Agricultural productivity in Paraguay: stochastic frontier estimation using rural household data

Abstract

This paper investigates the micro determinants of agricultural productivity at a farm level in Paraguay, using the Permanent Household Survey (EPHC) from 2017 to 2019. We estimate an agricultural production function where the main determinants of output are land, machinery, labor, and purchased variable inputs using Cobb-Douglas and a Translog production frontier. Econometric results show that the most productive areas such as Paraguari, Guaira, Caaguazu, and Caazapa from Southeastern Paraguay are the most efficient. These areas have more infrastructure and have traditionally been the main agricultural areas. We also see that Western Paraguay, predominantly devoted to livestock, is less efficient. More efficient farmers have observed a slowdown in the growth rate of Technical Efficiency (TE) relative to districts such as Ñembucu, Itapúa, or Alto Paraná. While the latter areas had lower TE, Total Factor Productivity (TFP) growth is higher than in the main production areas. The difference between regions is probably due to places that lagged previously and received most of the investments in the last years. From a policy perspective, Paraguay can obtain productivity improvements from the intensification of agriculture rather than incorporating new lands. Considering regional specialization in agriculture and livestock, public and private investment should boost farmers' comparative advantage by region.