Avaliação de impacto do Programa de Geração de Emprego e Renda (PROGER)

This article presents the results of an impact assessment of the Employment and Income Generation Program (Proger) - a loan program to small e medium sized firms conducted by the Brazilian government - on formal employment, payroll and unjustified dismissals. Two sources of data were used in the assessment: BG-Proger, which provides information on the program's loan contracts, and Rais, which contains information on the set of formal companies and their employees in the country. The assessment focused on companies that took credit from the program only once. The methodology employed used a recent expansion of the difference-in-differences method in the sense that it makes it possible to estimate heterogeneous impacts of the intervention for different groups entering the program (treatment groups) and over the time, both in the calendar dimension and in the time elapsed after the intervention. The application of the methodology was based on a balanced panel of "cells" formed by the interaction of the dimensions: Federation Units, industry and year of entry into the program. In general, the results show positive and statistically significant impacts on employment, payroll and unjustified dismissals for most treatment groups. The effects on employment and payroll show an inverted U pattern (incomplete at the end) for the calendar time dimension and more monotonic pattern for the relative time to entering the program; the effects on unfair dismissals are more stable in the first temporal dimension and more erratic in the second. The article also contains a partial analysis of the benefits and the implicit subsidies of the programs evaluated.