Sorting on-line and on-time

Abstract

Using proprietary data from a Chilean online job board, we compute sorting between workers and job positions during the application stage (ex ante) and predict sorting in the flow and stock of created matches (ex post) for different type measures. We find strong evidence for positive and procyclical correlations between workers and job types. Since ex ante and ex post sorting are very similar, we conclude that sorting is largely generated at the application stage. This suggests that theoretical models of sorting with directed search are a promising path for future research.