How Stable Has Growth Been in Latin America?

Abstract

In this paper, we characterize economic growth among a sample of Latin American countries by means of two panel stationarity tests: those of Carrion-i-Silvestre et al. (2005) and Hadri and Rao (2008). The latter has a number of advantages compared to the first, in that it allows us to control (i) non-observed heterogeneity, both in its form and dates of possible structural breaks in the tendency function; (ii) the cross-section dependency among the units with panel bootstrap methods; and (iii) serial correlation in the errors. Our results show that economic growth in most Latin American countries has slowed significantly following structural breaks. The main implication of this finding is the inability of countries to recover from negative impacts and thus return to the original path of balanced growth.