Workplace flexibility is believed to be a key factor for improving labour market outcomes among mothers and further reducing the gender pay gap, but less is known about other aspects of flexibility, such as whether fathers value it or whether other household members benefit from it. This column uses a Swedish social insurance reform to show that when more workplace flexibility is available, fathers use it, and that flexibility for fathers has positive spillover effects on maternal health.
Workplace flexibility is believed to be a key factor for improving labour market outcomes among mothers and further reducing the gender pay gap, but less is known about other aspects of flexibility, such as whether fathers value it or whether other household members benefit from it. This column uses a Swedish social insurance reform to show that when more workplace flexibility is available, fathers use it, and that flexibility for fathers has positive spillover effects on maternal health.
In many countries around the world, the majority of mothers and fathers of young children work outside the home and must balance the competing demands of job and family responsibilities. Policies that promote temporal flexibility in the workplace allow parents to rearrange their work hours in case of an unforeseen family need – such as a child's sickness or a snow day that forces school closure – while minimising work interruption. In other words, a key feature of workplace flexibility is that it generates flexibility in when to stay home from work. As mothers are more likely than fathers to be ‘on call’ for unanticipated domestic needs (Weeden et al. 2016), a burgeoning literature argues that workplace flexibility is a key factor for improving labour market outcomes among mothers and further reducing the gender pay gap (Bertrand et al. 2010, Goldin 2014, Goldin and Katz 2016).
But we know much less about other important aspects of workplace flexibility. First, do fathers value workplace flexibility? Put differently, if workplace flexibility is available, do fathers use it? Second, while a few studies show that work-related stress propagates throughout the family – for instance, individuals' work hours negatively affect their relationship quality and their partners' wellbeing (Shafer et al. 2018, Fan et al. 2019) – there is scarce evidence on whether workplace flexibility policies can buffer against these spillover effects. If one household member has access to a workplace flexibility policy, then do other household members benefit? Third, relative to our understanding of the role of workplace flexibility in reducing mothers' career costs of family formation, we know less about its impacts on the other costs associated with having children.
In a recent study (Persson and Rossin-Slater 2019), we aim to answer these questions by analysing fathers' responses to increased access to workplace flexibility and the spillover effects of fathers' access to workplace flexibility on maternal wellbeing. We focus on a critical period in family life, when spillovers may be especially important: the months immediately following childbirth. In this period, for a mother the major cost of having a family is not the cost to her career – which grows in magnitude and importance over time since childbirth (e.g. Kleven et al. 2018) – but instead the physical and mental health costs associated with postpartum recovery. Thus, we examine whether workplace flexibility for new fathers generates spillover benefits through improvements in maternal postpartum health.
To study this issue, we take advantage of a Swedish social insurance reform that effectively increased workplace flexibility for new fathers by relaxing a central restriction in the parental leave system. At the time of the reform, Swedish households were granted 16 months of job-protected paid leave (per child), to be allocated across the two parents. However, parents were generally not allowed to be on leave at the same time – in fact, simultaneous leave use was permitted for only 10 days around childbirth (hereafter, we refer to these 10 days as ‘baseline leave’). Since virtually all Swedish mothers take full-time leave in the months following childbirth, this rule effectively limited fathers' ability to use paid leave alongside the mother. In fact, in a typical household, the mother would be on leave alone for the first 14 months of a child's life, while the father would only take leave once the mother returned to work. The ‘Double Days’ reform, implemented on 1 January 2012, relaxed the restriction on simultaneous leave by allowing both parents to use full-time leave benefits at the same time for up to 30 additional days during the child's first year of life. Importantly, these Double Days could be taken intermittently; thus, fathers were granted more flexibility to choose, on a day-to-day basis, whether to claim paid leave to stay home together with the mother and child.
To generate predictions about household demand for father presence at home as well as the potential impacts of father presence on maternal wellbeing, we begin with a simple theoretical framework. In this framework, households must decide on a daily basis whether or not the father should stay home together with the mother. This decision reflects a trade-off between the household's benefit from father presence and the associated costs, which stem from his lost income (leave benefits are only partially wage-replaced) and the loss of the opportunity to use a future parental leave day. Importantly, the benefit of father presence at home varies from day to day; intuitively, additional support for the mother may be particularly valuable on some days, such as when she is not feeling well, needs medical care, or suffers from stress or anxiety. The central theoretical prediction is that households choose to keep the father at home alongside the mother on days when the maternal need for support is the highest.
Next, to provide a comprehensive empirical analysis of the effects of the Double Days reform on fathers' leave use and maternal health, we link multiple sources of Swedish administrative data, including birth records, parental leave claims, as well as inpatient, specialist outpatient, and prescription drug records. We use data on parents with first births of singleton children in 2008-2012 and implement a research design that compares the outcomes of parents of children born in the three months before and after the reform, relative to the analogous difference between these birth months in the three preceding years. Our empirical strategy thus exploits the change in eligibility for simultaneous leave for parents of children born shortly after the Double Days reform, while netting out other differences in family outcomes between October-December and January-March births that also exist in non-reform years (e.g. differences in characteristics of mothers who give birth in different seasons of the year, differences in holiday time off work, and differential birth timing due to school starting-age laws).
We first show that when more workplace flexibility is available, fathers use it. The Double Days reform raises the likelihood that fathers use more than the 10 days of baseline leave (hereafter referred to as ‘post-baseline leave’) in the first 60 and 180 days after childbirth by 3.9 and 5.9 percentage points, respectively, corresponding to 50% and 24% effects relative to the sample means. Interestingly, while the effects on any post-baseline leave use are substantial, we only observe a one-to-two-day average increase in the total number of leave days taken by fathers in the first six months post-childbirth. Thus, it appears that the reform primarily affects fathers' leave use on the extensive, rather than intensive, margin.
Next, we show that workplace flexibility for fathers has positive spillover effects on maternal health. We find that the reform leads to a 1.5 percentage point (14%) reduction in the likelihood of a mother having an inpatient or specialist outpatient visit for childbirth-related complications, and a 1.9 percentage point (11%) reduction in the likelihood of her having an antibiotic prescription drug in the first six months postpartum. Further, we find an improvement in maternal postpartum mental health – we observe a marginally significant 0.3 percentage point (26%) reduction in the likelihood of any anti-anxiety prescription drug in the first six months post-childbirth. When examining the timing of these effects, we find that the reduction in anti-anxiety drugs is particularly strong (and statistically significant at conventional levels) in the first three months after childbirth. These effects on maternal health are larger in both absolute and relative terms for particularly vulnerable mothers, who have a history of mental and physical health issues pre-childbirth.
The large maternal health effect magnitudes are consistent with the theoretical prediction that fathers take leave on days when the value of doing so is especially high. Further, we show that among families in which mothers have a history of health issues pre-childbirth, the reform increases the likelihood that the father takes leave on the same day as when the mother gets medical care. Thus, increased workplace flexibility for fathers allows them to stay home and care for their infants while mothers obtain medical care, in addition to averting health complications that necessitate medical intervention in the first place.
Our results have important implications for settings with limited workplace flexibility (such as the US, the only high-income country without a national paid family leave policy). Mothers bear the majority of the cost of a lack of workplace flexibility – not only directly through greater career costs of family formation (as documented in prior literature), but also indirectly as fathers' inability to respond to unforeseen domestic needs exacerbates the maternal health cost of childbearing. Our finding of a substantial maternal health benefit stemming from a small increase in the average number of leave days taken by fathers further implies that workplace flexibility policies can be highly cost-effective because they make use of families' private information about when leave is most beneficial and give families agency in choosing when and how to use their allotted time off work.
This article first appeared on www.VoxEU.org on September 1, 2019. Reproduced with permission.
