School entry cutoff and the timing of births : evidence from Argentina

Abstract

The distance between the date of birth and the school entry cutoff has been repeatedly used as an exogenous instrument to examine the impact of several educational programs. In this work we analyze the validity of this instrument for the case of Argentina. From a regression analysis, considering multiple waves of the Permanent Household Survey, we detect the existence of discontinuities in the distribution of births around the school entry cutoff (June 30). This suggests that parents act strategically. In particular, these defer the date of birth to days after the cutoff. This result is especially robust when considering a bandwidth of 7 days before and after the cutoff and in men, but not in women. The findings suggest a careful consideration of estimates that assume that the birth distribution is exogenous to the school entry cutoff.