On the incidence and effects of job displacement: Evidence from Sweden

Abstract

This paper investigates the effects of job displacement on outcomes such as annual earnings, unemployment, wages and hours worked. It relies on previously unexplored administrative data on all displaced workers in Sweden in 2002, 2003 and 2004 which are linked to employer-employee matched data at the individual level. By linking the data to military enlistment records, the paper assesses the selection into displacement and finds that workers with low cognitive and noncognitive skills are significantly more likely to be displaced than high-skilled workers. The analysis of displacement effects suggests large and long-lasting welfare costs of displacement. Moreover, studying the heterogenous impacts of job displacement across cognitive and noncognitive skills reveals that, although workers with high skills fare better than low-skilled workers in absolute terms, there are no significant differences in the recovery rates across skills. Finally, by using administrative data on displacements, it is possible to assess quantitatively the bias that results from previous studies not being able to separate quits from layoffs.