Empirical evidence has repeatedly demonstrated that countries do not convergence to the same level of development, they may diverge, and multiple equilibriums are possible. The existence of poverty traps has been offered as one explanation for these different experiences. Poverty traps are observable at the sub-national level, resulting in marginal territories within a country, ghettos within a city, or even at the individual level.
This question is most often analysed at the macroeconomic level, and it has nourished a rich and long lasting body of both theoretical and empirical literature. Similar bodies of literature have emerged at intermediate and micro levels, e.g., models of economic growth with multiple steady state equilibriums, or of nonlinearities in asset accumulation pathways, as well as numerous variants of the new economic geography explaining regional core-periphery divergences. The implications of these different explanations are far from trivial. They can contribute to the design of social policies, some of which may fail to eradicate poverty if the reasons why some individuals are trapped in poverty are not clearly identified. They contribute to our analytical understanding of why globalization does not benefit everybody everywhere, and can immiserate some populations.
Recent evidence highlights that a broader understanding of the causes for entrapment in poverty is needed. For instance, governance issues, rent seeking, and political economy issues may create a vicious circle where high logistics costs, low traded volumes, capture of rents in services, and increasing marginalization reinforce each other. At the microeconomic level, the lack of human capital has always been identified as one of the main obstacle to convergence. Further, in the case of least developed countries, a recent and fruitful strand of the literature shifts the focus to health issues. Recurrent and widespread diseases like malaria, HIV/AIDS, as well as neglected diseases such as diarrheas and respiratory diseases, are shown to alter dramatically children's cognitive capacities. This alteration has permanent effects, which affect the quantity/quality of school performance at the macro-level.