CRESCIMENTO ECONÔMICO E NÍVEL DE ESCOLARIDADE: TEORIA E ESTIMATIVAS DINÂMICAS EM PAINEL DE DADOS

In this paper, we develop a model that has as main outcome the existence of a non-linear relationship between growth and level of education (human capital). This non- linearity condition accrues from empirical estimation done at international level. Moreover, this non-linearity explains the contradictory results of the role of human capital, since a linear relationship has always been the assumption. The tests indicated that the basic econometric model should be a random effect model; however, this model assumes as a priori hypothesis the causality running from human capital. To solve the matter, we fit a dynamic panel data model for Brazilian States. The dynamic specification eliminates the causality problem and even considers the feed back effect of growth on education. The dynamic estimation found education lagged five periods to cause economic growth in the form of an inverted U relationship. The level of education (human capital) that generates the maximum growth rate lies between 4.5 and 4.7 years. The foremost implication of this result is that States with level of education below this range should have as priority education investment.