Abstract

Since we introduced the term “middle-income trap” in 2006, it has become popular among policy makers and researchers. In May 2015, a search of Google Scholar returned more than 3,000 articles including the term and about 300 articles with the term in the title. This paper provides a (non-exhaustive) survey of this literature. The paper then discusses what, in retrospect, we missed when we coined the term. Today, based on developments in East Asia, Latin America, and Central Europe during the past decade, we would have paid more attention to demographic factors, entrepreneurship, and external institutional anchors. We would also make it clearer that to us, the term was as much the absence of a satisfactory theory that could inform development policy in middle-income economies as the articulation of a development phenomenon. Three-quarters of the people in the world now live in middle-income economies, but economists have yet to provide a reliable theory of growth to help policy makers navigate the transition from middle- to high-income status. Hybrids of the Solow-Swan and Lucas-Romer models are not unhelpful, but they are poor substitutes for a well-constructed growth framework.