Luma de Oliveira

The aim of this research is to estimate two equations based on the Efficiency-Wage and Search and Matching theories with panel data available for the Brazilian states. Specification tests were made and were found the continuing differences among states, besides the presence of heteroscedasticity and/or autocorrelation, therefore the methodology used in the estimations was dynamic panels, which eliminates problems of endogeneity and omitted variables. The first step was estimating the aggregate matching function and making a prediction of the dependent variable. Then, the second step was estimating the wages’ equation on the baseline theories. One innovation of this study was the specification of the matching variable, along with the persistence of the employment rate and the methodology applied in some indicators such as the destruction rate of jobs. As regards the wages’ equation, there has been also the persistence of this rate, and the high significance of the variables that represent the theories, i.e., the productivity and the matching variables. The results are consistent with the new macroeconomic theories of work market. In other words, the higher the productivity and the better the conditions of the matching, the higher will be the wages’ rate and, thus, the lower will be the unemployment rate of the Brazilian States.