Modern models of economic growth have abandoned the standard neoclassical framework to consider it as an endogenous outcome of policy choices, which may therefore influence whether poor countries and regions catch up with richer ones or remain in poverty. Many empirical studies have found that poor and rich economies all appear to converge at a stable, uniform rate of about 2% per annum, and this striking result has been used to pronounce on such diverse issues as US world leadership in productivity, German reunification and regional redistribution, both within individual countries and across the European Union. In Discussion Paper No. 954, Research Fellow Danny Quah reinterprets well-known theoretical models of accumulation to show that the standard empirical methods produce misleading answers in precisely the cases most pertinent to convergence analysis. He then proposes a more straightforward method of studying convergence, based on a direct examination of the dynamics of the cross-sectional income distribution.

Quah applies this method to cross-country income data to reveal that persistence and immobility are more predominant than steady convergence at 2% or otherwise. The dynamics provide some evidence of an extreme form of `convergence club', but there is an overwhelming diversity of growth experiences. This model captures the experience of both the small number of countries undergoing apparently miraculous transitions from poor to rich and the great majority that simply remain rich or poor. Conditioning on variables that are often thought to explain growth physical capital investment, secondary school enrolment and a dummy for the African continent does not overturn the characterizations of these dynamics, but it does narrow the cross-sectional dispersion of incomes. This suggests that earlier findings of conditional convergence may arise from the incorrect interpretation of `convergence club' dynamics or polarization, as convergence and a slight reduction in the spread of the cross-country income distribution. In previous analyses, these have both obscured the actual and dominant features of persistence, immobility and wide diversity in cross-country growth experiences revealed here.

Convergence Empirics Across Economies with (Some) Capital Mobility
Danny Quah

Discussion Paper No. 954, May 1994 (IM)