Reassessment of the Proximate Determinants of Income Levels and Growth of Nations

Abstract

This paper provides further evidence of the inconsistency of the Solow growth model with the observed cross-country variation in income levels and growth rates. I tackle the assumption of constant and identical technical progress rates across countries by providing country-specific estimates derived from a growth accounting exercise. The estimates of total factor productivity (TFP), commonly used as proxies for technical progress, appear to vary considerably across countries and mostly fall far below the alleged identical rate of 2%. Using these country-specific rates of TFP, I show that the testable predictions of the Solow growth model are often not fulfilled. Deviating from the assumption of identical technical progress rates leads in most cases to poor compliance of the model with the observed disparities of income per capita across nations. The model performs relatively better when testing the out-of-steady-state dynamics and the implied rates of convergence of growth rates. The fragility of the Solow growth model is also documented for a varied array of data sources, country samples and time periods. I conclude that the proximate determinants of income levels and growth rates of nations should be complemented by the fundamentals prior to probing the effects of specific policies on income levels and growth.