Approaches to learn about employer learning

The empirical literature on employer learning assumes that employers learn about unobserved ability differences across workers as they spend time in the labour market. This article describes testable implications that arise from this basic hypothesis and how they have been used to quantify the contribution of job market signalling and human capital in measured returns to education. While the empirical basis is still thin, the results suggest that signalling contributes at most about 25% to the observed returns to education.