Spatial decomposition of the Brazilian agriculture Total Factor Productivity (TFP)

Abstract

The main objective of this research was to implement a spatial decomposition of Brazilian agriculture Total Factor Productivity (TFP), which includes direct (own) and indirect (spillover) components, in the context of an autoregressive spatial production (SAR) frontier analysis for a panel data for the censuses years 1995/96, 2006 and 2017, published by the Brazilian Institute of Geography and Statistics (IBGE). The SAR boundary is estimated using maximum likelihood methods considering the endogenous spatial lag term. An empirical analysis was conducted using data referent to 510 immediate geographic regions (IGR) as references for the production units. The results demonstrate that the TFP growth was 3.87% per year on average, considering the entire period. The variables rural credit, agricultural suitability, highway road, no-till, soil correction, higher education and technical assistance, were statistically significant to determine the TFP growth. The results suggest that the TFP growth is spatially and temporally correlated, and the inclusion of spatial spillover and its local and global effects influenced TFP in the censuses periods analyzed.