Do marginal employment subsidies increase re-employment probabilities?

The purpose of this paper is to test whether a Marginal Employment Subsidy program (MES) is effective in reducing the length of unemployment spells. Our test is constructed in such a way as to isolate the direct effect of the MES from the business cycle effect (i.e. the variation in unemployment that would have normally occurred in the absence of the program). Specifically, we estimate a Markov chain duration model with time-varying covariates and we test if eligibility for an MES increases the hazard rate of leaving unemployment. One of the time-varying covariates is an index of macro-economic performance that controls for the variation in the hazard rate due to business cycle fluctuations. The net effect of the MES offered by the Agenzia del Lavoro of Trento, Italy, is shown to be significantly positive: the Marginal Employment Subsidy program increases the re-employment probability and reduces the expected duration of unemployment by twenty one months.