Estimating the returns to educational mismatch with panel data: the role of unobserved heterogeneity

Using longitudinal data from the Swiss Household Panel, this analysis suggests that the cross-sectional estimates of the returns to educational mismatch are significantly biased when unobserved heterogeneity is omitted in the wage equation. The results of the standard fixed effects model indeed demonstrate that the wage returns to education are independent of the job requirements. Hence, this empirical analysis supports the human capital interpretation of the Swiss labour market.