Abstract

This project tries to analyze the allocation of workers to jobs and the wage distribution in Germany. It tries to reconcile prominent empirical models of wage dispersion (Abowd et al.,
1999; Card et al., 2013) with theoretical sorting models (Shimer and Smith, 2000; Eeckhout and
Kircher, 2011; Hagedorn et al., 2016). We check whether empirical fixed effect models provide a
valid approximation of observed wages and matching patterns for a large part of the data. For
low-type workers wages should decrease in the type of the firm a worker is matched
with. This prediction of theoretical sorting models is at odds with the monotonicity assumption
of fixed effect models. After ranking both workers and firms, we want to show that low-type workers
have become increasingly sorted into low-type firms over time, especially out of unemployment.
This increase could be driven by  selection into wage-maximizing matches at the bottom of the firm
type distribution.