The middle income trap (MIT) is a concept that is often used in the economic development literature but is not defined clearly enough. The purpose of this study is to revise previous discussions on MIT and to make an empirical investigation of the existence of MIT in 44 middle-income countries based on Ye and Robertson's (2016) time series approach. Unlike conventional time series analyses, this study takes into consideration non-linearity in time series and smooth transition of structural breaks. The empirical findings of this study show the following results. There is strong evidence of the existence of MIT in seven countries. The stochastic convergence is valid for twenty one countries, including Turkey. The trend findings of the countries show that the income gap between ten countries and the US is decreasing quite slowly or increasing steadily. These results also indicate that economic and institutional reform requirements maintain their priority in the political agenda in most middle income - especially in Latin American - countries. © 2020 Bogazici Universitesi. All rights reserved.

2-s2.0-85095884590