THE MIDDLE-INCOME TRAP FROM AN INTERNATIONAL OVERVIEW: A LOGIT PANEL DATA ANALYSIS

Abstract

This paper analyzes the probabilities of economies to lie into the middle-income trap by using a Logit Panel Data model. We define the trap with more encompassing criteria: average GDPs per capita relative to the US one and GDP per capita growth rates for ten-year averages in both cases. To determine those probabilities, we include two different sets of covariates. Firstly, a set of covariates that might be influenced by policymakers, e.g., inflation, exports, and education. Secondly, another set of covariates that are independent of policymakers’ influence; a trend, a dummy for the international financial crisis, growth rates of world leader economies (US and China), growth rates of continental leader economies, and dummies by continents. We find that domestic variables are essentialto determine the probability of lying into the middle-income trap. Also, the independent covariates are significant. The trend shows that the trap is a stable equilibrium where economies tend to stay for a long time. A global economic crisis can push the countries out of the trap, but the impulse makes them back to lower-income stages. Our model also finds that regional linkages with surrounding economies are essential for middle-income economies Keywords: Middle-Income Trap, International Economics, Logit Models, Panel Data