Heterogeneity in Growth Processes: Estimating Growth Regressions using Panel Data

This paper is concerned with estimating growth regressions for a panel of 104 countries with data spread over a 24-year period. The paper employs panel data estimation techniques. An important concern is whether growth regressions estimated for a large group can be replicated for smaller sub-groups of countries. The problem of parameter heterogeneity is investigated, and the results of the paper show that there is considerable parameter heterogeneity in the growth equations across groups. The major conclusion of the paper is that growth processes appear to be widely divergent across sub-groups of countries making the task of prescribing policy far more challenging and, hence, pointing to the need to incorporate country-specific institutional and political factors while recommending policies for growth.