Evidence on Maternal Health from Two Large Canadian Parental Leave Expansions: When is Enough Too Much?'

Exploiting unique administrative longitudinal data sets on medical services provided to mothers before- and after- delivery, we estimate the causal effects of two major distinct parental leave reforms on maternal health outcomes, over a period of 5 years postpartum. The health outcomes are objective measures based on all types of medical services provided by physicians. For mothers publicly insured by the public prescription drug plan we can also identify all drugs used, in particular those associated with depressive symptoms. The long time span of the longitudinal administrative data sets allows an assessment of short-run and long-run effects of maternity leave on mothers? health. The empirical approach uses a strict regression discontinuity design based on the day of regime change. The large samples of mothers, who gave birth three months before and three months after the two policy changes (in 2001 and 2006), are drawn randomly from the population of delivering women, all covered by the universal public health care program. We do not find any evidence that the reforms had sizeable impacts on maternal health care costs, either of a physical or of a mental in nature, as measured by physicians? fee-for-service billing costs, prescription drug costs, or the number of hospitalizations. The second expansion has given rise to large fiscal costs over time as well as socioeconomic inequities.