Identifying Sorting - In Data

We show theoretically how to identify, using wage data alone, whether assortative matching between workers and firms is positive or negative. The results of a Monte Carlo study of calibrated models featuring positive and negative sorting illustrate that the method performs well given the limitations (on sample size, frequency of labor market transitions, etc) of the commonly used matched worker-firm data sets. We apply the method to a large matched worker-firm data set from Germany.