Diversification and agricultural production in Brazil: an analysis using spatial models

Abstract

The objective of this study was to analyze the relationship between agricultural diversification and the production of permanent and temporary crops in the year 2017, in Brazilian microregions. For this purpose, some estimators were used, such as the ordinary least squares (OLS), and models that consider the spatial dependence in the data, as the spatial lag regression (SAR), the spatial autoregressive error (SEM), and the spatial autoregressive combined model (SAC), considering the maximum likelihood and the correction for the heteroscedasticity. The results show a negative relationship between diversification and agricultural production, evidencing that the specialization strategy to increase production was efficient in 2017.