Explaining the dynamics of stagnation: An empirical examination of the North, Wallis and Weingast approach

This paper analyzes periods of economic stagnation in a panel of countries. We test if stagnation episodes are predicted by institutional factors and external/internal shocks, as is implied by recent theoretical contributions, and compare the impacts of these variables with those of traditional macroeconomic variables. We examine the determinants of stagnation episodes using multivariate dynamic linear models, fixed-effects logit models, and dynamic random effects probit models. In addition, we analyze whether the included variables have different impacts on the onset of a stagnation episode than on its continuation. We find that inflation, negative regime changes, real exchange rate undervaluation, financial openness, and trade openness explain the incidence of stagnation spells. Only in the case of trade openness, there is robust evidence of a differential impact; it reduces the probability of falling into a stagnation spell, but has a weaker effect within a spell. All models account for unobserved heterogeneity and exhibit a moderate degree of positive state-dependence.

Keywords: growth episodes, stagnation, institutions, dynamic panel data

JEL Classification: O11, O43, C25