Moral Hazard and Liquidity Effects of Unemployment Benefits - Evidence from Germany (Thesis)

Abstract

The effect of unemployment benefits on unemployment durations is a long-studied topic in public finance. Most papers follow a reduced-form approach and show that higher benefits lead to longer unemployment durations. Few papers, however, make use of the estimated parameters to learn something about the trade-off between the intended and unintended effects of providing unemployment benefits. The conventional view - longer unemployment durations caused by higher benefits are purely due to a moral hazard effect - neglects the important role of the consumption-smoothing effects of unemployment benefits. Analyzing a sample of male job-losers aged 25-57 years extracted from the GSOEP and ranging from 1994-2007, I find significant evidence for the existence of both, a distortionary moral hazard effect and a non-distortionary liquidity effect. My baseline estimates suggest that the scope of the liquidity effect is rather limited, though. It is estimated to account for only about 16% of the additional duration length that is caused by higher benefits, while the moral hazard effect is calculated to account for the remaining 84%. The average level of benefits observed in the data is likely to be above the optimal level, but this might be different for the post-Hartz period from 2005 on.