Testing Acemoglu’s Critique: A Re-examination of the Mankiw-Romer-Weil Model Using GMM

Abstract

This article revisits the Mankiw et al.’s specification of the Solow growth model by addressing the endogeneity concern raised by Acemoglu. To do so, we re-estimate the Mankiw et al.’s equation of long-run cross-country income difference via generalized method of moments (GMM) by using the future-time reference of language as an instrument for physical capital and the Protestant missionary activities in 1923 and primary school enrolment in 1900 as instruments for human capital for 1970–2019. Additionally, the GMM estimates allow us to analyse the role of fundamental causes of economic prosperity, determined by cultural and institutional factors, in explaining cross-country income differences. Although the numerical similarity of the ordinary least squares (OLS) and GMM estimates shows that the benchmark Mankiw et al.’s results are robust to endogeneity criticisms, their interpretations differ. The OLS estimates primarily capture the impact of proximate causes on cross-country income differences, while the GMM estimates reveal the role of capital accumulation, inferred by institutions and cultural factors, in explaining long-term cross-country income differences.

JEL: O24, C26, E24