WHO BENEFITS FROM A STEM-EDUCATION? ESTIMATING MARGINAL TREATMENT EFFECTS FOR SWITZERLAND

Back to Page Authors: Aderonke Osikominu, Gregor Pfeifer, Tim Ruberg

Keywords: returns to education, STEM, IV, MTE

Abstract: This paper estimates average and marginal returns to a STEM-education using 2SLS regressions and marginal treatment effects (MTEs). Using representative data from college graduates in Switzerland, we model the STEM-major choice by using a measure of the relative distance to the next technical university as an exogenous instrument for identification. Regardless of the model specification, the relative distance to the next technical university has an economically and statistically strong negative effect on the probability of graduating from a STEM-major. The results from 2SLS regressions and MTE estimations consistently suggest positive returns to graduating from a STEM-major of 13.62% and between 6.90% and 40.90%, respectively. Statistical tests of (inverse) selection on gains cannot reject a flat MTE curve suggesting any effect heterogeneity based on unobserved characteristics.