Evaluating welfare reform under program heterogeneity and alternative measures of success

This paper aims at presenting an assessment of welfare reforms under a framework of program heterogeneity and alternative measures of success. We focus on a specific welfare program –Madrid’s Ingreso Madrileño de Integración (IMI)– which comprises heterogeneous subprograms. We test whether work-related subprograms perform better than general activities aimed at improving life skills. We also try to identify which work-related subprogram works best. The availability of a large database of administrative records (over 50,000 spells) matched with a special survey conducted for former welfare recipients makes possible to develop different types of evaluation strategies on the basis of multiple participation states. Our results show that intensive employment activities yield remarkably better results than general work-related schemes or life skills activities. However, increasing work participation does not automatically lift participants out of material hardship.