Mothers' work-to-family strain in single and couple parent families: the role of job characteristics and supports.

Baxter, Jennifer ; Alexander, Michael

There is considerable evidence that for parents, aspects of work 'spill over' into family life, sometimes in a positive way (work-family gains) and other times in a negative way (work-family strains). This analysis focuses on some negative aspects of these work-to-family spillover effects for single and couple mothers.

While there has been increased focus on the work-to-family strain of mothers, given the increased participation of women in employment, less is known about single mothers and their experience of work-to-family strain. We might expect that single parents would have more difficulty in combining work and family responsibilities, given that they do not have the support of a partner to assist with childrearing and other household responsibilities.

This paper explores the relationships between work-to-family strain of employed mothers and a number of factors, broadly grouped as demographic, job characteristics and supports. Amongst these factors, we consider how single motherhood makes a difference. We also consider whether there are certain factors that are particularly pertinent to single mothers, either making the work-family balance significantly worse or better than it is for otherwise similar couple-parent mothers.

Background

Overview

By its very nature, the research surrounding work-to-family gains and strains recognises that the interplay between the work and home environments is crucial to the understanding of individual and family wellbeing. The concept of 'work-to-family spillover' recognises that the demands and resources--emotional and structural--of one domain can flow into and affect a person's fulfilment of a role in another domain. This 'flow' or spillover can be positive or negative, depending on whether it enhances or deters a person's fulfilment in the other role. This paper focuses on mothers' experiences of negative spillover from the work domain to the family domain.

Negative measures of work-to-family spillover capture the degree to which parents believe that family time is compromised by work responsibilities, and can include how work affects the amount or quality of family time. Various theoretical models have been developed to demonstrate the nature of work-to-family spillover and the possible antecedents of this type of spillover (for example, Barnett, 1998; Frone, Russell, & Cooper, 1992; Hill, 2005; Voydanoff, 2005a), although differences across studies in the manner in which the concept has been operationalised has no doubt contributed to the inconsistencies in some of the published results (Keene & Reynolds, 2005). The antecedents of work-to-family spillover are likely to relate to the demands and resources that exist in both the work domain and the family domain. Job characteristics, however, are usually found to be the more important in explaining work-to-family spill0ver than family factors (Keene & Reynolds, 2005), although both have been found to play a part.

In the work environment, factors associated with more work-to-family spillover include longer hours of employment, non-standard work schedules (such as evening work, weekend work, shiftwork, or excessive overtime) (Barnett, 1998; Mennino, Rubin, & Brayfield, 2005; Roehling, Jarvis, & Swope, 2005), the quality, complexity and skill-level of jobs, as well as the degree of flexibility and schedule control a worker has over their tasks (Keene & Reynolds, 2005; Mennino, Rubin, & Brayfield, 2005; Roehling, Jarvis, & Swope, 2005). Higher quality jobs are often associated with higher status occupations, and therefore higher status occupations may be associated with less work-to-family spillover. However, higher status jobs have also been linked to higher stress, which is likely to increase work-to-family spillover (Roehling, Jarvis, & Swope, 2005).

The nature of the employment contract has also been shown to be a factor. In earlier analyses of the LSAC data, Alexander and Baxter (2005) found that for partnered mothers, those in self-employment experienced the least work-to-family spillover, followed by those in casual jobs, with those in permanent jobs experiencing the most work-to-family spillover.

Work-to-family spillover is expected to be greater in families in which there are greater family demands. For example, the presence of young children or others needing higher levels of care is associated with more work-to-family spillover (Barnett, 1994; Barnett & Marshall, 1992a, 1992b). This strain can also be greater if a mother's ability to cope with the demands of work and family is reduced. Factors likely to be related to a reduced ability to cope with these demands are poor health, lower income and having language difficulties (Ciabattari, 2007).

Based on the existing literature, it is expected that having a higher level of support within the home would assist in managing both work and family responsibilities. One obvious source of support for women is a partner--whether married or cohabiting. Voydanoff (2005a) identifies spousal support (as well as spousal demands) as elements in a model of how work and family domains interact. Other research indicates that support received from partners and other family members appears to moderate the negative aspects of work-to-family spillover (Adams, King, & King, 1996; Eagle, Icenogle, Maes, & Miles, 1998; Westman & Etzion, 1995). This support can be invaluable to parents in jointly managing work and family responsibilities.

The obvious difference between employed single and couple mothers is that single mothers do not have a resident partner to share in the day-to-day management of the balancing of childrearing responsibilities with the demands of employment. Of course, couple mothers do not all receive the same amount of support from their partner in relation to childrearing. Likewise, there is some variation in the relationship single parents have with children's non-resident parent, with some receiving considerable support or sharing childrearing duties and others none at all. On balance though, as a result of the absence of a resident partner, we might expect that an employed single mother might experience more work-to-family strain than might an employed partnered mother.

Apart from partner support, single and couple mothers are also likely to differ in the types of jobs in which they are employed, as well as some demographic and family characteristics, and their levels of other types of support. For example, the education levels of single mothers are, on average, lower than of couple mothers (Australian Institute of Family Studies, 2005), and as a result single mothers are more likely to be employed in lower status occupations (Baxter, Gray, Alexander, Strazdins, & Bittman, 2007). In the context of this analysis, differing job characteristics and demographic and support variables may also contribute to differences in the overall levels of work-to-family strain experienced by those single and couple mothers who are in employment.

These differences between single and couple mothers are usually not explored in the existing literature. When single parents are included, they are not always separately identified in the analyses or the results (Frone, Russell, & Cooper, 1992; Hill, 2005). An exception is Mennino et al (2005), who analysed work-family strain for employed men and women, with and without children. They predicted that having a partner would increase work-family strain, as having a partner would make for a more complicated life. However, this analysis did not interact partner status with child-related variables (as childless workers were also included), so the results do not differentiate between single and couple mothers.

Morehead (2002) also noted the lack of research on work and family for single mothers. Her qualitative analysis of work and family arrangements experienced by single mothers working in a hospital identified a number of resources and supports that were important to single mothers in managing their work and domestic circumstances. These resources and supports were "jobs that are family-friendly, secure, and that provide a reasonable income; support from ex-partners and new partners that includes minding the children; help from family and friends particularly with child care and to provide emotional support; and access to further education" (Morehead, 2002, p. 61). Ciabattari's (2007) research on work-family conflict amongst single mothers in the US also demonstrated the importance of social, family and workplace supports. She emphasised the importance of workplace supports to single mothers, in particular access to flexible work hours (Ciabattari, 2007), while also noting the role played by extended family members who often live with single mothers. Hughes and Gray (2005) used New South Wales data collected in 2000 to compare the use of family-friendly work arrangements of single and couple mothers. As they discuss, these two groups of mothers are not only likely to vary, in the extent of support they have, they also differ in the types of jobs they work in. Their research showed that single and couple mothers reported about the same usage of flex-time as a means of combining work and family, but that single mothers were more likely than couple mothers to say they would have liked to make greater use of flex-time.

To help address some of the gaps in the literature, the above-mentioned work, individual, family and supports variables are considered in this analysis to identify factors that ameliorate or exacerbate negative work-to-family spillover for employed mothers. This analysis then leads to a more detailed examination of differences in work-to-family strain for single and couple mothers, to see if the same factors are at play, or if some factors have a more prominent effect on employed single mothers relative to their partnered counterparts.

By addressing these issues, we hope to provide valuable information with regard to how the work-family balance could be improved for all mothers, but also, if there are particular supports that could be encouraged to reduce work-to-family strain for single mothers. Reducing work-to-family strain would have benefits to employed parents, but also might encourage greater participation in employment, if some of the not-employed elect to remain out of employment because they wish to avoid work-to-family strain.

Data and method

Data

This paper uses data from the first wave of the Longitudinal Study of Australian Children (LSAC), collected in 2004 (see Soloff, Lawrence, & Johnstone, 2005 for details). LSAC aims to examine the impact of Australia's unique social, economic and cultural environment on children growing up in today's world. The essential focus of the study is on the early years of children's lives, and therefore 'the child' is the primary sampling unit of interest, although the collection of detailed personal and employment information from the child's resident parent(s) makes this a very useful dataset for work and family research.

The LSAC sample is broadly representative of all Australian children in each of two selected age cohorts: children born between March 2003 and February 2004 (infants) and children born between March 1999 and February 2000 (children aged 4-5 years). For this paper, both the infant and 4-5 year old cohorts were used, with mothers of these children the subject of the analysis.

The focus in this paper is on mothers who were at work and not on leave at the time of the survey (46 per cent of all mothers in the sample), with comparisons made between partnered mothers and single mothers. Partnered mothers include married and cohabiting mothers. Single mothers are mothers who are not living with a partner.

The sample of employed mothers available for analysis was 3,850. As single mothers were less prevalent than couple mothers, as well as being less likely to be employed than couple mothers (across both cohorts, 48 per cent of couple mothers were employed and 29 per cent of single mothers were employed), there were far fewer single mothers than couple mothers in this sample (289 compared to 3,561). (We discuss this difference in employment rates further below.)

The LSAC survey contains a number of subjective measures of how the work domain of parents 'spills over' into their family domain. Two standard items, originally taken from Barnett (1994), measure negative perceptions of how work spills over to family, or work-to-family strain. These are:

"Because of my work responsibilities I have missed out on home or family activities that I would like to have taken part in" and "Because of my work responsibilities nay family time is less enjoyable and more pressured".

Voydanoff (2005b) refers to two aspects of spillover as due to time-based demands and strain-based demands. The first of the measures used here relates to a time-based demand, in which work responsibilities result in an absence from the family and therefore a greater sense of missing out. The second relates more to a strain-based demand, in which time that is available to share with the family is negatively affected by the demands of work. Levels of agreement with these statements were measured on five-point Likert scales, from 'strongly disagree' (1) to 'strongly agree' (5) such that a higher value of the score indicates more work-to-family strain.

To first identify the differences between single and couple mothers, these measures of work-to-family strain for single and couple mothers were compared. Multivariate techniques were then used to determine which factors were associated with work-to-family strain. In addition to the single parent indicator, the models included a number of variables capturing job characteristics, demographics and supports (described below). Ordinary least squares regression was used and the analysis took into account the sample design and initial selection probabilities. Additional model specifications were tested, for example, using an ordered logit instead of OLS, to account for the categorical and ordered nature of the data. The results of the ordered logit analyses yielded comparable results, but as the OLS models were more easily interpreted, they were retained.

To identify, whether coefficients differed according to family type, these models were then re-estimated separately for single and couple mothers, as well as for all mothers but with all variables interacted with the single-parent indicator.

These findings were then considered in respect of the different characteristics of single and couple mothers, which are described below to consider whether any apparent difference in work-family spillover was due to those different characteristics, rather than single-parenthood per se.

Explanatory variables

A range of variables that characterise the domains of work and family were used to investigate the distribution of the two measures of work-to-family strain and included job characteristics, demographics and measures of support available to employed mothers.

Job characteristics pertaining to working hours were the usual weekly hours worked (grouped into 1 to 15, 16 to 24, 25 to 34, 35 or more hours), as well as scheduling information on whether work was sometimes done on weekends and whether work was sometimes done after 6pm or overnight. Information was also included on the flexibility of start and finish times (inflexible--could not change start and finish times, can change hours with approval, and flexible hours--can change hours without first seeking approval). Other job attributes included were job autonomy (level of agreement with "I have a lot of freedom to decide how I do my own work"), job security (rating of "how secure do you feel in your present job?"), the type of employment contract (self-employed, permanent or ongoing, casual, and fixed term or other), and the employee's broad occupation group (five groups were: (1) manager, professional, associate professional; (2) tradespersons and related workers, advanced clerical and service workers; (3) intermediate clerical, sales and service workers; (4) intermediate production and transport workers; and (5) elementary clerical, sales and service workers, labourers and related workers).

Single and couple mothers worked similar hours, had similar proportions sometimes working weekends, and similar ratings of their job's autonomy (Table 1). There were more differences in other job characteristics. Self-employment was a common employment type for employed couple mothers, but much less so for single mothers, who were more likely to be casual employees than were couple mothers. Single mothers were also less likely to work after 6pm or overnight and had lower perceived job security. Similar proportions of single and couple mothers reported to work inflexible hours, but single mothers were somewhat more likely to need approval to change their hours, rather than the greater degree of flexibility. Almost half of the employed couple mothers were in the higher status occupations (manager, administrator, professional or associate professional). While a sizable 31 per cent of single mothers were in this category, they were more likely than couple mothers to be in the lower status occupations.

Demographic information included mother's age, as well as age of the youngest child (under 1 year old, 1 year, 2-3 years, 4-5 years), the number of children (1, 2, 3 or more), the presence or otherwise of a child with a long-term medical condition, mothers' ratings of personal health (poor or fair health versus good, very good or excellent health), and main language spoken (English versus another language). The perceived financial wellbeing of the family was assessed by responses to the question "given your current needs and financial responsibilities, how would you say you and your family are getting on?". Those who responded "just getting along", "poor" or "'very poor" were classified as 'poor' and contrasted with those who indicated they were "reasonably comfortable", "very comfortable" or "prosperous".

A number of variables were included as possible sources of support, capturing potential support from the child's father, from other household members, and from a broader network of family and friends. The extent to which the father of the child assisted in the care of the child was incorporated by identifying families in which the father regularly cared for the child. The information source for this item depended on whether the father was resident or non-resident. For resident fathers, the source was mothers' responses, when asked if there were any regular times during the week that the father cared for the child while she was not there (for example, when at work or doing the shopping). Those who indicated "yes" were coded as 'other parent cares'. Note that this doesn't necessarily capture all times the father helps, since a subtext to this question says 'not just casual sharing of care that parents do for each other'. For non-resident fathers, the indicator was based on single mothers' responses, when asked how often the child's other parent looks after the child when she needs to do other things such as working, studying or attending appointments. Those who answered sometimes or often were also coded as 'other parent cares'. To capture possible support from other household members, the household composition was used to identify if a grandparent or another adult (but not grown-up children) lived with the family. For support from the wider network of family and friends, respondents were also asked about how they felt about the amount of support or help they got from family or friends living elsewhere. Those who said they didn't get enough help or didn't get any help at all were said to have an unmet need for help or support (as opposed to those who said they got enough help or didn't need any help).

There were some marked differences between employed single and couple mothers in their demographic and supports variables (Table 2). Employed single mothers were more likely to be in poor health, to be doing less well financially, and to have a child with a long-term medical condition. Employed single mothers were more likely to have only one child and to have older children compared to employed-partnered mothers.

Single mothers were expected to have less support available to them compared to couple mothers, given that they don't have a partner to share in the childrearing responsibilities. Based on the measure of 'other parent cares', over one-third of single mothers received assistance with childrearing from the child's non-resident parent (34 per cent). In comparison, 61 per cent of the partnered mothers indicated that their partner regularly spends time caring for the child. As outlined earlier, these data were based on different questions, so are not directly comparable. Single mothers were more likely to have the child's grandparent or another adult (not including adult siblings of the study child) living with them. The proportion of employed mothers who reported having an unmet need for help was the same for single and couple mothers (20 per cent).

An argument could be made that these supports variables are not useful in explaining variation in work-family strain, because lower levels of support would always be expected to be associated with greater work-family strain. This, however, is not necessarily the case, especially since we are considering support from a range of places. Understanding how these supports relate to differences in the perception of work-family strain is valuable.

A note about employed single and couple mothers

While the difference in employment rates of single and couple mothers is not a focus of this paper, to some extent, this difference is likely to be explained by the different characteristics of single and couple mothers, such as single mothers having, on average, lower levels of education. In the context of an analysis of work-to-family strain, however, employment rates are also likely to be influenced by the capacity of mothers to manage work-to-family strain, which in turn will be influenced by the nature of the jobs available, the supports a mother has both in the workplace and at home, and individual circumstances (Keene & Reynolds, 2005). At the extreme, difficulties with managing both work and family responsibilities would be evidenced by some women withdrawing from employment altogether, to avoid the high work-to-family strain that would occur if they did work. This may be more influential in the employment decision for single mothers, and may be one reason for the lower employment rate amongst single mothers compared to couple mothers (Ciabattari, 2007; Hughes & Gray, 2005).

As a result, it is possible that the employed mothers in LSAC are a select group of mothers who are able to satisfactorily balance work and family responsibilities, with the higher employment rate of couple mothers reflecting that it is easier for them to combine work and family responsibilities. As the analyses presented in this paper focus only on employed women, this issue is not explored further. It is, however, an important point in considering how these analyses are applicable to the wider population of not employed as well as employed single and couple mothers with young children.

Results

Work-family strain: employed single and couple mothers

To consider whether employed single mothers experience more work-to-family strain than couple mothers, the distributions of the work-to-family strain measures were initially compared. Single mothers were more likely than couple mothers to agree that because of their work responsibilities they had missed out on home or family activities that they would have liked to have taken part in. Also, single mothers were more likely to 'strongly agree' that their family time was less enjoyable and more pressured because of their work responsibilities (see Table 3). In both cases, the difference in the mean scores of single and couple mothers was significant at least at the 0.05 level. Considered in isolation from other factors, these data suggest that single mothers experience more work-to-family strain on average than couple mothers. However the differences are not large in size. Considering that a 1-unit difference in the scores would signify the difference between adjacent response categories, a difference of 0.2 to 0.3 appears fairly marginal.

Multivariate analysis of work-family strain

To test whether differences in work-to-family strain were evident after controlling for different characteristics of single and couple mothers, both measures of work-to-family strain were regressed against single parent status, along with the other demographic variables, supports variables and job characteristics. This section discusses the associations between each of these factors and work-to-family strain and the results are shown in Table 4.

Demographics

Some demographic characteristics were very important in explaining work-to-family strain, although single parenthood is not one of them. After controlling for other demographic variables, supports variables and job characteristics, there was no longer a significant difference between single and couple mothers for either measure of work-to-family strain. In fact, the reduction in the significance of single parent status was primarily due to the inclusion of the other demographic variables, rather than the support or job characteristic variables (analysis not shown). This suggests there is little evidence that once other factors are controlled, single parenthood in itself is associated with large differences in work-to-family strain.

The associations between other demographics and work-to-family strain indicated that there was more strain when the family perceived they were poorer and the mother in worse health. Older mothers and those with a child with a long-term medical condition also were more likely to perceive that their family time was more pressured or less enjoyable because of work.

Work-to-family strain was greatest in families with older children for the 'missed out at home' item. This may be a genuine effect: perhaps once children reach older ages, parents are more aware of what they are missing out on when at work. There are, however, other plausible reasons for this effect. It may be a selection effect: that mothers of very young children stay at home, rather than go to work, if the available job offers a poor fit between work and family. The effect may also be due to the question wording which, being in the past tense, might elicit more positive reports amongst those who have been caring for children for longer.

Supports

The extent of support received by mothers was also important in explaining work-to-family strain, with mothers who reported having an unmet need for help or support having higher levels of work-to-family strain. This was highly significant, and a relatively large effect across both measures.

Having the child's father sometimes helping with care had different associations with the two measures. Mothers reported being more likely to have missed out at home when the father sometimes helped with care. This may reflect that fathers are more likely to help with the care when the mothers' work-family strain is greater. However, having the father sometimes provide care was not associated with higher levels of work-family strain as measured by the measure of time pressure--it was associated with reduced levels of work-family strain. Another possible reason for the first finding is that mothers are more aware of the family time they are missing out on when the children spends time with the father. The finding for the second measure of work-family strain then is likely to reflect that this assistance with childcare from the father helps to reduce the sense of time pressure.

Residing with the child's grandparent, or having another adult in the home, did not have a significant association with work-family strain. Variations on these two measures were also explored, taking into account whether or not these people had long-term medical conditions, but even then, no significant relationships were evident.

Job characteristics

Overall, most of the job characteristics were important in explaining the variation in work-to-family strain. In particular, working longer hours was associated with more work-to-family strain. Having access to flexible hours, having a job with more control and being in casual employment were factors related to less work-to-family strain. Other factors were important to one of the measures of work-to-family strain but not the other.

Specifically, working longer hours was strongly and positively associated with both work-to-family strain measures, with mothers working full-time (compared to mothers working short part-time hours) reporting strain at a whole category higher in terms of 'missing out at home' and half a category higher in terms of 'family time is less enjoyable'. As one of the strongest effects in the model, this result identifies hours worked as one of the key influences on these measures of work-to-family strain.

Potential family time can be cut short if mothers are working weekends, so it is not surprising that working weekends was also associated with a greater likelihood of reporting to be missing out at home. It was not, however, related to family time being less enjoyable or more pressured, which is consistent with weekend work being more associated with an absence froth the family domain rather than being time pressured within the home.

Evenings are usually busy times for families with young children, and if parental employment encroaches into this time, it is easy to imagine how this can increase the time pressure that parents are under to get done those aspects of parenting that are fixed and not easily substituted with alternative supports or market solutions. It is therefore not surprising that those who sometimes worked evenings or nights were more likely to agree that their work responsibilities meant their family time was less enjoyable and more pressured. However, evening work was not a significant influence on parents feeling liked they had missed out on home or family activities. This may be because parents who work evenings or nights have time available at other parts of the day in which they can spend time with children, such that they do not experience a sense of missing out.

In contrast to the difficulties of weekend or evening work, the ability to easily change start and finish times can provide a useful support for employed mothers in managing both the absence from the family and the additional pressures (particularly time pressures) that employment can bring. Working flexible hours (and not having to seek approval to change times) was associated with less strain on both work-to-family spillover measures (compared to having inflexible start and finish times) and was stronger in ameliorating the effects of family absence. This is presumably because some flexibility in the work schedule allows mothers to more easily schedule parenting responsibilities.

In a similar vein to flexible hours, higher levels of job autonomy were associated with less work-to-family spillover, since greater autonomy is likely to be associated with more capacity for and control over scheduling of work to fit around family demands. In contrast, job security, whilst important for other aspects of wellbeing (Baxter, Gray, Alexander, Strazdins, & Bittman, 2007), did not have a significant relationship with these two measures of work-to-family spillover.

Self-employment, relative to permanent employment, was associated with a lower reporting of missing out at home, presumably because some of this work can be done at home or with children. Relative to permanent employment, casual work was also associated with less spillover from work to family, both in terms of missing out at home and home being less enjoyable and more pressured. Casual employment for mothers of young children may offer greater flexibilities around managing work-to-family strain than does permanent ongoing employment. This is an interesting result because in other research, for fathers in these same families, casual employment was found to increase work-to-family strain (Alexander & Baxter, 2005).

The inclusion of broad occupational groupings in the model is designed to examine the association with work-to-family strain of job of different status, as well as act as a control and capture any residual effect of job characteristics not already controlled for by the proceeding variables. The analysis finds that relative to low status jobs, mothers employed in the highest status jobs (managers, professionals, associate professionals) were more likely to report their family time to be less enjoyable and more pressured because of work responsibilities independent of the other job characteristics already controlled for. It is possible that this reflects these jobs involving more job stress and pressure.

Single-couple mother interactions

To determine whether any effects were stronger or weaker for single mothers than couple mothers, these two models were re-estimated with each factor interacted with single parent status. There were very few significant effects, which is very likely a consequence of the relatively small sample size for employed single mothers compared to employed couple mothers. These results are given in Table 5 (in the Dif column). To see the extent of differences in effect between single and couple mothers, separate models were also estimated for each measure. The results of these models are also shown in Table 5. Just as the interaction terms were often not significant, the terms in the single mother models were often not significant, again, no doubt due to small sample size.

Only three sets of variables were significantly different between single and couple mothers, and were primarily associated with one measure--that of missing out on activities at home. Firstly, for single mothers, access to flexible hours or the ability to change hours with approval had a significantly stronger effect (relative to having inflexible hours) on reducing this type of work-to-family strain. Secondly, the difference between the highest occupation group and the lowest occupation group in terms of missing out at home was significantly greater in single mother families. And thirdly, the age of youngest child effects were greater for single than couple mothers.

Unfortunately the small sample size appears to have reduced the ability to identify other factors that differ between single and couple mothers. The effects that were identified, however, are interesting. The fact that flexible hours had a more beneficial effect on single mothers' experience of missing out at home is consistent with Ciabattari's (2007) findings, that such workplace supports are likely to be particularly beneficial to single mothers.

Implications of differences in single and couple mothers' characteristics

Given these associations between demographic, supports and job variables and work-family strain, we can then consider whether differences in characteristics of single and couple mothers could lead to different experiences of work-to-family strain. (These differences were identified in the Data section.)

There were quite large difference between single and couple mothers in relation to their demographic and family characteristics. Single mothers had poorer health and lower perceived prosperity than couple mothers, and these factors were both significantly related to work-to-family strain. The poorer status (in health and perceived prosperity) of single mothers therefore increases their likelihood of experiencing work-to-family strain relative to couple mothers. Also, single mothers in this sample were more likely to have older children, compared to couple mothers, and having missed out on family activities due to work was more likely amongst mothers with older children. This also means single mothers would experience more of this work-family strain.

Looking at support, the greatest difference was that couple mothers were more likely to have the child's other parent provide care sometimes, than were single mothers. However, the results above showed that for couple mothers this support was associated with a greater sense of missing out at home, although was associated with a lower reporting of time pressure. Single mothers would be more likely to be time pressured, in part because of this lower level of parental support. Couple mothers, however, may have a greater sense of missing out because of work, since they are more likely to be in a position of having the other parent sometimes provide care.

Single mothers were more likely to have the child's grandparent or another adult (not including adult siblings of the study child) living with them. However, the multivariate analyses showed that mothers who had another adult living in the home did not have any less (or more) work-to-family strain than those who did not. Surprisingly, in total these supports measures did not appear to contribute to differences in work-to-family strain of single and couple mothers. It was expected that for single mothers the lack of support of the father and the added support of other resident family members would make a difference to work-to-family strain, but as measured here, these variables had little or no association.

In terms of job characteristics, there were a few differences between employed single and couple mothers. In particular, single mothers were more-often casual employees and less-often self-employed, compared to couple mothers; they were more likely to be employed in lower status jobs; they were less likely than couple mothers to work evenings and were less likely to have flexible working hours, being more likely to need to seek approval to change work times. These differences contribute in various ways to work-to-family strain differences. For couple mothers, according to the multivariate analyses, having higher use of self-employment and flexible hours would be of benefit in reducing their work-to-family strain, but being more likely in higher status occupations and undertaking work in the evenings increases strain for some measures. These differences are of particular interest in ensuring single mothers have access to the workplace supports that help in managing work and family. In particular, having access to more flexible hours, it appears, would be of benefit.

Discussion

This paper set out to explore the differences in work-to-family strain of employed single and couple mothers, both in regard to work responsibilities meaning missing out on home of family activities, and in terms of family time being less enjoyable and more pressured. Unexpectedly, before controlling for other variables, the difference in the experience of work-to-family strain between single and couple mothers was not large.

One possible explanation for the smaller than expected difference lies in the fact that employed single mothers are a more select subgroup of all single mothers than employed couple mothers are of all couple mothers. Perhaps the potential work-to-family strain for single mothers with young children is more of a barrier to employment than it is for couple mothers with young children. On the other hand, those in paid employment have somewhat similar experiences of work-to-family strain because those who face potentially excessive strain may withdraw from or not enter employment in the first instance.

The employment characteristics of single and couple mothers were similar, especially in regard to some of the variables that had strong relationships With work-to-family strain, such as hours of employment and level of job autonomy. There were, however, other differences that were more important in explaining the work-to-family strain of single mothers relative to couple mothers. One of these was access to flexible working hours that could be changed without needing prior approval. Improving access to more flexible hours may well reduce experiences of work-to-family strain for all mothers, but for single mothers in particular.

Another important factor in describing the work-to-family strain of employed mothers was whether they expressed an unmet need for help or support--those who reported an unmet need experienced more strain. Of all the supports variables included in this analysis, this had the strongest association with work--family strain, but it is interesting that the proportion expressing this unmet need for help was similar for single and couple mothers. The degree of supports in the community or extended family may also have strong associations with the likelihood of being employed, as some women may not enter employment unless they have help or support to back them up to attempt combining work and family. There was, not surprisingly, more difference between single and couple mothers in the proportion who had assistance from the child's other parent in providing care sometimes. While this assistance appeared to alleviate some of the pressure within families, it was associated with a heightened sense of missing out.

It should be noted that the measures of work-family strain used in this paper are subjective, capturing the degree to which mothers perceive their work spills over to family, as experienced as 'missing out', or being "more pressured and less enjoyable' in the family domain. Responses to such items may vary across individuals according to a range of characteristics not captured in this analysis, and may indeed vary according to recent events that elicit more positive or negative responses.

Overall, the work-to-family strain of employed mothers is much more strongly determined by employment characteristics, demographics, and selected supports variables, than it is by single parent status. Having a resident partner does not automatically reduce the work-to-family strain for employed mothers. However, having a resident partner is associated with a higher likelihood of being employed. It is possible that potential work-to-family strain has a stronger impact on those with no partner, which contributes to decisions about staying out of employment.

References

Adams, G. A., King, L. A., & King, D. W. (1996). Relationships of job and family involvement, family social support, and work-family conflict with job and life satisfaction. Journal of Applied Psychology, 76, 60-74.

Alexander, M., & Baxter, J. (2005). Impacts of work on family life among partnered parents of young children. Family Matters(72), 18-25.

Australian Institute of Family Studies. (2005). House of Representatives Standing Committee on Family and Human Services Inquiry into Balancing Work and Family submission from the Australian Institute of Family Studies. Melbourne, Vic.: Australian Institute of Family Studies.

Barnett, R. C. (1994). Home-to-Work Spillover Revisited: A Study of Full-Time Employed Women in Dual-Earner Couples. Journal of Marriage and the Family, 56(3), 647-656.

Barnett, R. C. (1998). Toward a review and reconceptualization of the work/ family literature. Genetic, Social & General Psychology Monographs 124(2), 125-182.

Barnett, R. C., & Marshall, N. L. (1992a). Men's Job and Partner Roles: Spillover Effects and Psychological Distress. Sex Roles, 27(9-10), 455-472.

Barnett, R. C., & Marshall, N. L. (1992b). Worker and Mother Roles, Spillover Effects, and Psychological Distress. Women and Health, 18(2), 9-40.

Baxter, J., Gray, M., Alexander, M., Strazdins, L., & Bittman, M. (2007). Mothers and fathers with young children: Paid employment, caring and wellbeing (Social Policy Research Paper No. 30). Canberra: Department of Families, Community Services and Indigenous Affairs.

Ciabattari, T. (2007). Single Mothers, Social Capital, and Work-family Conflict. Journal of Family Issues, 28(1), 34-60.

Eagle, B. W., Icenogle, M. L., Maes, J. D., & Miles, E. W. (1998). The Importance of Employee Demographic Profiles for Understanding Experiences of Work-family Interrole Conflicts. Journal of Social Psychology, 138(6), 690-709.

Frone, M. R., Russell, M., & Cooper, M. L. (1992). Antecedents and outcomes of work-family conflict: Testing a model of the work-family interface. Journal of Applied Psychology, 77(1), 65-78.

Hill, E. J. (2005). Work-family Facilitation and Conflict, Working Fathers and Mothers, Work-family Stressors and Support. Journal of Family Issues, 26(6), 793-818.

Hughes, J., & Gray, M. (2005). The use of family-friendly work arrangements by lone and couple mothers. Family Matters(71), 18-23.

Keene, J. R., & Reynolds, J. R. (2005). The Job Costs of Family Demands: Gender Differences in Negative Family-to-Work Spillover. Journal of Family Issues 26(3), 275-299.

Mennino, S. E, Rubin, B. A., & Brayfield, A. (2005). Home-to-job and job-to-home spillover: The impact of company policies and workplace culture. Sociological Quarterly, 46(1), 107-135.

Morehead, A. (2002). Behind the paid working hours of single mothers: managing change and constructing support. Family Matters(61), 56-61.

Roehling, P. V., Jarvis, L. H., & Swope, H. E. (2005). Variations in Negative Work-family Spillover Among White, Black, and Hispanic American Men and Women: Does Ethnicity Matter? Journal of Family Issues, 26(6), 840-865.

Soloff, C., Lawrence, D., & Johnstone, R. (2005). Sample Design (LSAC Technical Paper No. 1). Melbourne: Australian Institute of Family Studies.

Voydanoff, P. (2005a). Toward a Conceptualization of Perceived Work-family Fit and Balance: A Demands and Resources Approach. Journal of Marriage & the Family, 67, 822-836.

Voydanoff, P. (2005b). Work Demands and Work-to-family and Family-to-Work Conflict: Direct and Indirect Relationships. Journal of Family Issues, 26(6), 707-726.

Westman, M., & Etzion, D. (1995). Crossover of Stress, Strain and Resources from One Spouse to Another. Journal of Organizational Behavior, 16(2), 169-181. Table 1: Job characteristics, employed couple and single mothers Couple Single Total Per cent Usually work 1 to 15 37.2 34.2 37.0 hours per week 16 to 24 hours 26.1 26.9 26.2 25 to 34 hours 14.3 14.6 14.3 35 hours or more 22.4 24.3 22.6 Sometimes works 54.2 52.1 54.0 weekends Sometimes works 52.1 42.6 51.4 evenings or after 6pm Permanent employee 48.7 51.7 49.0 Casual employee 19.5 30.8 20.4 Fixed-term or 5.7 7.7 5.9 other employee Self-employed 26.0 9.8 24.7 Can change work times: 56.9 47.2 56.1 flexible hours Can change times with 26.0 35.5 26.8 approval Cannot change work hours 17.1 17.3 17.1 Manager, administrator, 47.3 31.4 46.0 professional, associate professional Tradespersons and 15.1 10.0 14.7 related workers; advanced clerical and service workers Intermediate clerical, 22.9 34.9 23.8 sales and service workers, intermediate production and transport workers Elementary clerical, 14.8 23.7 15.5 sales and service workers, labourers and related workers Mean Level of job security 3.0 2.7 2.9 (from least to highest security, 1 to 5) Autonomy at work (from 3.7 3.6 3.7 least control to most control over the work you do, 1 to 6) Table 2: Demographic characteristics and supports, employed couple and single mothers Couple Single Total Mean Age (years) 33.8 32.9 33.8 Per cent Single mother 0.0 100.0 8.0 Has poor health 6.6 10.5 6.9 Financially 'just getting along', 'poor' 'or very poor' 281.0 53.9 30.1 Main language is not English 9.9 7.1 9.7 A child has long-term medical condition 22.5 33.3 23.3 Youngest child aged <1 37.4 16.1 35.7 Aged 1 15.4 9.4 14.9 Aged 2-3 14.8 12.8 14.6 Aged 4-5 32.4 61.8 34.8 1 child in the family 25.1 36.8 26.0 2 children 47.7 40.7 47.1 3 or more children 27.2 22.6 26.8 Other parent helps with caring for child sometimes 60.8 34.4 58.6 Has unmet need for help/ support 20.4 19.5 20.3 A grandparent lives with the family 3.3 12.6 4.1 Other adults (not older siblings) live with the family 2.1 13.6 3.0 N 3,558 289 3,847 Table 3: Work-to-family spillover, employed single and couple mothers Strongly disagree Disagree Neither (1) (2) (3) Per cent Because of my work responsibilities ... I have missed out on home or family activities that I would like to have taken part in Couple 17.3 30.8 12.5 Single 11.8 26.3 14.3 difference My family time is less enjoyable and more pressured Couple 22.2 37.8 17.8 Single 19.0 34.6 19.0 difference Strongly Agree Agree (4) (5) Total Mean SD Because of my work responsibilities ... I have missed out on home or family activities that I would like to have taken part in Couple 28.9 10.5 2.84 1.30 Single 33.2 14.5 3.12 1.28 difference *** My family time is less enjoyable and more pressured Couple 18.1 4.2 2.44 1.14 Single 17.2 10.2 2.64 1.25 difference * Legend: # p<0.1, * p<0.05, ** p<0.01, *** p<0.001. Table 4: Work-to-family and family-to-work spillover main effects (OLS coefficients), employed mothers Work-to-family spillover Family I have time missed is less Work-to-family spillover out at home enjoyable Constant 2.38 *** 1.96 *** Couple mother, partner employed (reference) Couple mother, partner not employed 0.19 -0.02 Single mother 0.08 0.11 Just getting along, poor or very poor 0.28 *** 0.23 *** Poor health mother 0.40 *** 0.44 *** A child has medical condition 0.09 (#) 0.15 ** English is not main language -0.05 0.00 Age mother (centred at 33.6 years) 0.00 0.02 *** Youngest child aged <1 (reference) Youngest child aged 1 -0.04 0.05 Youngest child aged 2-3 0.05 0.03 Youngest child aged 4-5 0.22 *** 0.08 1 child in the family (reference) 2 children in the family 0.03 0.08 (#) 3 or more children in the family 0.09 0.13 * Has unmet need for help/support 0.23 *** 0.35 *** Father helps with caring for child sometimes 0.10 * -0.11 ** Grandparent lives with family 0.10 -0.03 Other adult lives with family -0.08 -0.04 1 to 15 hours (reference) 16 to 24 hours 0.41 *** 0.17 *** 25 to 34 hours 0.65 *** 0.29 *** 35 hours or more 0.94 *** 0.54 *** Sometimes works weekends 0.22 *** 0.08 (#) Sometimes works nights -0.05 0.10 * Level of job security (centred at 2.95) 0.01 -0.02 Autonomy at work (centred at 3.73) -0.09 *** -0.07 *** Permanent employee (reference) Self-employed -0.34 *** -0.06 Casual employee -0.19 *** -0.13 * Fixed-term and other employee -0.15 0.05 Cannot change work times (reference) Can change work times: flexible hours -0.39 *** -0.19 *** Can change times with approval -0.11 (#) -0.09 Manager, administrator, professional, associate professional 0.05 0.18 ** Tradespersons and related workers, advanced clerical and service workers -0.08 0.04 Intermediate clerical, sales and service worker intermediate production and transport workers -0.11 0.09 Elementary clerical, sales and service workers, labourers and related workers (reference) N 3698 3680 R-square 0.23 0.14 Legend: # p<0.1, * p<0.05, ** p<0.01. *** p<0.001 Table 5: Work-to-family spillover, employed single and couple mother I have missed out at home Couple Single Dif Constant 2.40 *** 2.44 *** Couple, partner is not employed 0.18 n. a. Just getting along, poor, very poor 0.28 *** 0.23 Poor health mother 0.40 *** 0.45 (#) A child has medical condition 0.10 (#) -0.12 English is not main language -0.03 -0.36 Age mother (centred at 33.6 years) 0.00 0.00 Youngest child aged <.1 (reference) Youngest child aged 1 -0.07 0.50 (#) * Youngest child aged 2-3 0.02 0.55 (#) * Youngest child aged 4-5 0.19 *** 0.61 ** * 1 child in the family (reference) 2 children in the family 0.05 -0.10 3 or more children in the family 0.13 * -0.13 Has unmet need for help/support 0.23 *** 0.21 Father helps with caring for child 0.11 * 0.02 sometimes Grandparent lives with family 0.07 0.11 Other adult lives with family -0.15 0.08 1 to 15 hours (reference) 16 to 24 hours 0.41 *** 0.41 * 25 to 34 hours 0.67 *** 0.54 * 35 hours or more 0.95 *** 0.87 *** Sometimes works weekends 0.19 *** 0.44 * Sometimes works nights -0.03 -0.25 Level of job security (centred at 2.95) 0.01 0.02 Autonomy at work (centred at 3.73) -0.10 *** 0.00 Permanent employee (reference) Self-employed -0.35 *** -0.10 Casual employee -0.18 ** -0.23 Fixed-term and other employee -0.16 (#) -0.13 Cannot change work times (reference) Can change work times: flexible -0.35 *** -0.82 *** * hours Can change times with approval -0.05 -0.60 * * Manager, administrator, professional, 0.01 0.43 * * associate professional Tradespersons and related workers; -0.13 0.35 (#) advanced clerical and service workers Intermediate clerical, sales and -0.16 * 0.25 (#) service workers, intermediate production and transport workers Elementary clerical, sales and service workers, labourers and related workers (reference) N 3427 263 R-square 0.23 0.29 Family time is less enjoyable Couple Single Dif Constant 1.93 *** 2.09 *** Couple, partner is not employed -0.02 n. a. Just getting along, poor, very poor 0.26 *** -0.01 Poor health mother 0.43 *** 0.53 A child has medical condition 0.17 *** -0.16 (#) English is not main language 0.02 -0.11 Age mother (centred at 33.6 years) 0.02 *** 0.00 Youngest child aged <.1 (reference) Youngest child aged 1 0.02 0.50 (#) (#) Youngest child aged 2-3 0.00 0.35 Youngest child aged 4-5 0.06 0.36 1 child in the family (reference) 2 children in the family 0.12 * -0.28 * 3 or more children in the family 0.15 * 0.12 Has unmet need for help/support 0.33 *** 0.72 ** Father helps with caring for child -0.10 * -0.17 sometimes Grandparent lives with family 0.03 -0.40 Other adult lives with family -0.17 0.24 1 to 15 hours (reference) 16 to 24 hours 0.17 ** 0.18 25 to 34 hours 0.31 *** 0.13 35 hours or more 0.53 *** 0.65 * Sometimes works weekends 0.07 (#) 0.05 Sometimes works nights 0.10 * 0.07 Level of job security (centred at 2.95) -0.02 -0.08 Autonomy at work (centred at 3.73) -0.07 *** -0.02 Permanent employee (reference) Self-employed -0.07 0.21 Casual employee -0.15 ** 0.02 Fixed-term and other employee 0.06 -0.23 Cannot change work times (reference) Can change work times: flexible -0.19 ** -0.14 hours Can change times with approval -0.07 -0.27 Manager, administrator, professional, 0.17 ** 0.20 associate professional Tradespersons and related workers; 0.06 -0.12 advanced clerical and service workers Intermediate clerical, sales and 0.08 0.24 service workers, intermediate production and transport workers Elementary clerical, sales and service workers, labourers and related workers (reference) N 3411 261 R-square 0.14 0.23 Legend: (#) p<0.1, * p <0.05, ** p<0.01, *** p<0.001.