Agricultural Productivity Convergence in Latin America: The Role of Research and Development, Knowledge Spillovers, and Education Spending

Abstract

This study examines the evolution and drivers of agricultural productivity, measured as Total Factor Productivity (TFP), across 10 Latin American countries from 1981 to 2012. Earlier studies using traditional time series unit root and cointegration methods have assumed common short‐ and long‐run parameters and therefore tended to confirm convergence across countries. This study estimates TFP using a stochastic production frontier model, then tests for convergence using a Panel Error Correction Model that allows for differences between countries and includes variables influenced by policy: cumulative R&D investment (as a proxy for knowledge), trade in capital goods (to capture knowledge spillovers), and education spending (as a proxy for human capital). The study finds no evidence of absolute convergence, that is countries are not all heading toward the same productivity level. However, conditional convergence toward different steady states is observed for all countries except Guatemala. Although most countries are moving toward separate productivity levels, investment in R&D, trade openness, and improved education can both close productivity gaps and raise overall long‐run productivity. The findings suggest that targeted policies in these areas are essential to support productivity growth in lagging countries.