Association for Public Policy Analysis & Management

In 2007, while seeking to increase female labor force participation and ease the burden on working women, the Mexican government introduced the childcare program: Estancias Infantiles para Apoyar a Madres Trabajadoras (EI). The roll-out of EI was so aggressive that by 2010 it had offered 352,000 childcare spaces, which doubles that of the 25-year-old social security childcare program. Using quasi-experimental variation of childcare, a difference-in-difference-in-differences approach and Synthetic Control Methods, this paper analyzes women’s labor outcomes and their spouses’ responses. The sharp expansion and uneven roll-out of EI allows to exploit the variation in program's availability over time, across municipalities, and between eligible and similar ineligible families. EI resulted in an increase in female's employment, job stability and income. Husbands within eligible families, spent less time on child rearing and housework, and those who lived with women that started working, were more likely to switch to a higher-paying job.

Results suggest that women increase their labor force participation when they have an exposure to 10 child care spaces per 100 eligible children from 1 to 5 percentage points. The significant economic results, which are between 5 to 17 percentage points, are perceived when municipality have more than 30 child care spaces per 100 EI-eligible children. Men responded increasing their labor force participation as well, but not as much as women (less than 2 percentage points). Hours worked, also increased significantly for this population: 5% for those who faced an EI exposure of less than 10 child care spots, and 17% for those who faced an exposure greater than 40 child care spaces per 100 eligible children. Finally, the IV estimates suggest that the probability of men switching to a better paid job increased by 30 percentage points.

The principal contributions of this paper are four-fold. First, we evaluate one of the most ambitious child care programs in the developing world. Second, we develop a procedure for using Synthetic Control Methods in applications in which the data comes in repeated cross-sections and in which people move in and out of eligibility for treatment over time. This procedure is helpful to many other researchers because we are often not in suitable conditions for which Synthetic Control Methods were originally design – that is, fully longitudinal data in which units do not switch from treatment to control or vice versa. Third, we explore whether men took a higher risk by measuring the change in their propensity to work when their spouses increase the probability of being employed. Fourth, we provide empirical evidence of the impact of female labor force participation on her partner's probability to find a better paid job.