Heterogeneous impacts on earnings from an early effort in labor market programs

We study whether a labor market program, previously shown to lower unemployment duration, affects job quality. The empirical analysis is based on a randomized controlled trial, conducted in two different counties. We find no effects on women but positive effects for men. In one county, the program increased men's earnings in the short term by 9%, possibly by taxing leisure. In the second county, earnings also increased in the longer run, by about 9%, possibly because of a removal of labor market frictions. The positive effects of the program are heterogeneous, with taxing of leisure time primarily affecting low income earners while removing labor market frictions affecting high income earners.