Dynamic models of knowledge-based growth: evidence of bidirectional causality between GDP and R&D efficiency in Latin America

Published: 17 Jan 2026, Last Modified: 09 Feb 2026IWHICI 2026 ACCEPTWITHMAJORREVISIONSEveryoneRevisionsBibTeXCC BY-NC-ND 4.0
Keywords: Economic growth; Gross Domestic Product; GDP; Knowledge efficiency; R&D; Bidirectional causality; Latin America.
TL;DR: This study analyses the bidirectional relationship between economic growth and R&D expenditure efficiency in Latin America between 1990 and 2022, using VAR/VECM, ARDL and simultaneous equation models to assess its causal dynamics.
Abstract: This study analyses the bidirectional relationship between economic growth and the efficiency of Research and Development (R&D) expenditure in five Latin American economies: Colombia, Chile, Brazil, Mexico, and Argentina. Annual series from the Science and Technology Indicators Network (RICYT) were used for the period 1990–2022, applying unit root tests (Augmented Dickey-Fuller, Phillips–Perron), Granger causality models, Vector Autoregression/Vector Error Correction Model, Auto Regression Distributed Lag/Error Correction Model, and simultaneous equations Two/Three Stage Least Squares. The variables included Gross Domestic Product adjusted for Purchasing Power Parity (GDP PPP) and two efficiency indicators: publications and patents per unit of R&D expenditure. The results show significant reciprocal causality in Colombia and Chile, partial relationships in Brazil and Mexico, and unstable effects in Argentina. The estimates reveal positive and significant elasticities in both directions, with an effect of 0.2–0.3% of GDP for 1% increases in expenditure efficiency. These findings confirm the existence of a dynamic, endogenous equilibrium between knowledge and growth, though it is conditioned by institutional maturity and expenditure orientation. The study contributes to the academic debate on endogenous growth by introducing a measure of knowledge efficiency that combines scientific and technological productivity, demonstrating that the quality of R&D expenditure is more decisive than its magnitude.
Submission Number: 4
Loading