Research

Publications



Working Papers


In the Netherlands, an immediate baby boom followed the end of WWII and the baby bust of the 1930s. I propose a novel application of the bunching methodology to examine whether the war shifted the timing of fertility or changed women's completed fertility. I disaggregate the number of births by age for cohorts of mothers, and estimate counterfactual distributions of births by exploiting that women experienced the war at different ages. I show that the rise in fertility after the liberation did not make up for the "missed" births that did not occur prior to the war, as fertility would have been 9.4% higher in absence of WWII.


This paper presents new causal evidence on the “power” of oral contraceptives in shaping women’s lives, leveraging the 1970 liberalization of the Pill for minors in the Netherlands and demand- and supply-side religious preferences that affected Pill take-up. We analyze administrative data to demonstrate that, after Pill liberalization, minors from less conservative areas were more likely to delay fertility/marriage and to accumulate human capital in the long run. We then show how these large effects were eliminated for women facing a higher share of gatekeepers – general practitioners and pharmacists – who were opposed to providing the Pill on religious grounds.



This paper aims to better understand the persistence of gender gaps in adult economic outcomes by studying the influence of pre-birth factors on educational performance early in life. We exploit natural variation in prenatal testosterone exposure between fraternal twins. Such exposure has been shown to affect the organization of the brain in the developing fetus. The twin design helps control for postnatal factors and we also examine interactions between prenatal exposure and postnatal factors. Girls with a twin brother have higher mean birth weights than girls with a twin sister, leading us to predict that they should have higher test scores. However, they lag behind their brothers in terms of math scores by the end of primary school. Our findings suggest that post-birth factors are driving this result, and that these offset any direct biological advantage of prenatal testosterone on test scores. Overall, these results suggest that gender differences in test scores are strongly socially determined.


Works in Progress


Publications in Dutch