|
Growth
Theory Charles Leung and Research Fellow Danny Quah raise some simple theoretical warnings on reading too much into the link between convergence and endogeneity of the growth process. In Discussion Paper No. 1383, they strengthen previous research on stochastic growth models, which shows that cross-country convergence can occur even in the presence of increasing returns to scale, by showing that without auxiliary, purely statistical and non-economic assumptions on cross-country interaction, the notion of convergence in a representative-economy model cannot be sensibly addressed. These assumptions, however, which to date have been left only implicit, call into question some widely accepted empirical findings.Leung and Quah argue that the way out of these difficulties requires two departures. Theorists should spell out explicitly the nature of cross-country interaction, and empirical researchers should use models of distribution dynamics instead of regression-based analyses. The main lesson taken from the theoretical manipulations is that the link between convergence empirics and theoretical notions of exogenous and endogenous growth is tenuous, but not uninteresting. It calls for theoretical modelling that is explicit about the relations between individual cross-section units, and will thus mesh well with recent theoretical work on social and economic interaction.
Discussion Paper No. 1383, April 1996 (IM) |