Leave Duration After Childbirth

Abstract

In this article, we use longitudinal data on Australian two-parent families to assess the impact of mothers’ leave duration after childbirth on their subsequent mental health, quality of parenting, and couple relationships. We found that 2 to 3 years after the birth, psychological distress was significantly less likely among mothers who took >13 weeks paid leave or >26 to 52 weeks leave in total. However, those who took >26 weeks in total were more likely to register as “distant” in their parenting relationship. The analysis also highlighted the importance of leave supports at the workplace: Mothers who experienced workplace problems during their pregnancy were more likely to report feeling rushed and to rate their couple relationship as unhappy or argumentative 2 to 3 years later. Overall, our study underlined the complexity of leave impact but affirmed the importance of leave duration for maternal mental health and leave takers’ workplace supports for maternal and family outcomes.

Keywords

parental leave maternal mental health parenting couple relationship Australia

Among the most pervasive social changes over the past half century are widespread increases in women’s labor force participation during childbearing years and associated declines in sole (male) breadwinner arrangements in couple families. Concerns about the possible impact of mothers’ labor force participation on maternal and child health and well-being have been among many pressures leading governments to adopt policies that enable parents to take leave from employment around the time of their child’s birth. However, the impact of such leave arrangements on mothers and children remains unclear, perhaps unsurprisingly given the complexity of the relationships involved, particularly for children’s outcomes. We argue that an essential step in achieving a better understanding of the impact of leave is to know more about its associations with aspects of maternal and family functioning—outcomes important in themselves as well as potentially for child health and well-being. Using three waves of data from Australian two-parent families, we seek to contribute to this field of research by investigating whether the duration of paid and total leave taken by mothers is associated with mothers’ mental health, quality of parenting, and couple relationships 2 to 3 years after the child’s birth.

Impacts of Parental Leave on Maternal and Child Health and Well-being: Review of Evidence

Although a fundamental objective of maternity leave since its earliest manifestations has been the enhancement of maternal and child health (Kamerman, 2000), there are considerable gaps in the evidence to date on links between leave taking and such outcomes. Much research on children’s outcomes focuses on the impact of maternal employment during the child’s first year, without directly assessing the type and duration of leave taken during this period. Where research does focus directly on leave, analysis is sometimes limited to the presence of national entitlements rather than individual-level leave usage, while microlevel studies of leave impact have delivered mixed findings. We elaborate first on research relating to children’s development and physical health before turning to studies of maternal well-being and family functioning.

The influence of parental leave on child development has proved particularly difficult to demonstrate given the number of potential influences and the extended time period necessary for meaningful assessment. Negative associations between maternal employment in the first year of a child’s life and children’s subsequent cognitive and behavioral development have been found in some studies, but results overall have been mixed (e.g., Brooks-Gunn, Han, & Waldfogel, 2002; Gregg, Washbrook, Propper, & Burgess, 2005; Han, Waldfogel, & Brooks-Gunn, 2001; Lucas-Thompson, Goldberg, & Prause, 2010; Ruhm, 2004; Waldfogel, Han, & Brooks-Gunn, 2002). A recent U.S.-based study (Brooks-Gunn, Han, & Waldfogel, 2010) underlined this complexity, showing that the impact of maternal employment in the first year on children’s cognitive, social, and emotional development to 7 years of age was effectively neutral because of the combination of positive and negative effects. The influence of parental leave remains largely unobserved in such studies, although its importance is sometimes suggested. For example, while a recent study of British and American longitudinal data found limited adverse effects of early maternal employment, the one negative association identified (with reading comprehension at age 4-14 years) was in the United States, leading the authors to observe that this may reflect the “shorter and less generous” maternity leave arrangements in the United States (Joshi, Cooksey, & Verropoulou, 2009, p. 19). However, this was not tested and direct evidence of association between parental leave duration and child development is limited. In Canada, where paid leave was extended from 25 to 50 weeks in 2000, analysis identified no significant associations between mothers’ increased time away from work and measures of children’s motor and social development up to 2 years of age (Baker & Milligan, 2010).

While the impact of parental leave has been more directly assessed in relation to children’s physical health, results have also been mixed. Cross-national analyses using aggregated data have consistently identified associations between national provisions for paid and job-protected parental leave and lower infant mortality rates (Ruhm, 2000; Tanaka, 2005; Winegarden & Bracy, 1995). The role of breast-feeding, which is recognized as both conducive to maternal and child health and positively associated with the generosity of parental leave entitlements (Galtry, 2003), may help explain these results. Several microdata analyses have linked longer leave periods with higher breast-feeding incidence and duration, and in some cases with health-related indicators, such as immunization (e.g., Baker & Milligan, 2008; Baxter, Cooklin, & Smith, 2009; Berger, Hill, & Waldfogel, 2005; Khanam, Nghiem, & Connelly, 2009). However, Baker and Milligan’s (2008) analysis of Canadian policy advances found that, although extended maternity leave provisions significantly increased mothers’ leave and breast-feeding duration, associations between these changes and outcomes, such as reduced incidence of asthma and allergies, were neither robust nor persistent.

Turning to maternal well-being, a number of microdata studies have investigated associations with leave duration. McGovern et al.’s (1997) examination of mothers’ vitality, mental health, and role function in the United States produced a nonlinear pattern in which these outcomes improved with leave duration greater than 12, 15, and 20 weeks, respectively. Other U.S.-based studies have indicated reductions in the incidence of depressive symptoms with increased periods of maternal leave (Chatterji & Markowitz, 2008), with Feldman, Sussman, and Zigler (2004) showing higher maternal depression among those taking fewer than 12 weeks leave. Again, Canadian results differ: Baker and Milligan (2008) found no significant impact of extended leave usage in Canada on mothers’ postpartum depression or self-rated health up to 2 years after the birth.

Parenting and family functioning have also, although less frequently, been investigated for associations with parental leave. Clark, Hyde, Essex, and Klein (1997) found that shorter leave periods (6 weeks or less) in the United States were associated with greater likelihood of negative interactions with infants at 4 months of age, as observed during feeding. In Canada, however, Baker and Milligan (2010) found no significant associations between increased leave duration and measures of positive and effective/hostile parenting and family functioning.

Overall, research to date emphasizes the complexity of the relationships under investigation and highlights the need for deeper understanding of the pathways through which leave taking might affect children’s outcomes. While mothers’ mental health has received some attention, less has been paid to parenting quality and couple relationships in dual-parent families, or the impact of fathers’ leave (Kamerman, 2006; O’Brien, 2006). Microdata studies allow the most meaningful investigations, with the observed cross-national differences in findings from such analyses highlighting the importance of national policy contexts. As Lero (2003) has observed, a high proportion of microdata analyses to date are from the United States, where parental leave arrangements are atypical, leaving the impact of longer and paid forms of leave underassessed. Contrasting findings from the Canadian studies reported above underline this issue, suggesting that extensions beyond a relatively generous entitlement (25 weeks in Canada prior to the changes under investigation) may have limited measurable impact on the outcomes of interest. In the following sections, we outline our development of an Australian microdata study designed to address some of these gaps in understandings of leave impact.

Conceptual Model

Our analysis seeks a deeper understanding of the impact of parental leave by focusing on potential pathways connecting leave taking and child outcomes. Specifically, we examine three aspects of maternal and family functioning: mothers’ mental health, parenting quality, and couple relationships.

These variables have all been shown to have an impact on child well-being. Poor mental health in parents is associated with increased rates of depression, anxiety, and conduct problems in children (Bayer, Sanson, & Hemphill, 2006; Gelfand & Teti, 1990). The lack of a warm, positive relationship with parents increases the risk that children will develop major behavioral and emotional problems and hinders the development of prosocial skills (Bayer et al., 2006; Paterson & Sanson, 1999). Conversely, parenting that is angry and irritable, or marked by elevated levels of criticism, negativity, and emotional reactivity has been linked to the incidence of conduct-related behavior problems in children (Chang, Schwartz, Dodge, & McBride-Chang, 2003; McMahon, 1994). In addition, the degree of confidence parents have in their own parenting skills is highly correlated with both quality of parenting and outcomes, such as parental psychosocial well-being, family conflict, and children’s well-being (Jones & Prinz, 2005). Finally, children have been shown to be particularly vulnerable to family conflict. Children exposed to high levels of inter-parental conflict and whose parents are unhappy in their couple relationship have elevated rates of socioemotional and behavioral problems (Erel & Burman, 1995).

The focus of our analysis is the impact of leave taking on these maternal and family functioning outcomes. Our conceptual model accepts fundamental assumptions implicit in policy approaches and health literature that “women need time to recover from pregnancy and childbirth and adjust to changes in marital relationships and role identities” (McGovern et al., 1997, p. 508). We thus propose relationships between leave taking and maternal and family outcomes as depicted in Figure 1, with our key hypotheses on the left side of the diagram.

Figure 1.

Use of leave and maternal and family functioning: A conceptual model.

Of central concern is the impact of mothers’ leave duration on maternal mental health: We see this as the most direct potential impact of leave taking by providing mothers with time to recover from the birth and adapt to their new situation. In line with the U.S. studies reviewed earlier, we hypothesize a positive relationship, at least with leave periods of around 12 to 15 weeks. The other two key relationships hypothesized are the impact of mothers’ leave duration on quality of parenting and couple relationships. While both these outcomes should benefit from mothers’ access to a period of leave to establish parenting routines and allow couples time to adapt, these are likely to be more complex relationships than the one proposed between mothers’ leave and mental health and more difficult to demonstrate.

Although leave duration is the key influence under examination, the literature also emphasizes the importance of paid leave and job protection (e.g., Ruhm, 2000; Tanaka, 2005). We therefore distinguish between the effects of paid and total leave duration and include a set of “leave support” variables as indicators of the degree of job protection for leave takers. Even with a formal leave entitlement to job protection, the reality of mothers’ experience depends on the security of job contracts and levels of support at the workplace for those taking parental leave. Fathers’ leave is also shown to be an important support for mothers in some studies (e.g., Chatterji & Markowitz, 2008), and we include it initially for this reason.

In addition, as Figure 1 shows, we include factors that may confound the relationships under investigation: These include characteristics of the child, mother, family, and community environment. Of particular concern is the need to control for child temperament and household income, which potentially affect both length of leave and our outcomes of interest. Having a child with a difficult temperament may either increase or decrease mothers’ leave duration, and may detrimentally affect parents’ mental health, parenting behaviors, and couple relationships. Mothers with low income may also be more likely to return to work sooner for financial reasons, and are more likely to experience psychological distress and relationship dissatisfaction. Further details on these and other potential confounders are given in the Data and Method section.

A particular contribution of our analysis is the application of longitudinal Australian data, which have hitherto been unavailable for investigation of these issues. While Australian findings, such as those from the other countries reported earlier, must be interpreted in the context of the prevailing policy framework, the varied patterns of leave taking in Australia at the time of data collection for this study provide considerable potential to extend understanding of leave duration impacts. More broadly, our analysis will add to the range of national studies available for comparison.

The Australian Context

A distinctive feature of the Australian context is the absence of a national paid parental leave scheme until January 2011. At the time of data collection for this study (2003-2006), national legislation provided only unpaid parental leave. Eligible parents (those with the same employer for 12 months prior to the child’s birth) were entitled to 52 weeks unpaid, although job protected, leave which could be shared between the mother and her partner. Although paid maternity leave was not universally available, some mothers had access, either under dedicated provisions for public service employees (for periods up to 14 weeks) or through company policies or industrial agreements (usually providing 6 or 12 weeks). Similarly, some fathers had access to paid paternity leave, typically for 1 to 2 weeks. Payments were normally at salary replacement level. For a summary of Australian parental leave arrangements, including recent and pending changes, see Alexander, Whitehouse, and Brennan (2010).

In this context, mothers frequently combined different forms of leave following the birth of a child, using paid (where available) and unpaid maternity leave, recreation leave, and sometimes other leave entitlements. Most fathers took only paid (usually recreation) leave, for an average duration of 2 weeks (Whitehouse, Baird, Diamond, & Hosking, 2006). Another feature of the Australian context is the comparatively low employment rate among women of childbearing age. In 2005, at the time our survey data were collected, Australia ranked 25th among Organisation for Economic Co-operation and Development (OECD) countries in labor force participation of women aged 25 to 34 and 35 to 44 years (OECD, 2012).

These issues have implications for our study design. First, given our intention to focus on leave duration, we are interested only in mothers who are in the labor force, use leave around the time of the birth of their child, and subsequently return to work. In Australia, a considerable proportion of mothers do not fall into these categories. Second, given the frequent combination of different forms of leave, our analysis includes all leave taken in a continuous bloc rather than only that formally designated as “parental.” Although access to an appropriate period of paid parental leave may be more beneficial for well-being than having to combine different forms of leave, this is a separate question not examined here.

Data and Method

Sample

Our analysis uses data from the infant cohort of the Longitudinal Study of Australian Children (LSAC), a national longitudinal study of Australian children and their families (for full sample details, see Misson & Sipthorp, 2007; Soloff, Lawrence, & Johnstone, 2005). Data are from Wave 1 (collected in 2004; 5,107 infants aged 3-19 months) and Wave 2 (2006; 4,606 2- to 3-year-olds; 90.2% retention), and the Parental Leave in Australia Survey conducted midway between these waves (Wave 1.5; 2005; 3,573 1- to 2-year-olds; Whitehouse et al., 2006). Waves 1 and 2 data were collected by face-to-face interviews and self-complete questionnaires with the child’s primary carer. Wave 1.5 was a mail-back questionnaire. The infant cohort was broadly representative of the Australian population at Wave 1 on most demographic characteristics (Soloff, Lawrence, Misson, & Johnstone, 2006).

For the current analyses, the infant cohort sample was restricted to primary carers who were their child’s mother and were eligible for parental leave, that is, had been with the same employer for 12 months prior to the birth of their child. To enable accurate estimation of leave duration, the sample was further restricted to those mothers who had returned to work by Wave 1.5. Thirty-nine percent of Wave 1.5 participants (n = 1,393) met these employment-related inclusion criteria. Applying these selection criteria helps to reduce the risk of selection bias because of unobserved differences between women who take maternity leave and those who do not. By narrowing our sample to those women who access maternity leave and return to work, we can focus on a more homogeneous group, albeit noting that our findings can only be generalized to this subset. In addition, we argue that focusing on these labor force–engaged women limits the extent to which preexisting mental conditions might affect our results.

Given our focus on relationships and parenting, mothers were included only if they were in a couple relationship. Although the experiences of single mothers are of interest for some of our outcomes, we wished to avoid the problems of interpretation that would accompany the use of different samples. Furthermore, the number of single mothers in our sample, given other exclusions, was too small for meaningful analysis.

To further minimize confounders of the relationship between parental leave and outcomes, we restricted the sample to mothers of children who were singleton births, born in Australia at 32 weeks or later gestation, and who did not have special health care needs (Bethell et al., 2002) at Wave 1 or Wave 2. These exclusions reduced the sample to 987 mothers. Of mothers who met all inclusion criteria, 807 had complete data on all variables.

Measures

Mothers’ mental health (Wave 2)

Psychological distress was assessed using the K6 measure of the frequency of nonspecific symptoms of psychological distress over the previous 4 weeks (Furukawa, Kessler, Slade, & Andrews, 2003): a well-validated measure, widely used in population studies. In this, as in other Australian studies (e.g., Cooklin, Canterford, Strazdins, & Nicholson, 2011; Strazdins, Shipley, & Broom, 2007), a threshold of ≥8 was used to identify those with significant psychological distress. Time pressure was measured with a single item assessing the frequency of feeling rushed or pressed for time (Prior, Sanson, Smart, & Oberklaid, 2000), rated on a 5-point scale (1 = never to 5 = always). Scores were dichotomized, with women reporting often or always feeling rushed (≥4) classified as rushed.

Parenting (Wave 2)

Distant parenting was measured using six items (5-point scale, 1 = never/almost never to 5 = always/almost always) assessing how often mothers expressed affection for their child, had warm and intimate times with the child, and felt close to the child (Paterson & Sanson, 1999). Scores were dichotomized in line with other recent studies using this measure (e.g., Baxter, Weston, & Qu, 2011; Lucas, Nicholson, & Maguire, 2011). Women whose score fell into the lowest 25% were classified as distant in their parenting. Hostile parenting was measured with five items (10-point scale, 1 = not at all to 10 = all of the time) assessing the frequency of hostile behaviors and feelings toward the child (Statistics Canada, 2000). Scores were the mean of individual items with higher scores indicating more hostile parenting. Parental self-efficacy was measured using four items (5-point scale, 1 = never/almost never to 5 = always/almost always), assessing how often mothers felt competent in parenting tasks (National Center for Education Statistics, 2004). Final scores were the mean of individual items with higher scores indicating greater parenting self-efficacy.

Couple relationship (Wave 2)

Happiness was based on the single item: “Which best describes the degree of happiness, all things considered, in your relationship?” (Sharpley & Cross, 1982). Responses were on a 7-point scale ranging from 1 = extremely unhappy to 7 = perfectly happy. Scores were dichotomized, with the 20% of women with the lowest scores (≤4) classified as unhappy with their relationship. Couple argumentativeness was assessed using four items (5-point scale, 1 = never to 5 = always) from the Quality of Coparental Interaction Scale (Ahrons, 1981), assessing the frequency of arguments and conflict with their partner. Final scores were the mean of individual items with higher scores indicating more frequent conflict.

Parental leave (Wave 1.5)

Mothers’ paid leave was the sum of paid maternity leave, paid recreation leave, paid long-service leave, paid sick leave, and other paid leave. As the distribution of paid leave taken showed peaks at 0, 6, and 12 weeks, duration was categorized into (a) no paid leave, (b) ≤6 weeks, (c) 7 to 13 weeks inclusive, and (d) >13 weeks, which corresponded to the quartile value cutoff points. Mothers’ total leave was the sum of all forms of paid and unpaid leave taken by mothers, categorized according the sample’s distribution, into (a) ≤26 weeks, (b) 27 to 52 weeks inclusive, and (c) >52 weeks. Fathers’ leave was the sum of any leave (paid or unpaid) that the father took around the time the child was born. It was measured in weeks and categorized as none versus any. It was excluded from the final models as no associations with outcomes were evident.

Workplace conditions

Workplace problems during pregnancy included items that indicated lack of support at the workplace level. Mothers were asked whether they encountered any of the following during pregnancy: (a) treated with less respect, (b) uncomfortable about taking full leave entitlement, (c) trouble negotiating return-to-work hours, (d) trouble getting information on maternity leave, (e) other problems. The measure was dichotomized to distinguish women who encountered none versus one or more of these problems. Job security was assessed as permanent employment (secure) versus casual or fixed-term contracts (insecure). Mothers’ workplace supports during pregnancy, mothers’ workplace size and sector, and work hours (part-time vs. full-time) were also examined, but excluded from the final models because of lack of associations with the outcomes.

Potential confounders

Child characteristics were gender, age (at Wave 2), and temperament (at Wave 1) assessed using the irritability subscale of the Short Infant Temperament Scale (Smart & Sanson, 2005). Infants with scores above the 70th percentile were considered to have a difficult temperament. Mothers’ characteristics (Wave 1) were age at time of birth centered on the mean age of the sample, use of a language other than English at home and marital status (married/cohabiting). Family characteristics (Wave 1) included whether the birth in question was a first child and equivalized household income from all sources, a measure commonly used in OECD publications on income equality (e.g., Burniaux, Dang, Fore, Mira d’Ercole, & Oxley, 1998). Household income was measured in Australian dollars and categorized into ≤$30,000 per annum, >$30,000 to ≤$50,000, and >$50,000, with sensitivity tests based on quintiles showing no difference in results. (Data for this item were collected in 2004 when the Australian dollar ranged in value between $US0.70 and $US0.80. However exchange rates have varied considerably over time and the Australian dollar has mostly been at, or above, parity with the U.S. dollar during 2011 and 2012.). Community characteristics (Wave 1) were represented in mothers’ rating of five items ranging from 1 to 4, assessing neighborhood livability (Stone & Hughes, 2002). Final scores were the mean of the items, with higher scores indicating less livable neighborhoods.

Analysis

After excluding mothers who did not meet the inclusion criteria and those with missing data, summary statistics were estimated for the outcome, leave, and potentially confounding variables. The relationship between duration of leave around the birth of the child and the outcome measures was investigated using generalized linear models: logistic models for the binary outcomes and linear models for the continuous measures. Separate models were used to investigate the effect of paid leave and total leave. Variables were entered into the models in three stages: (a) maternal leave variables, (b) measures of leave support, and (c) potential confounders. Data were analyzed in Stata 11.0. As the analyses focused on a subsample of the LSAC cohort selected for specific characteristics, study weights accounting for the larger cohort’s clustered sample design and sample attrition were not applied.

Results

Fifty percent of the infants in the analysis sample were female, and 52% were the mother’s first child. At Wave 2, the children were aged between 28 and 45 months (M =35 months, SD = 3 months). At the birth, mothers’ ages ranged from 19 to 45 years (M = 32 years, SD = 4 years), 9% spoke a language other than English at home, and 11% were unmarried. Below one third had an equivalized household income ≤$30,000 per annum (28%) or >$50,000 (31%), with the remaining 41% in the >$30,000 to ≤$50,000 category. Neighborhood livability ratings ranged from 1 to 3.6 (M = 2.0, SD = 0.4).

Duration of paid leave used by the mothers ranged from 0 to 58 weeks. Around 25% of mothers took each of: no paid leave, ≤6 weeks, 7 to 13 weeks inclusive, and more than 13 weeks. Total leave taken ranged from 1 to 163 weeks. Thirty-four percent of mothers took 6 months or less, 48% took between 26 weeks and 1 year, and 18% took more than 1 year. Twenty-one percent of mothers reported one or more problems at work during their pregnancy, and 13% were not permanent employees.

Maternal Mental Health

Longer paid leave duration was associated with lower maternal psychological distress 2 to 3 years after the infant’s birth after controlling for workplace environment and child, maternal, family, and neighborhood characteristics. As Table 1 shows, mothers who took >13 weeks paid leave had around one quarter the odds (odds ratio [OR] = 0.24) of being distressed as those who took no paid leave. In this model, problems at work and nonpermanent employment status were unrelated to maternal distress. However, mothers showed higher odds of being distressed if they described their child as having a difficult temperament or if they reported using a language other than English at home (OR = 3.17 and 3.21, respectively).

Table 1.

Mothers’ Paid Leave and Other Factors on Maternal Mental Health.

	Psychological Distress		Time Pressured
	OR [95% CI]	p	OR [95% CI]	p
Paid leave^a
>0 weeks and ≤6 weeks	0.55 [0.20, 1.50]	.24	1.36 [0.84, 2.21]	.22
>6 weeks and ≤13 weeks	0.57 [0.21, 0.26]	.26	0.92 [0.56, 1.50]	.74
>13 weeks	0.24 [0.07, 0.84]	.03	1.56 [0.95, 2.57]	.08
Workplace conditions^b
Workplace problems	1.80 [0.80, 4.07]	.16	3.05 [1.99, 4.66]	<.001
Nonpermanent employee	0.57 [0.17, 1.94]	.37	2.14 [1.22, 3.77]	.008
Child characteristics
Female	1.98 [0.94, 4.21]	.07	1.17 [0.86, 1.61]	.32
Age at Wave 2 (months)^c	0.97 [0.85, 1.11]	.64	0.98 [0.93, 1.04]	.60
Difficult temperament	3.12 [1.49, 6.53]	.003	1.45 [1.03, 2.04]	.04
Mother characteristics
Age at birth of child (years)^d	1.00 [0.91, 1.10]	.95	1.11 [1.06, 1.16]	<.001
Language other than English at home	3.20 [1.34, 7.65]	.009	1.08 [0.61, 1.90]	.79
Unmarried	1.10 [0.31, 3.85]	.88	0.79 [0.48, 1.32]	.37
Family characteristics
First child	0.55 [0.24, 1.24]	.15	0.64 [0.45, 0.91]	.01
Joint household income^e
≤$30,000 per annum	0.85 [0.36, 1.99]	.71	0.96 [0.65, 1.42]	.85
>$50,000 per annum	0.86 [0.33, 2.25]	.76	0.87 [0.59, 1.30]	.50
Community characteristics
Neighborhood livability^f	1.39 [0.58, 3.30]	.46	1.15 [0.81, 1.63]	.44

Note: OR = odds ratio; 95% CI = 95% confidence interval.

Baseline category: No paid leave.

During pregnancy.

Centered on mean age of 35 months.

Centered on mean age of 32 years.

Figures are in Australian dollars. Baseline category: >$30,000 to ≤$50,000.

Centered on mean of 2, with higher scores indicating less livable neighborhoods.

Mothers who took a total of 27 to 52 weeks’ leave also had slightly lower odds of distress than those who took a total leave period of 26 weeks or less (OR = 0.43, 95% confidence interval [CI] = 0.18-0.99). However, the odds of distress among mothers who took >52 weeks total leave was not significantly different from those taking 26 weeks or less (OR = 0.68, 95% CI = 0.25-1.82). The total leave model produced similar results to those reported for the paid leave model for the work characteristics and demographic variables. (For space reasons, only the paid leave models are included in the article. Differences between paid and total leave models are noted in the text. Full results for models using total leave are available from the authors.)

After controlling for workplace environment factors and other potential confounders, there was no association between paid leave duration and our measure of time pressure 2 to 3 years later. However, mothers had three times the odds (OR = 3.05) of feeling rushed if they experienced at least one problem in their workplace during pregnancy than if they experienced no problems, and mothers who were nonpermanent employees also had significantly higher odds (OR = 2.14) of feeling rushed than those in permanent jobs. In addition, mothers showed higher odds of feeling rushed if they described their child as having a difficult temperament and if they were older (OR = 1.45 and 1.11, respectively), but less likely to feel rushed if this was their first child (OR = 0.64). No association between total leave duration and time pressure was found. The effects for the other covariates in the total leave model were similar to those reported for the paid leave model.

Quality of Parenting

After controlling for workplace environment and other potential confounders, there was no association between paid leave duration and any of the parenting quality outcomes measured 2 to 3 years after the birth (Table 2). Nor were problems at the workplace and nonpermanent status associated with later parenting quality. The risk of maternal distance was increased if the child was described as having a difficult temperament and for older children, and decreased for first-time mothers (OR = 1.62, 1.07, and 0.60, respectively). Maternal hostility was found to be increased for mothers who reported that their child had a difficult temperament and reduced for older mothers. Joint household income was also found to be related to maternal hostility, with mothers found to be less hostile if their joint income was >$50,000 per annum compared with >$30,000 to ≤$50,000. No difference was found between those whose income was ≤$30,000 compared with the baseline income category. Only child’s temperament was found to be significantly related to self-efficacy, with mothers scoring lower on this scale if they described their child as having a difficult temperament.

Table 2.

Mothers’ Paid Leave and Other Factors on Quality of Parenting.

	Distant Parenting		Hostile Parenting		Parental Self-Efficacy
	OR	p	β [95% CI]	p	β [95% CI]	p
Paid leave^a
>0 weeks and ≤6 weeks	1.23 [0.71, 2.12]	.47	0.09 [−0.19, 0.36]	.54	−0.06 [−0.19, 0.06]	.33
>6 weeks and ≤ 13 weeks	1.13 [0.65, 1.96]	.66	0.00 [−0.28, 0.28]	1.00	−0.08 [−0.21, 0.05]	.22
>13 weeks	1.37 [0.80, 2.37]	.25	0.12 [−0.16, 0.40]	.39	−0.05 [−0.18, 0.07]	.41
Workplace conditions^b
Workplace problems	0.70 [0.45, 1.09]	.12	0.12 [−0.10, 0.34]	.29	−0.04 [−0.14, 0.06]	.43
Nonpermanent employee	1.25 [0.69, 2.26]	.47	−0.10 [−0.41, 0.20]	.51	0.02 [−0.12, 0.16]	.75
Child characteristics
Female	0.99 [0.70, 1.40]	.96	−0.09 [−0.27, 0.08]	.30	0.07 [−0.01, 0.15]	.11
Age at Wave 2 (months)^c	1.07 [1.00, 1.13]	.04	0.03 [−0.01, 0.06]	.13	0.01 [−0.01, 0.02]	.44
Difficult temperament	1.62 [1.14, 2.32]	.008	0.32 [0.13, 0.51]	<.001	−0.16 [−0.25, −0.07]	<.001
Mother characteristics
Age at birth of child (years)^d	0.99 [0.95, 1.04]	.68	−0.05 [−0.07, −0.03]	<.001	0.00 [−0.01, 0.01]	.98
Language other than English at home	0.82 [0.45, 1.50]	.52	−0.06 [−0.38, 0.25]	.71	−0.09 [−0.24, 0.05]	.20
Unmarried	0.87 [0.49, 1.54]	.62	−0.09 [−0.38, 0.19]	.52	0.07 [−0.06, 0.20]	.31
Family characteristics
First child	0.60 [0.41, 0.88]	.009	−0.10 [−0.30, 0.09]	.31	0.05 [−0.03, 0.14]	.23
Joint household income^e
≤$30,000 per annum	0.70 [0.46, 1.09]	.11	−0.03 [−0.24, 0.19]	.80	0.06 [−0.03, 0.16]	.20
>$50,000 per annum	1.20 [0.79, 1.84]	.39	−0.26 [−0.48, −0.03]	.02	0.08 [−0.02, 0.18]	.12
Community characteristics
Neighborhood livability^f	1.12 [0.77, 1.65]	.55	0.18 [−0.01, 0.38]	.07	−0.05 [−0.14, 0.04]	.26

Note: OR = odds ratio; 95% CI = 95% confidence interval.

Baseline category: No paid leave.

During pregnancy.

Centered on mean age of 35 months.

Centered on mean age of 32 years.

Figures are in Australian dollars. Baseline category: >$30,000 to ≤$50,000

Centered on mean of 2, with higher scores indicating less livable neighborhoods.

Total leave was related to distant parenting, although not to hostile parenting or parental self-efficacy. Contrary to expectations, longer total leave was associated with higher odds of feeling distant. Compared with those who took 26 weeks or less, mothers who took between 27 and 52 weeks inclusive or more than 52 weeks, had higher odds of distant parenting (OR = 1.76, 95% CI = 1.16-2.66; OR = 1.78, 95% CI = 1.07-2.97, respectively). The effects for the other covariates in the total leave model were similar to those for the paid leave model.

Couple Relationship

After controlling for workplace environment and child, maternal, family and neighborhood covariates, no association was found between paid leave taking and couple relationship 2 to 3 years later. However, mothers who experienced workplace problems during their pregnancy had higher odds of unhappiness (OR = 1.64) and more couple argumentativeness (β = 0.19) than mothers who did not experience such problems (Table 3). Several other factors were associated with couple outcomes. Odds for relationship unhappiness were higher among mothers who were older, spoke a language other than English at home, were cohabiting, and lived in less livable neighborhoods. Couple conflict was more likely for women whose child had a difficult temperament, for those who were older or cohabiting, and less likely for first-time mothers. Total leave was found to be unrelated to either outcome and similar results were found for the other covariates in the model.

Table 3.

Mothers’ Paid Leave and Other Factors on Couple Relationship.

	Unhappiness		Argumentativeness
	OR (95% CI)	p	β (95% CI)	p
Paid leave^a
>0 weeks and ≤6 weeks	1.72 [0.93, 3.18]	.08	0.01 [−0.11, 0.12]	.90
>6 weeks and ≤13 weeks	1.15 [0.61, 2.17]	.67	−0.06 [−0.17, 0.05]	.30
>13 weeks	1.16 [0.61, 2.19]	.65	−0.04 [−0.15, 0.07]	.49
Workplace conditions^b
Workplace problems	1.64 [1.04, 2.58]	.03	0.19 [0.10, 0.28]	<.001
Nonpermanent employee	1.21 [0.61, 2.41]	.58	−0.10 [−0.22, 0.03]	.13
Child characteristics
Female	1.25 [0.84, 1.85]	.27	−0.01 [−0.08, 0.06]	.76
Age at wave 2 (months)^c	0.97 [0.90, 1.04]	.39	−0.01 [−0.02, 0.00]	.12
Difficult temperament	1.31 [0.87, 1.98]	.19	0.09 [0.01, 0.17]	.03
Mother characteristics
Age at birth of child (years)^d	1.09 [1.04, 1.15]	.001	0.01 [0.00, 0.02]	.01
Language other than English at home	2.13 [1.19, 3.83]	.01	0.11 [−0.02, 0.23]	.10
Unmarried	1.84 [1.02, 3.31]	.04	0.14 [0.02, 0.27]	.02
Family characteristics
First child	0.82 [0.53, 1.26]	.36	−0.08 [−0.17, 0.00]	.04
Joint household income^e
≤$30,000 per annum	0.88 [0.54, 1.42]	.60	−0.01 [−0.10, 0.08]	.90
>$50,000 per annum	0.99 [0.61, 1.62]	.97	0.04 [−0.06, 0.13]	.45
Community characteristics
Neighborhood livability^f	1.92 [1.22, 3.01]	.004	0.08 [0.00, 0.16]	.05

Note: OR = odds ratio; 95% CI = 95% confidence interval.

Baseline category: No paid leave.

During pregnancy.

Centered on mean age of 35 months.

Centered on mean age of 32 years.

Figures are in Australian dollars. Baseline category: >$30,000 to ≤$50,000.

Centered on mean of 2, with higher scores indicating less livable neighborhoods.

Discussion

Our analyses examined new longitudinal data from a national cohort to investigate the impact of mothers’ paid and total leave duration and leave supports on maternal and family outcomes 2 to 3 years after the birth. Two of our outcomes (parenting and couple relationships) have previously received limited analytical attention, and our inclusion of workplace supports also extends previous research. We examined a clearly defined group (two-parent families), excluding those for whom special circumstances were likely to confound results. Our focus was on mothers who took leave and returned to work and our results are generalizable to this group.

In combination, the analytical models suggest a complex picture, with mothers’ leave duration and workplace conditions associated with some aspects of maternal and family functioning, and some differences emerging between paid and total leave duration. While having a child with a difficult temperament showed the strongest and most consistent associations across the outcomes measures, several associations with mothers’ leave duration and working conditions remained after adjusting for this and other contextual influences.

The significant benefit of paid leave duration for lessening the odds of psychological distress was evident only for mothers who took >13 weeks paid leave compared with those who had no paid leave, suggesting that Australian patterns of paid leave usage (clustered around 6 and 12 weeks at the time of these surveys) fell below the optimum at the time. Total leave duration of >26 to ≤52 weeks was also associated with lower odds of maternal distress (compared with those who took <26 weeks), but this was not the case for leave periods >52 weeks. This may reflect financial security pressures associated with longer leave duration, as this would almost certainly have been largely unpaid leave.

Although neither paid nor total leave duration was associated with the second measure of maternal mental health (feeling rushed), there was a significant and strong positive association between this outcome and the extent to which mothers had encountered problems at their workplace and lacked permanent status during pregnancy (both stronger associations than with child temperament). These results (and those obtained in the models examining couple relationship) underline the need to extend analyses beyond leave duration to the degree of support and security around leave taking as additional factors relevant for maternal and family functioning.

The associations between longer total leave durations (>26 to ≤52 weeks and >52 weeks) and one aspect of poorer parenting (distant parenting) further underlines the complex determinants of family functioning. No relationship was evident with paid leave, and it may be the case that it is only when leave is composed of a combination of forms and largely unpaid (as would be the case for most mothers taking >26 weeks leave in Australia) that longer duration has the effect illustrated in Table 2. Other possible explanations are that mothers who have taken shorter leave periods may engage in more intensive parenting in the time they do have with their children, or that those who have taken longer periods are subsequently more sensitized to being apart from their children.

Although mothers’ leave duration was not associated with couple relationship outcomes, the presence of workplace problems was associated with both relationship unhappiness and argumentativeness, again suggesting that problems at the workplace have far-reaching implications. Although further research is needed to elucidate interactions between such variables and the maternal and family functioning outcomes examined in this article, our analysis is an important first step in the Australian context.

Conclusion

Our analysis sought a more nuanced understanding of the impact of parental leave by focusing on maternal and family functioning variables that potentially constitute a pathway between mothers’ use of leave and child outcomes. We used longitudinal microdata that for the first time in Australia provided detailed individual characteristics, including leave duration and workplace supports, to test our hypothesized associations rigorously.

Key findings of our analysis were associations between mothers’ leave duration (paid and total) and reduced maternal distress levels. Our findings on this measure suggest an optimal leave duration for mothers of more than 6 months (although not necessarily more than 12 months), with more than 3 months being paid leave. The implementation from January 2011 of a national government scheme providing 18 weeks paid parental leave may shift patterns of leave taking closer to this suggested optimum, with future research required to assess changes in leave duration and implications for maternal well-being.

Ultimately, elaboration of pathways from leave usage through intermediate variables to children’s outcomes will require detail on a wider range of potential influences on maternal and family functioning as well as precise information on what parents actually do while on leave (Kamerman, 2006; Lero, 2003). Nevertheless, our analysis adds to the evidence base for policy makers, affirming the value of mothers’ leave for maternal mental well-being and highlighting the importance of security and support in the workplace for maternal and family outcomes.

Footnotes

Acknowledgements

This article uses unit record data from Growing Up in Australia, the Longitudinal Study of Australian Children (LSAC). LSAC is conducted in partnership between the Department of Families, Housing, Community Services and Indigenous Affairs (FaHCSIA), the Australian Institute of Family Studies (AIFS), and the Australian Bureau of Statistics (ABS).

Authors’ Note

The findings and views reported are those of the authors and should not be attributed to FaHCSIA, AIFS, the ABS, or any funding bodies.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research on which this article is based.

The Parental Leave in Australia Survey conducted in conjunction with LSAC was designed by Gillian Whitehouse and Chris Diamond (University of Queensland) and Marian Baird (University of Sydney) and funded through the Australian Research Council (LP0453613) with support from the Australian Human Rights Commission, New South Wales (NSW) Office of Industrial Relations, Queensland Department of Industrial Relations, NSW Office for Women, and the Women’s Electoral Lobby. Research for this article was also supported by the National Health and Medical Research Council (Career Development Award 390136 to Jan Nicholson) and the Australian Research Council (DP0774439).

References

Ahrons

C. R.

(1981). The continuing coparental relationship between divorced spouses. American Journal of Orthopsychiatry, 51, 415-429.

Alexander

Whitehouse

Brennan

(2010). Australia. In Moss

(Ed.), International review of leave policies and related research 2010 (pp. 42-51: Employment Relations Research Series, No 115). London, England: Department for Business Innovation and Skills.

Baker

Milligan

(2008). Maternal employment, breastfeeding and health: Evidence from maternity leave mandates. Journal of Health Economics, 27, 871-887.

Baker

Milligan

(2010). Evidence from maternity leave expansions of the impact of maternal care on early child development. Journal of Human Resources, 45, 1-32.

Baxter

Cooklin

Smith

(2009). Which mothers wean their babies prematurely from full breastfeeding? An Australian cohort study. Acta Paediatrica, 98, 1274-1277.

Baxter

Weston

(2011). Family structure, co-parental relationship quality, post-separation paternal involvement and children’s emotional wellbeing. Journal of Family Studies, 17, 86-109.

Bayer

J. K.

Sanson

A. V.

Hemphill

S. A.

(2006). Parent influences on early childhood internalizing difficulties. Journal of Applied Developmental Psychology, 27, 542-559.

Berger

Hill

Waldfogel

(2005). Maternity leave, early maternal employment and child outcomes in the US. The Economic Journal, 115(February), F29-F47.

Bethell

Read

Stein

Blumberg

Wells

Newacheck

(2002). Identifying children with special health care needs: Development and evaluation of a short screening instrument. Ambulatory Paediatrics, 2, 38-48.

10.

Brooks-Gunn

Han

Waldfogel

(2002). Maternal employment and child outcomes in the first three years of life: The NICHD Study of Early Child Care. Child Development, 73, 1052-1072.

11.

Brooks-Gunn

Han

Waldfogel

(2010). First year maternal employment and child development in the first seven years. Monographs of the Society for Research in Child Development, 75, 1-142.

12.

Burniaux

J. M.

Dang

T. T.

Fore

M.F.

Mira d’Ercole

Oxley

(1998). Income distribution and poverty in selected OECD countries (OECD Economics Working Paper No. 189). Paris, France: OECD.

13.

Chang

Schwartz

Dodge

K. A.

McBride-Chang

(2003). Harsh parenting in relation to child emotion regulation and aggression. Journal of Family Psychology, 17, 598-606.

14.

Chatterji

Markowitz

(2008). Family leave after childbirth and the health of new mothers (National Bureau of Economic Research [NBER] Working Paper No. 14156). Cambridge, MA: NBER.

15.

Clark

Hyde

Essex

Klein

(1997). Length of maternity leave and quality of mother-infant interactions. Child Development, 68, 364-383.

16.

Cooklin

A. R.

Canterford

Strazdins

Nicholson

J. M.

(2011). Employment conditions and maternal postpartum mental health: Results from the Longitudinal Study of Australian Children. Archives of Women’s Mental Health, 14, 217-225. doi:10.1007/s00737-010-0196

17.

Erel

Burman

(1995). Interrelatedness of marital relations and parent-child relations: A meta-analytic review. Psychological Bulletin, 118, 108-132.

18.

Feldman

Sussman

Zigler

(2004). Parental leave and work adaptation at the transition to parenthood: Individual, marital and social correlates. Journal of Applied Developmental Psychology, 25, 459-479.

19.

Furukawa

Kessler

R. C.

Slade

Andrews

(2003). The performance of the K6 and K10 screening scales for psychological distress in the Australian National Survey of Mental Health and Well-Being. Psychological Medicine, 33, 357-362.

20.

Galtry

(2003). The impact on breastfeeding of labour market policy and practice in Ireland, Sweden and the USA. Social Science & Medicine, 57, 167-177.

21.

Gelfand

D. M.

Teti

D. M.

(1990). The effects of maternal depression on children. Clinical Psychology Review, 10, 329-353.

22.

Gregg

Washbrook

Propper

Burgess

(2005). The effects of a mother’s return to work decision on child development in the UK. The Economic Journal, 115(February), F48-F80.

23.

Han

Waldfogel

Brooks-Gunn

(2001). The effects of early maternal employment on later cognitive and behavioural outcomes. Journal of Marriage and Family, 63, 336-354.

24.

Jones

T. L.

Prinz

R. J.

(2005). Potential roles of parental self-efficacy in parent and child adjustment: A review. Clinical Psychology Review, 25, 341-363.

25.

Joshi

Cooksey

Verropoulou

(2009). Combining childrearing with work: Do maternal employment experiences compromise child development? (Centre for Longitudinal Studies [CLS] Working Paper 2009/1). London, England: CLS.

26.

Kamerman

(2000). From maternity to parental leave policies: Women’s health, employment and child and family well-being. Journal of the American Medical Women’s Association, 55, 96-99.

27.

Kamerman

(2006). Parental leave policies: The impact on child well-being. In Moss

O’Brien

(Eds.), International review of leave policies and related research 2006 (pp. 16-21; Employment Relations Research Series No. 57). London, England: Department of Trade and Industry.

28.

Khanam

Nghiem

H. S.

Connelly

L. B.

(2009). Does maternity leave effect child health? Evidence from parental leave in Australia survey. Evidence from parental leave in Australian Survey. In 31st Australian conference for Health Economists, Hobart, Australia. (University of Southern Queensland e-Prints). Retrieved from http://eprints.usq.edu.au/5822/

29.

Lero

(2003). Research on parental leave policies and children’s development: Implications for policy makers and service providers. In Tremblay

Barr

Peters

(Eds.), Encyclopedia on early childhood development (pp. 1-9). Montreal, Quebec, Canada: Centre of Excellence for Early Childhood Development.

30.

Lucas

Nicholson

Maguire

(2011). Longitudinal Study of Australian Children Annual Statistical Report 2010: 5 Parenting practices and behaviours. Melbourne, Victoria: Australian Institute of Family Studies. Retrieved from http://www.aifs.gov.au/growingup/pubs/asr/2010/asr2010e.html

31.

Lucas-Thompson

R. G.

Goldberg

W. A.

Prause

(2010) Maternal work early in the lives of children and its distal associations with achievement and behavior problems: A meta-analysis. Psychological Bulletin, 136, 915-942.

32.

McGovern

Dowd

Gjerdingen

Moscovice

Kochevar

Lohman

(1997). Time off work and the postpartum health of employed women. Medical Care, 35, 507-521.

33.

McMahon

R. J.

(1994). Diagnosis, assessment, and treatment of externalizing problems in children: The role of longitudinal data. Journal of Consulting and Clinical Psychology, 62, 901-917.

34.

Misson

Sipthorp

(2007). Wave 2 weighting and non-response (LSAC Technical Paper, No 5). Melbourne, Victoria: Australian Institute of Family Studies.

35.

National Center for Education Statistics. (2004). Early childhood longitudinal study, Birth cohort 9-month restricted-use data files user’s manual (NCES 2004-093). Washington, DC: Author.

36.

O’Brien

(2006). Parental leave policies for mothers and fathers: Children’s perspectives and well-being. In Moss

O’Brien

(Eds.), International review of leave policies and related research 2006 (pp. 22-30; Employment Relations Research Series No. 57). London, England: Department of Trade and Industry.

37.

Organisation for Economic Co-operation and Development. (2012). OECD.StatExtracts, labour force statistics by sex and age—indicators. Retrieved from http://stats.oecd.org/Index.aspx

38.

Paterson

Sanson

(1999). The association of behavioural adjustment to temperament, parenting and family characteristics among 5 year old children. Social Development, 8, 293-309.

39.

Prior

Sanson

Smart

Oberklaid

(2000). Pathways from infancy to adolescence: Australian temperament project 1983-2000. Melbourne, Victoria: Australian Institute of Family Studies.

40.

Ruhm

(2000). Parental leave and child health. Journal of Health Economics, 19, 931-960.

41.

Ruhm

(2004). Parental employment and child cognitive development. Journal of Human Resources, 39, 155-192.

42.

Sharpley

C. F.

Cross

D. G.

(1982). A psychometric evaluation of the Spanier Dyadic Adjustment Scale. Journal of Marriage and the Family, 44, 739-741.

43.

Smart

Sanson

(2005). A comparison of children’s temperament and adjustment across 20 years. Family Matters, 72, 97-104.

44.

Soloff

Lawrence

Johnstone

(2005). Sample design (LSAC Technical Paper, No. 1). Melbourne, Victoria: Australian Institute of Family Studies.

45.

Soloff

Lawrence

Misson

Johnstone

(2006). Wave 1 weighting and non-response (LSAC Technical Paper No 3). Melbourne, Victoria: Australian Institute of Family Studies.

46.

Statistics Canada. (2000). National Longitudinal Survey of Children and Youth (NLSCY) Cycle 3 survey instruments: Parent questionnaire. Ottawa, Ontario: Author.

47.

Stone

Hughes

(2002). Social capital: Empirical meaning and measurement validity (Research Paper No. 27). Melbourne, Victoria: Australian Institute of Family Studies.

48.

Strazdins

Shipley

Broom

(2007). What does family-friendly really mean? Well-being, time and the quality of parents’ jobs. Australian Bulletin of Labour, 33, 202-222.

49.

Tanaka

(2005). Parental leave and child health across OECD countries. The Economic Journal, 115(February), F7-F28.

50.

Waldfogel

Han

Brooks-Gunn

(2002). The effects of early maternal employment on child cognitive development. Demography, 39, 369-392.

51.

Whitehouse

Baird

Diamond

Hosking

(2006). The Parental Leave in Australia Survey: November 2006 report. Retrieved from http://www.uq.edu.au/polsis/parental-leave/level1-report.pdf

52.

Winegarden

Bracy

(1995). Demographic consequences of maternal-leave programs in industrial countries: Evidence from fixed-effects models. Southern Economic Journal, 61, 1020-1035.