Macroeconomics Seminar - Kerstin Holzheu (University of Chicago)

Title: 'How Much Should we Trust Estimates of Firm Effects and Worker Sorting? (with Stephane Bonhomme, Thibaut Lamadon, Elena Manresa , Magne Mostad and Bradley Setzler)'.

Abstract: A large number of studies use matched employer-employee data to estimate the additive worker and firm effects wage model proposed by Abowd, Kramarz, and Margolis (1999; hereafter, AKM). These studies tend to draw two conclusions (see the review in Card et al, 2016). First, firms play an significant role in the wage determination, with a typical finding that about 10-20 percent of the variance of wages is attributable to firm effects. Second, the correlation between firm and worker effects tend to be positive but relatively small, indicating modest sorting of high wage workers to high wage firm. While influential, these findings have been subject to much controversy. In particular, several concerns have been raised about model specification and estimation (Card et al, 2016). Building on recent methodological developments that aim to investigate or address these issues (Andrews et al, 2008, 2012; Bonhomme et al., 2017; Kline et al, 2018), we re-examine the contribution of workers, firms, and worker-firm sorting to earnings inequality. Our analysis uses matched employer-employee data from Austria, Italy, Sweden, and the United States. By comparing the results across countries, we shed light on whether the patterns of worker sorting and firm effects are common across several developed economies that differ considerably in labor market institutions.