Understanding the Sources of Earnings Losses After Job Displacement: A Machine-Learning Approach

Andreas Gulyas.

(University of Mannheim).

Understanding the Sources of Earnings Losses After Job Displacement: A Machine-Learning Approach.

Abstract:

We document the sources behind earnings losses after job displacement adapting the generalized random forest due to Athey et al. (2019). Using administrative data from Austria over three decades, we show that displaced workers face large and persistent earnings losses. We identify substantial heterogeneity in losses across workers. A quarter of workers face cumulative 11-year losses higher than 2 times their pre-displacement annual income, while another quarter experiences losses less than 1.1 times their income. The most vulnerable are older high-income workers employed at well-paying firms in the manufacturing sector. Our methodology allows us to consider many competing theories of earnings losses prominently discussed in the literature. The two most important factors are the displacement firm’s wage premia and the availability of well paying jobs in the local labor market. Our overall findings provide evidence that earnings losses can be understood by mean reversion in firm rents and losses in match quality, rather than by a destruction of firm-specific human capital.

Attendance confirmation.