The Augmented SOLOW Model And The OECD Sample

Abstract

In their influential work on the augmented Solow model, Mankiw, Romer and Weil (1992) showed that cross-section evidence was reasonably consistent with the Solow growth model augmented to include human capital for a wide range of countries.  However, for the sample of OECD countries, they found that the model had low explanatory power and underestimated the output elasticity of physical capital. We revisit their seminal work using data from the recently released version 6.1 of the Penn World Table. We find that the ability of the augmented Solow model to explain the cross-country variation in income per capita in the OECD sample improves significantly. Our results highlight the importance of taking into account changes that take place over time in the collection and measurement of national accounts data in estimating and testing the augmented Solow model.