Convergence in Colombia’s Total Factor Productivity: a spatial analysis 1992–2022

Abstract This study identifies key determinants of manufacturing value added and estimates Total Factor Productivity (TFP) across Colombia’s departments. It also examines manufacturing productivity convergence, testing the absolute β-convergence hypothesis and estimating the half-life and speed of convergence to assess how quickly productivity disparities diminish. Using data from 1992 to 2022, the study applies the Levinsohn-Petrin (LP) method and the Generalized Method of Moments (GMM) following Wooldridge within a KLHEMIS framework. After estimating input elasticities, TFP is computed as the difference between value added and factor contributions. The research further evaluates productivity convergence through spatial and non-spatial analyses using a Dynamic Spatial Durbin Model and a Unit Root Test for panel data. Results indicate that non-qualified workers, human capital, electricity consumption, and investment in R&D and advertising significantly impact manufacturing value added. Both LP and GMM models suggest decreasing returns to scale. TFP growth follows a cyclical pattern, averaging 2.5% annually. The convergence analysis reveals significant interdepartmental productivity alignment, supporting absolute β-convergence. Additionally, estimates of half-life and speed of convergence provide insights into the rate at which productivity disparities decline. This study offers a detailed regional TFP estimation for Colombia’s manufacturing sector, incorporating spatial analysis. By testing absolute β-convergence and quantifying convergence dynamics, it provides novel insights into productivity growth and regional integration. JEL Codes : O40, O47, O54, L60.