A desigualdade reginal da renda no Brasil: uma análise da hipótese de convergência

Abstract

This paper analyses the convergence hypothesis among per capita income of Brazilian states and counties from 1970 to 2000. By means of cross-sections regressions, temporal analysis of dispersion, non-parametric estimates of probability distribution, as well as, Markov’s matrices of transition, the hypothesis of absolute convergence among state and municipal geographic units has been rejected. During the period of analysis, the polarization of per capita income was found in two convergence clubs, both in relation to states and counties. The first is comprised of geographic units of Northern and Northeastern regions, the second of Southeastern, Southern and Central-Western areas. Except in the 1980’s, a strong process of absolute convergence among the units of those subgroups was indentified. The hypothesis of conditional convergence of municipal per capita income was also found. The human capital measured by means of levels of education and rate of municipal urbanization indicates that these are the conditioning variables of greater relevance for the convergence process.