The determinants of displaced workers’ wages: Sorting, matching, selection, and the Hartz reforms

We present a simple new method to decompose the wage effects of displacement into components due to differences in the way that displaced and non-displaced workers are sorted across higher- and lower-paying employers (a sorting effect), differences in the quality of worker–employer matches that they enter into (a matching effect), and differences in their unobservable characteristics (a selection effect). In an extended application, we apply our decomposition to understand how the determinants of displaced workers’ wages in Germany changed following the 2003–2005 Hartz reforms. We find that the wages of displaced workers fell substantially after the reforms, and that over 80 percent of the decline was because they found re-employment at lower-paying employers. Sorting into worse matches explains a smaller 5–9 percent of the wage decline experienced by men, and 12–23.5 percent of the female wage decline. Collectively, the sorting and matching channels explain almost all of the post-reform decline in displaced workers’ wages, and selection played little role.