Growth in Latin America: Models and Findings

The performance of the Latin American economies in the 20th century has been disappointing. Using per capita GDP as a measure, Latin America has roughly maintained its income (relative to the world) over the last 100 years. If, instead the post-1950 performance is considered, the results are more discouraging. Using data from Maddison (2003), it follows that the Latin American economies had a level of income per capita that was 18% higher than the world average in 1950. After 30 years, Latin America’s per capita income exceeded the world average by almost 20%. However, in the early 1990s per capita income in the region was equal to the world average. Moreover, in the last fifteen years it has not improved. Even though Latin America is not the only region that has underperformed relative to the world (e.g. Africa’s performance is even worse), it stands out because its endowment of natural and human resources, its institutions, and the (relative to the world) low level of social conflicts would have made it a natural candidate for a ‘high growth’ region at the turn of the 20th Century. It is this divergence between reasonable expectations and outcomes that makes the study of the determinants of growth in Latin America such a fascinating topic. The objective of this paper is twofold. First, using the results produced by individual country studies as well as region-wide analysis, I discuss the proximate reasons for the poor economic performance of Latin America. I summarize the major findings of both growth and development accounting exercises that allocate differences in output per capita (or its growth rate) to differences in factor inputs and total factor productivity (TFP). Even though it is not easy to generalize from the sample of