Family leave policy and child health: Evidence from 19 OECD countries from 1969 to 2010

Abstract

This study examines the effects of family leave policy on eight child health outcomes - five age specific child mortality rates (infant, perinatal, neonatal, post-neonatal, and child mortality rates), low birth weight, and immunization rates for measles and DPT (diphtheria, pertussis, and tetanus) across 19 Organisation for Economic Co-operation and Development (OECD) countries from 1969 to 2010. In addition, this dissertation investigates the extent to which the effects of leave policy vary by period and across welfare regimes. This research contributes to the existing literature (Ruhm, 2000; Tanaka, 2005) by including one additional country, South Korea, a highly developed but considerably understudied country, and by incorporating data from 2001 to 2010. I use data on family leave policy from Ruhm (2000) and Tanaka (2005) and extend it using data from the Max Planck Institute for Demographic Research (MPIDR), Organization for Economic Co-operation and Development (OECD), World Health Organization (WHO), International Labour Organization (ILO), and World Bank. Additional data sources include the United States Social Security Administration (SSA), International Social Security Association (ISSA), and various government sources. I estimate the effects of family leave policy (specially, number of weeks provided) - considering both job protected paid leave and other leave (unpaid or non-job protected leave) - on child health using ordinary least squares (OLS) models. I control for other relevant variables including gross domestic product (GDP) per capita, health expenditures, healthcare coverage, dialysis patients, and fertility and female employment rates. I also include: (1) country fixed effects; (2) year fixed effects; and (3) country-time trend interactions. Missing values are imputed 20 times using the predictive mean matching method. The results suggest job protected paid leave significantly reduces infant mortality (deaths less than 1 year of age) and post-neonatal mortality (deaths between 1 month and 1 year of age). In particular, the largest effects of job protected paid leave are found in reducing post-neonatal mortality; the effects are robust throughout all model specifications. Comparing the effects of other leave (unpaid or non-job protected) and job protected paid leave, other leave has no significant effects on any of the outcome indicators. This suggests that parents do not respond to leave provided without adequate payment benefits or job protection, and mothers may return to work early. As a result, other leave does not have any significant effects on infant health. When investigating the effects of family leave policy by period with models estimated separately by two time periods, somewhat larger effects of job protected paid leave on post-neonatal mortality are found in the earlier period (1969-1989) compared to the later period (1990-2010); however, the difference in the policy effects between the two periods is not statistically significant. This difference may be explained by the fact that it was during the earlier period when most OECD countries provided leave for the first critical weeks and months after birth. In addition, when examining the effects of leave policy by welfare regime type with models estimated separately by regime type, larger effects of job protected paid leave on post-neonatal mortality are found in the Social Democratic and Conservative regimes than in the other regime types; however, the difference in the policy effects across regime types is not statistically significant. This difference may be explained partly by the fact that overall Social Democratic and Conservative welfare state countries provide more generous payment benefits for parents on leave. The concluding section discusses how these findings compare to previous research and explores future research and policy implications.