Assortative Matching Using Soccer Data

Abstract

In labor economics, the theory of assortative matching focuses on the mutual selection of workers and firms. Empirical studies using Abowd, Kramarz, and Margolis’ methodology (an estimation of a two-way fixed effects (FE) model) do not provide clear evidence of the existence of assortative matching between workers and firms, that is, there is no clear evidence that good employees work in good firms. In fact, negative or very small correlations are found in the literature using this approach. Andrews, Gill, Shank, and Upward suggest that positive assortative matching has not been observed in the empirical literature because of a limited mobility bias. This article uses Italian soccer championship data to test positive assortative matching. Because job turnover is significantly high in professional soccer, we use numerous movers to detect sorting. Our results indicate that estimating a standard Mincerian wage equation with numerous movers enables us to find a positive correlation between players’ and firms’ FE, meaning that good players tend to move to good teams. This suggests that superstar effects are at work as the unobserved component of the player FE is far more correlated with the wage than the observed component.