In pursuit of excellence Four (natural) experiments in the economics of education

In this thesis I report on the causal effects of a common policy for gifted and talented (GT) students in Dutch education. Exploiting a regression discontinuity (RD) empirical design, I find that the GT policy does positively effects student performance in math, languages and in other subjects. I also find a positive income effect for GT-students because of their subject choice in tertiary education.
A replication of this GT-policy at three Dutch schools produces similar positive effects on student performance. A difference-in-difference (DD) design, a RD strategy and a combination of both, show that more able students seem to profit more than the marginal GT-student. I also find that exposed students are more likely to continue their school career in a more ambitious scientific profile.
In chapter 4, I rule out that feedback of the used GT-selection test drives the GT-effect on student performance.
Chapter 5 evaluates a pilot program on a new mathematical curriculum. In this new curriculum context learning is replaced with more emphasize on concepts and mathematical thinking. The theory of stereotype threat predicts that the performance of girls could be harmed by a mathematical curricula change, especially when girls perceive this change as making math more difficult. I find that girls tend to shy away from advanced (wiskunde B) math at HAVO level. This effect is not present at VWO level. Girls do not perform worse than boys when exposed to the new curriculum.