The Sources of Growth at Different Levels of Development

Cross-country growth regressions have become an increasingly common tool in empirical development research. But these regressions typically do not attempt to distinguish among countries in different stages of development. Two empirical methods are used to test for such differences. Several of the factors known to affect economic growth are shown to operate differently for countries in different portions of the global income distribution. The results have implications for the role of financial markets, openness and human capital in promoting growth. 1What separates wealthy nations from poor ones is one of the cornerstones of economic theory.1 But it has only been in recent years, with the development of powerful computational techniques and data sets, that empirical analysis of growth and why it happens (or, more revealingly, does not happen) has taken off. A common technique in this analysis has been cross-country regressions in which national economic performance, typically growth in per capita product, has been used as the dependent variable. Among the pioneering work in this regard is Barro (1991). The technique has also been applie