The Impact of Paid Maternity Leave on Maternal Health
New working paper "The Impact of Paid Maternity Leave on Maternal Health" from Aline Bütikofer, Julie Riise and Meghan Skira.
The impact of maternity leave
Aline Bütikofer, Julie Riise and Meghan Skira have a new working paper "The Impact of Paid Maternity Leave on Maternal Health" where they examine the impact of the introduction of paid maternity leave in Norway in 1977 on maternal health.
Across OECD countries, there is substantial variation in maternity leave benefits. While paid maternity leave benefits were increased in most high-income countries over the last several decades, mothers are only eligible to 12 weeks of unpaid leave in the United States. Several previous papers estimate the effects of maternity leave reforms on maternal employment and earnings as well as a variety of outcomes of children. However, there is little evidence on the causal effects of maternity leave on maternal health outcomes, which is surprising given one of the main motivations for maternity leave provisions is to allow women to recover from childbirth.
The paper exploits the introduction of paid maternity leave in Norway in 1977. Before the policy reform, mothers were eligible for 12 weeks of unpaid leave. Mothers giving birth after July 1, 1977 were entitled to 4 months of paid leave and 12 months of unpaid leave. Bütikofer, Riise and Skira combine Norwegian administrative data with survey data on the health of women around age 40 and estimate the medium- and long-term impacts of the reform using regression discontinuity and difference-in-regression discontinuity designs.
Larger impact for low-resource mothers
The paper suggests that the introduction of paid maternity leave is protective of maternal health. Various aspects of metabolic health improved for mothers who were eligible for the reform. That is, access to maternity leave benefits improved mothers’ BMI, blood pressure, pain, and mental health, and it increased health-promoting behaviors, such as exercise and not smoking. The reform had significantly larger effects on mothers who experienced complications at delivery, first-time mothers, single mothers and mothers with below-median household income.
The authors find that the reform did not crowd out unpaid leave and did not significantly affect maternal income. This implies more time at home after childbirth, not income effects, drives the health improvements. Furthermore, the improvements were larger for women who would have taken little unpaid leave in the absence of the reform. Thus, the additional time at home was particularly beneficial for disadvantaged women who could not afford to take much unpaid time off work after childbirth.
Consistent with the idea that more time at home is an important channel, the authors hypothesize that increased breastfeeding duration plays a role in generating the health improvements. In particular, they find suggestive empirical evidence that the reform decreased the risk of breast and ovarian cancer. This result fits well in a large public health literature documenting a negative association between breastfeeding and breast and ovarian cancer.
Bütikofer, Riise and Skira analyze the impact of maternity leave on many aspects of health including self-reported measures as well as biomarkers from medical examinations. These biomarkers predict well a variety of future health conditions, and they allow the authors to learn more about the mechanisms through which maternity leave affects maternal health. In addition, the authors analyze the effects of maternity leave several years after the woman has given birth. For the most part, the prior literature has focused on maternal health outcomes shortly after childbirth. The results in this paper are informative for understanding the medium and long-term effects of paid maternity leave, which are important for policy-makers to consider when designing family leave schemes.