. Measures marked (+) indicate that a higher score means 'better' wellbeing. For those measures marked (-), a higher score means 'worse' wellbeing.

Wellbeing and other family characteristics

This section presents associations between wellbeing and some key personal and family characteristics that were used in the multivariate analyses. While these variables were primarily included as control variables, and therefore not central to the analysis, they provide important contextual information for the findings on wellbeing. As with the results presented in the previous sections, these results are based on Model Set 1, and the full regression results are provided in Tables D1 to D4.

For both mothers and fathers, having a long-term medical condition was significantly associated with poorer wellbeing (with the exception of mothers' work–family strains and fathers feeling time pressured).

Parents who were not born in Australia showed some differences in wellbeing compared to Australian-born parents, and among the non-Australian-born parents, some differences emerged according to the proficiency of spoken English. Being non-Australian-born and having poor language proficiency was associated with worse wellbeing on some indicators. In some cases, however, it was the Australian-born parents who fared worse. For example, Australian-born mothers and fathers had more difficulties and were more often rushed than other mothers and fathers.

Higher educational attainment was not necessarily associated with improved wellbeing for mothers and fathers. Indeed, for mothers, higher educational attainment was associated with having more difficulties, more distress, more work–family strains (but also more work–family gains) and more time pressure. For fathers, those who were more highly educated reported more difficulties and time pressure, although interestingly, those with certificate or diploma qualifications had better health outcomes than those with both lower and higher qualifications. Couple fathers with bachelor degrees or above had higher quality relationships than fathers with low educational attainment.

There were some clear associations between parents' wellbeing and the number of children, often with worse outcomes being associated with more children in the family. Compared to parents with more than one child, parents with one child were less likely to report difficulties or problems coping, and they reported less psychological distress and were less rushed for time (time pressure). These effects were stronger for mothers than for fathers.67

Finally, parental income68 was associated with several dimensions of wellbeing. Both mothers and fathers with higher incomes reported fewer difficulties, less problems coping and lower psychological distress. Mothers also reported better health at higher levels of income. Parental income was unrelated to relationship quality or work–family gains or strains for mothers, but for fathers, higher income was associated with more conflict and less work–family gains. Mothers (but not fathers) were more likely to report being rushed in families with higher parental income.

back to top

8.3 Wellbeing of employed parents

This section is based on the results of Model Set 2 (see Box 8.2), which focus on employed parents only, to assess the associations between wellbeing, work hours and other job characteristics. The analyses include work hours (coded into categories, see below), job type (permanent/ongoing, self-employed and casual) and work conditions (flexible hours, working evenings/nights or weekends, job security, and job autonomy). In addition to these job characteristics, the models include the control variables in Model Set 1. As in Model Set 1, couple and single parents were included in each model, with a control variable to identify single-parent families.

As analyses are multivariate, the associations described are independent of other relationships, and not due to co-relationships among work-related factors. Further, the analysis adjusts for having a medical condition, if born in Australia or not (and English proficiency), level of education, parental income, parent age, family type, number of children, and age of youngest child. These analyses are presented in Tables D5 to D8.

Hours worked

The categories part-time and full-time work analysed in Section 8.2 mask a wide diversity in actual hours worked. Consequently, a more detailed breakdown of work hours and their association with wellbeing has been used in this second set of models (focusing on employed mothers and fathers). For mothers, this meant breaking part-time work hours into a number of categories: short part-time hours (less than 16 hours per week), medium part-time hours (16 to 24 hours) and long part-time hours (25 to 34 hours). Relatively few mothers worked very long hours, so full-time hours were kept as one category (35 or more hours). Fathers who worked full-time were classified as short full-time hours (35 to 44 hours), medium full-time hours (45 to 54 hours) and long full-time hours (55 hours or more), with all part-time hours grouped together (less than 35 hours).

For both mothers and fathers, those working the longest hours were the reference group, contrasting their wellbeing with those working shorter hours. This meant that for mothers the reference group was full-time employed mothers. For fathers the reference group was those working long full-time hours (55 or more hours per week). A significant association, as indicated in Table 8.3 or Table 8.4, means that the wellbeing measure for that category was significantly different from the reference group (the longest hours category). While these significance levels cannot be used alone to assess differences between non-comparison groups (for example, to compare less than 16 hours with 16 to 24 hours in Table 8.3), such differences were formally tested, and where appropriate the results are discussed.

Looking first at mothers, Table 8.3 presents predicted wellbeing scores across the work hour categories. Full-time employed mothers showed poorer wellbeing across all indicators, although the differences between full-time employed mothers and mothers working longer part time-hours (25 to 34 hours per week) were not always significant. This was the case for measures of health, work–family gains, and relationship quality (including arguments with partner). For these measures, mothers working less than 25 hours showed better wellbeing (although they did not differ significantly between those working short or medium part-time hours).

Mothers were least likely to report poor health if they worked 16 to 24 hours compared to all other categories of mothers. The likelihood of mothers having more difficulties or of not coping increased as their work hours increased, although within the different categories of mothers working part-time hours, these differences were often not statistically significant. Similarly, distress scores also varied across categories of part-time hours, but these differences were not significant, although all were significantly lower than those of full-time working mothers.

Work–family strains varied significantly across categories of part-time hours, with those working fewer hours reporting less work–family strain, and mothers working full-time reporting the most. Similarly, working longer hours was associated with a greater likelihood of reporting time pressure, and for mothers who were employed, the least rushed worked short part-time hours.

Table 8.3: Predicted wellbeing: employed mothers by hours worked
Wellbeing measures(a)

Mothers' work hours

1–15 16–24 25–34 35 or more
 

Per cent

Fair or poor health (-) 5.5** 4.1*** 7.2 8.3
Some to very many difficulties (-) 42.0*** 43.0*** 47.1** 54.6
Problems coping (-) 42.3*** 43.4*** 46.3* 52.9
Time pressure (-) 44.8*** 50.3*** 58.6* 64.6

Scale (1–5)

Psychological distress (-) 1.60* 1.56*** 1.59** 1.69
Arguments with partner (-) 2.17*** 2.20*** 2.23 2.29
Relationship quality (+) 4.35* 4.32** 4.26 4.23
Work–family gains (+) 3.68** 3.70*** 3.65 3.59
Work–family strains (-) 2.41*** 2.59*** 2.71*** 2.91
Source: LSAC 2004, Wave 1.
Note: Figures refer to the predicted percentage or the predicted score, holding all control variables at their mean value (for employed mothers only) and varying only hours worked.
(a) Wellbeing measures as detailed in Box 8.1. Measures marked (+) indicate that a higher score means 'better' wellbeing. For those measures marked (-), a higher score means 'worse' wellbeing.
Significance refers to difference from the reference category, 35 hours or more.
*** p < 0.001, ** p < 0.01, * p < 0.05.

Table 8.4 presents fathers' wellbeing across work hour categories. The most striking result from this table is that among those working full-time hours,69 almost all of the predicted wellbeing measures were at their most favourable for short full-time hours (35 to 44) and at their least favourable for long full-time hours (55 or more). For example, fathers' self-reported health was at its most favourable (that is, least poor) among those working 35 to 44 hours, somewhat poorer for those working 45 to 54 hours, and at its poorest for those working 55 hours or more. The only exceptions were with respect to the amount of arguing with a partner and the quality of the relationship, both of which did not vary significantly across the range of full-time hours.

Table 8.4: Predicted wellbeing: employed fathers by hours worked
Wellbeing measures(a)

Fathers' work hours

1–34 35–44 45–54 55 or more
 

Per cent

Fair or poor health (-) 10.5 8.3** 9.6* 12.0
Some to very many difficulties (-) 47.2 41.6* 46.0 48.3
Problems coping (-) 48.4 44.3* 45.8 48.0
Time pressure (-) 39.7*** 37.4*** 46.4*** 56.9
 

Scale (1–5)

Psychological distress (-) 1.61 1.55* 1.57 1.60
Arguments with partner (-) 2.25** 2.17 2.20 2.17
Relationship quality (+) 4.35* 4.41 4.39 4.43
Work–family gains (+) 3.69 3.71** 3.65 3.60
Work–family strains (-) 2.60*** 2.75*** 2.82*** 2.97
Source: LSAC 2004, Wave 1.
Note: Figures refer to the predicted percentage or the predicted score, holding all control variables at their mean value (for employed fathers only) and varying only hours worked.
(a) Wellbeing measures as detailed in Box 8.1. Measures marked (+) indicate that a higher score means 'better' wellbeing. For those measures marked (-), a higher score means 'worse' wellbeing.
Significance refers to difference from the reference category, 55 hours or more.
*** p < 0.001, ** p < 0.01, * p < 0.05.

Unlike for mothers, Table 8.4 shows that part-time work was not associated with better wellbeing for fathers, but was either similar to or worse than the wellbeing of fathers working very long full-time hours (55 hours or more). This was the case for self-reported health, having difficulties, problems coping and psychological distress (all of which were not significantly different from fathers working 55 hours or more); and for more arguments and relationship quality (where part-time hours were associated with the least favourable outcomes and were significantly worse than for fathers working 55 hours or more).

The only exceptions to this for fathers were time pressure and their work–family strains. For these two wellbeing measures, fathers who worked less hours (including part-time) had more favourable outcomes.70

Job type

Job type describes the distinction between the self-employed, permanent/ongoing employees (including fixed-term contract, and a small number employed on 'some other basis') and casual employees.

For mothers, self-employment had strong, and where significant, consistently negative association with wellbeing, relative to permanent/ongoing employment. These differences in association were observed for psychological wellbeing (difficulties, coping and distress), relationship quality, and work–family balance (less work–family gains and more rushed). Self-employed mothers did not, however, have significantly worse work–family strains compared to permanent/ongoing employees.

For fathers, self-employment was associated with more difficulties, more distress and being more rushed. However, it was also associated with more work–family gains.

Comparing casual to permanent/ongoing employees, there were fewer differences in wellbeing. Among mothers, casual employees had lower work–family strains. Among fathers, casual employees reported better coping.

Flexible hours

Working more flexible hours (that is, being readily able to change start and finish times) was associated with positive wellbeing on almost all measures for both mothers and fathers. For mothers, having flexible start and finish times (relative to those with no flexibility) was associated with better health, less difficulties, better coping, less distress, and less work–family strain. Mothers who were only able to change their hours with approval did not have such positive wellbeing, and were not significantly different to those mothers who could not change their hours.

Similarly, fathers who worked flexible hours had less problems coping, less distress, less arguments, less work–family strains and were less rushed.

Working evenings/nights or weekends

For mothers, working either evenings/nights or weekends was associated with more work–family strain and more time pressure, while working evenings/nights was associated with mothers reporting more difficulties. Unexpectedly, working evenings/nights was also associated with better health outcomes for mothers.

For fathers, working weekends had no clear associations with wellbeing, with the one exception being couple fathers regularly working weekends who reported slightly more conflict with their partners. Working evenings/nights was associated with poorer wellbeing including having more difficulties, having less work–family gains and more work–family strains, and time pressure (being more rushed).

Job security and autonomy

The extent to which parents felt secure about their job was significantly associated with some measures of wellbeing. For mothers, greater job security was related to better coping, less distress and less work–family strain.

Greater job security showed strong and consistently favourable associations with all measures of fathers' wellbeing. For fathers, greater job security was associated with better health, less difficulties, better coping, less distress, being less rushed, more work–family gains and less work–family strains. For the relationship wellbeing measures, fathers reporting more security had significantly better scores; that is, less conflict and better relationship quality.

Job autonomy assesses how much freedom parents perceived they had to decide how they do their work. In line with previous research on job control or autonomy, consistently favourable associations were observed with wellbeing, for both mothers and fathers. Higher autonomy was associated with better health, psychological wellbeing (fewer difficulties, better coping and less distress), less arguments and better work–family spillover (fewer strains and more gains). Job autonomy was also associated with being less rushed and having a better quality relationship for employed fathers.

Based on this analysis, job security and job autonomy appear to have important favourable associations with wellbeing, particularly for fathers. Not only were the size of the effects for fathers some of the largest in the multivariate analyses, they were easily the strongest effects in terms of significance. For mothers, job autonomy appeared the more important of the two variables, with a more universal impact across all the measures of wellbeing compared to job security. Compared to fathers, though, the impacts of these variables were rarely as large or significant.

Summary

The analyses in this section compared variations in employed mothers' and fathers' wellbeing by work hours, job type and work conditions. All analyses were adjusted for a range of family, socio-demographic and work-related variables in order for independent associations with wellbeing to be explored. In reporting these associations, however, it needs to be kept in mind that the analyses are cross-sectional in nature and causality cannot be inferred. Indeed for some aspects of wellbeing, especially parents' health, it is likely that reciprocal relationships exist.

The key findings are outlined below.

back to top

8.4 Couple labour supply and wellbeing

This analysis extends the work in the previous section, and for couples only, assesses if mothers' or fathers' wellbeing is related to their partner's work hours. The approach is well founded in the literature, which shows that within families, the work patterns of one parent may affect the wellbeing of the other (termed cross-over effects) (Almeida 2004; Wethington 1999). For example, there is evidence that wives' mortality risk is linked to their husband's occupational risks (Fletcher 1991). Further, spouses' work stress can predict increased distress and negative mood in partners, pointing to the transmission of distress to other family members from work-related stresses (Bolger et al. 1990).

This section is based on multivariate analyses, in which associations with wellbeing were estimated, controlling for individual and family characteristics (as in previously presented models). Analyses also use the detailed work hours classification for each parent, and introduces their partner's work hours (detailed classification). Any partner effects, therefore, are adjusted for possible associations with own hours worked.71 This refers to Model Set 3 (See Box 8.2) and the detailed results are provided in Tables D9 to D12.

Table 8.5 presents the associations between mothers' wellbeing and partners' work hours, using predicted values, as before. Partner's hours were associated with some, although not all, measures of mother's wellbeing. Significant effects were evident for relationship quality, increased time pressure, having difficulties and higher distress. For all these indicators of wellbeing, mothers whose partner worked short full-time hours (35 to 44 hours) had better wellbeing than those whose partner worked other hours. Mothers whose partner was not employed were also less likely to report being time pressured, although they were also more likely to report poorer health.

Table 8.5: Predicted wellbeing: couple mothers, by partner's hours worked
Mothers' wellbeing measures(a)

Father's hours worked

Not
employed		 1–34 35–44 45–54 55 or more
 

Per cent

Fair to poor health (-) 9.9* 9.3* 8.4 6.9 6.7
Some to very many difficulties (-) 42.7 47.2 38.7*** 41.3* 45.7
Problems coping (-) 44.1 47.9 44.5 45.8 47.3
Time pressure (-) 41.1* 47.9 41.5** 44.9 47.7
 

Scale (1–5)

Psychological distress (-) 1.61 1.64 1.59** 1.64 1.65
Arguments with partner (-) 2.26 2.24 2.19 2.21 2.19
Relationship quality (+) 4.29 4.29 4.36* 4.32 4.30
Work–family gains (+) 3.48 3.60 3.63 3.62 3.58
Work–family strains (-) 2.74 2.76 2.71 2.74 2.73
Source: LSAC 2004, Wave 1.
Note: Figures refer to the predicted percentage or the predicted score, holding all control variables at their mean value (for all fathers) and varying only hours worked.
(a) Wellbeing measures as detailed in Box 8.1. Measures marked (+) indicate that a higher score means 'better' wellbeing. For those measures marked (-), a higher score means 'worse' wellbeing.
Significance refers to difference from the reference category, 55 hours or more.
*** p < 0.001, ** p < 0.01, * p < 0.05.

For self-rated health, poorer maternal health was likely in couples if the partner worked part-time hours or if they were not working, compared to mothers with a partner who worked longer hours. While this does not mean that partner's working part-time hours or not working at all caused poorer health of mothers, Baxter (2005b) found that women with partners who are not working because of poor health often have an ongoing medical condition themselves.

A similar analysis was undertaken for fathers, looking at whether partner's work hours were associated with their wellbeing (see Table 8.6). Compared to the findings for mothers, fathers' wellbeing was less consistently associated with their partner's work hours. All findings are in comparison with households where mothers worked full-time. Fathers reported less problems coping when mothers worked shorter part-time hours (either 1 to 15, or 16 to 24) or where mothers were not employed. In households where mothers worked between 16 and 24 hours each week, fathers reported fewer arguments and better relationship quality, more work–family gains and less work–family strains. However, fathers' time pressure was significantly worse when mothers worked either short part-time hours (1 to 15) or long part-time hours (25 to 34).

Fathers reported significantly less psychological distress when mothers worked short part-time hours, but their distress was unrelated to any other category. As well as the better relationship outcomes for fathers in households where mothers worked between 16 and 24 hours per week, compared to fathers whose partners worked full-time, better relationship quality was also reported when partners worked 1 to 15 hours per week or if partners were not employed. Similarly, fathers reported fewer arguments when partners were not employed.

Table 8.6: Predicted wellbeing: couple fathers, by mother's hours worked
Fathers' wellbeing measures(a)

Mother's hours worked

Not
employed		 1–15 16–24 25–34 35 or more
 

Per cent

Fair to poor health (-) 11.1 10.6 9.4 12.8 11.9
Some to very many difficulties (-) 45.3 45.0 44.1 39.8 45.8
Problems coping (-) 45.3** 43.4 *** 46.2* 49.5 52.8
Time pressure (-) 41.6 45.4* 39.4 45.3* 39.8
 

Scale (1–5)

Psychological distress scale (-) 1.58 1.55* 1.54 1.56 1.60
Arguments with partner (-) 2.17* 2.16 2.15** 2.21 2.22
Relationship quality (+) 4.41* 4.43* 4.45*** 4.32 4.34
Work–family gains (+) 3.65 3.69 3.72* 3.71* 3.64
Work–family strains (-) 2.83 2.76 2.73* 2.76 2.81
Source: LSAC 2004, Wave 1.
Note: Figures refer to the predicted percentage or the predicted score, holding all control variables at their mean value (for all mothers) and varying only hours worked.
(a) Wellbeing measures as detailed in Box 8.1. Measures marked (+) indicate that a higher score means 'better' wellbeing. For those measures marked (-), a higher score means 'worse' wellbeing.
Significance refers to difference from the reference category, 55 hours or more.
*** p < 0.001, ** p < 0.01, * p < 0.05.

back to top

8.5 Preferred hours and wellbeing

Following the approach developed by Gray et al. (2004), this section analyses employed parents' wellbeing according to their work hour preferences. The analyses include all employed parents (single and couple) based on Model Set 4 (Box 8.2). These models contains the same variables as Model Set 2, but also include one additional variable to test whether the wellbeing measures of employed mothers and fathers are associated with a parent's preference for more or less hours than they were working. The model details are provided in Tables D13 to D16.

As shown in Section 3.5, many parents preferred to be working different hours to those that they were working. This might have been because their current work hours were having a detrimental effect on some aspect of wellbeing. In the case of parents who wanted to work longer hours (underemployed parents), this preference may have been related to the financial situation of these families, specifically that they preferred more hours to increase their income. In fact, of primary carers (mostly mothers) who preferred to work more hours, 46.9 per cent rated their prosperity as 'just getting along', 'poor' or 'very poor', compared to 31.6 per cent of all primary carers. On the other hand, wanting to work fewer hours (overemployed parents), is unlikely to be driven by financial difficulties, but a consequence of other strains, such as those associated with time pressure.

Table 8.7 shows that preferred working hours were strongly related to wellbeing, even after controlling for these other variables. For mothers and fathers, preferring to work fewer hours was strongly associated with poorer wellbeing, compared to those who preferred the hours they worked. Mothers who preferred fewer hours had lower wellbeing on all measures, compared to those who preferred the hours they worked, although the association was not significant for self-reported health. Fathers had lower wellbeing on all measures also, but the effects on the relationship measures (arguments and relationship quality) were not significant.

Table 8.7: Predicted wellbeing: employed parents, by preferred working hours
Wellbeing measures(a) Prefer fewer Prefer same Prefer more
 

Mothers

 

Per cent

Fair to poor health (-) 6.4 5.2 8.4*
Some to very many difficulties (-) 53.5*** 41.3 53.7***
Problems coping (-) 53.0*** 41.2 52.4***
Time pressure (-) 64.4*** 49.0 50.5
 

Scale (1–5)

Psychological distress (-) 1.68*** 1.56 1.71***
Arguments with partner (-) 2.24* 2.19 2.31***
Relationship quality (+) 4.26** 4.35 4.10***
Work–family gains (+) 3.42*** 3.73 3.78
Work–family strains (-) 2.92*** 2.48 2.72***
 

Fathers

 

Per cent

Fair to poor health (-) 11.3** 8.1 12.1*
Some to very many difficulties (-) 49.5*** 40.3 53.5***
Problems coping (-) 49.7*** 42.4 51.7**
Time pressure (-) 54.3*** 38.3 40.9
 

Scale (1–5)

Psychological distress (-) 1.63*** 1.52 1.68**
Arguments with partner (-) 2.18 2.17 2.33***
Relationship quality (+) 4.40 4.42 4.30*
Work–family gains (+) 3.53*** 3.75 3.79
Work–family strains (-) 3.01*** 2.66 2.81**
Source: LSAC 2004, Wave 1.
Note: Figures refer to the predicted percentage or the predicted score, holding all control variables at their mean value and varying only the hours preference.
(a) Wellbeing measures as detailed in Box 8.1. Measures marked (+) indicate that a higher score means 'better' wellbeing. For those measures marked (-), a higher score means 'worse' wellbeing.
Significance refers to difference from the reference category, 55 hours or more.
*** p < 0.001, ** p < 0.01, * p < 0.05.

Preferring to work more hours was also associated with lower wellbeing, relative to a preference for current hours. The only measures in which the association was not significant (for both mothers and fathers) were work–family gains and being rushed. On these measures, those who preferred more hours were not significantly different to those who preferred the same hours. Preferring to work more hours also significantly increased work–family strains. This is possibly explained by there being some parents whose working hours are constrained by their family responsibilities. Such parents might have indicated they preferred more hours, and also have agreed more strongly with the statement 'I have to turn down work activities that I would prefer to take on', which is one of the elements of the work–family strain measure.

back to top

8.6 Child care use and wellbeing

In Section 4, the patterns of child care use for each cohort were described in some detail. Using LSAC, it is possible to look for associations between child care use and wellbeing, to investigate whether parents report having more difficulties, or other reduced (or increased) wellbeing, when they have a particular care arrangement. A preliminary analysis of these relationships showed that no such relationships existed for either cohort. This may be related to the vast majority of parents being satisfied with the child care they used, if they used any, and also, that the majority of those not using child care had no need for such care. Although such information indicates a high degree of satisfaction with child care use, LSAC was not designed to measure any discordance between preferred and actual use of child care (for example, using informal care but preferring formal care), which may have a stronger association with wellbeing.

This analysis was not progressed, given the lack of associations found in these data and lack of suitable data on child care preferences.

back to top

8.7 Summary

Maximising parent participation in the workforce will be increasingly important as the population ages. Given recent policy initiatives, the question of whether some work conditions or arrangements can further benefit families is timely. On the one hand, aspects of work that conflict with and strain family life or affect parent health could have adverse consequences for parents' wellbeing and family relationships. Poorer mental and physical health, or strains on family relationships could in turn serve as disincentives for sustained participation, as well as increasing health care costs, reducing productivity, and potentially affecting children. On the other hand, there is evidence that work can have positive effects on family life (Section 6).

Sections 6, 7 and 8 address these links between employment and family wellbeing, focusing in depth on family financial wellbeing (Section 6), the spillover (positive and negative) from work to family life (Section 7), and parents' health, psychological wellbeing, marital relationship, work–family strains and gains, and time pressures (Section 8). Section 6 shows the financial benefits that employment delivers to families. Jobless families with neither parent working reported the most financial hardship compared to families with one or both parents employed full-time.

For wellbeing, the key findings are:

These findings illustrate two key points. First, work, either in terms of the hours, conditions or arrangements between mothers and fathers, shows pervasive associations with wellbeing. In some instances the work patterns of one partner is also associated with the wellbeing of their partner. Second, the relationship between full-time employment and wellbeing differs for mothers and fathers, being largely positive for fathers but not for mothers. This suggests that mothers and fathers face different pressures at work and at home, but these data could not be used to assess why this was so. The types of jobs available to mothers compared to fathers and the sharing of the housework and care may all contribute; the findings underline the need for further research.

In Section 6 and in numerous other research, it has been shown that adequate family income is important to wellbeing. Findings from this section indicate that jobs offering parents flexible work hours, security, and autonomy are also likely to yield further payoffs to parent mental health and family relationships.

Parents' work hours were also associated with wellbeing, but in more complex ways. A key element is parents' preferences and control over when they work. Allowing parents to tailor their work hours to their circumstances, rather than mandating more or less hours, may be a preferable policy option. In this section and throughout the report, the mothers of young children show stronger preferences for, and better wellbeing outcomes from, working short to medium part-time hours (less than 25 hours per week). For fathers, the best and preferred outcomes are from jobs with short full-time hours (35–44 hours per week).

The association between mothers' poorer wellbeing and full-time employment suggests ongoing differences between men and women. This is possibly related to mothers' greater share of caring and domestic work, and supporting fathers to be more active carers could be a focus of policy initiatives. On a cautionary note however, although part-time work appears to be one way to support family wellbeing and help share the caring, it could also impose significant financial costs to mothers or fathers in the longer term in the form of lower superannuation benefits, which may work against other policy objectives of more self-funded retirement as the population ages.

Work–family strain is another measure of how difficult it is to combine work with caring responsibilities, and the analysis in this report suggests several ways strain could be reduced. For mothers and fathers with young children, working fewer hours, having flexible work hours, job security and autonomy appear to reduce strain. Findings suggest that evening/night work, and for mothers (but not fathers) weekend work, might also make it harder to combine work with caring. An analysis of labour market regulation may need to conduct cost–benefit analyses that include family outcomes along with business outcomes.

It is important to keep in mind that the analyses show correlations, not causality, and so interpretations should be made with caution. Reverse and reciprocal relationships between employment and wellbeing are possible. Furthermore, many of the work characteristics are measured with single items, reducing their reliability and potentially underestimating the strength of the association. Finally, although our findings are consistent with previous research on employment, job characteristics and mental and physical health, unobserved characteristics of the parents or the families may underlie the relationships identified here.

Return to top

9. Conclusion

7. Work–family spillover