Does childcare matter for maternal labor supply? : pushing the limits of the regression discontinuity framework

We use an extension of the RD approach based on a kindergarten enrollment cutoff date and a new resampling design to estimate the causal impact of subsidized childcare availability on Hungarian mothers' labor market participation around the 3rd birthday of the child. Besides standard fuzzy RD, which holds calendar time constant, we apply an alternative version where child's age is held constant, which enables us to (a) separate the childcare effect from other, age-specific effects, and (b) consider the effect of not only point, but interval cutoffs for eligibility. We combine RD with a difference-in-differences approach using a comparison group of mothers with children aged 4-5 to control for seasonal effects (parent selection, child development, within-year labor market fluctuations). Our estimates indicate that a mother with a 3 year old is 15% more likely to be active if her child is eligible for subsidized kindergarten, corresponding to previous estimates of labor supply elasticity of 0.3-0.75. This suggests that increased subsidized childcare availability and parental leave alone cannot explain the sharp increase in the rate of maternal participation seen around children's 3rd birthday, highlighting the importance of other factors such as separation preferences and flexible work forms.