Economic growth dynamics of medium cities in the northeast: 1991 to 2016

Abstract

The growth of Brazilian urbanization in recent decades has pointed to a spatial deconcentration in the sense that medium-sized cities have grown more than the large metropolises. In this context, the study sought to identify these growth patterns, using models of spatial econometrics and tested the hypothesis of conditional convergence of these cities in Northeast Brazil. Estimates were made considering two periods: 1991-2000 and 1991-2016. The robustness of the tests pointed to the need to incorporate the neighborhood effect in econometric models. Thus, spatial error models (SEM) and spatial lag model (SLM) were estimated. The tests indicated the SEM model as more appropriate. Thus, the conditional convergence hypothesis was accepted. That is, smaller medium-sized cities grow faster than larger ones and all in the direction of long-term equilibrium. Better infrastructure, health and per capita income act to increase economic growth, whereas illiteracy and income concentration have the opposite effect.