Re-Examining Mankiw–Romer–Weil (MRW) Growth Model: Evidence and Policy Implications

Abstract

This study aims to re-examine the Mankiw–Romer–Weil (MRW) growth model using Bayesian nonlinear regression with prior information specific to the physical and human capital elasticity coefficients. The Bayesian approach shows superiority over the frequency approach when it flexibly combines prior information about parameters with information from the data, thereby providing more reliable results. The results of re-estimating the MRW model show that the Bayesian estimate of the elasticity coefficient of physical capital is similar across the three samples but smaller than that of MRW. In comparison, the estimate of the elasticity coefficient of human capital is similar for the two samples of non-oil and intermediate countries. It is more significant than that of MRW. Moreover, the elasticity of human capital is more significant than physical capital's across the three samples.