ESSAYS ON LONG-RUN ECONOMIC GROWTH AND DEVELOPMENT

This dissertation examines two issues pertaining to economic growth and development in the long run. Chapter 1 contributes to the understanding of the Neolithic transition, a milestone in early economic development, by investigating the role of species diversity in the adoption and diffusion of agriculture. Using high resolution geo-referenced data on the distribution of terrestrial animal species around the world, we show that in a global sample of preindustrial societies from the Ethnographic Atlas, there is a robust inverted-U relationship between species richness and reliance on agriculture for subsistence. Our tentative exploration of the mechanism behind this relationship points to the important role of sedentism as a mediating link. Furthermore, we show that a similar inverted-U association holds between species diversity and the timing of agricultural transition across Neolithic sites in Europe and the Near East. The non-monotonic relationship revealed in our study suggests that competing theories connecting the abundance and diversity of food sources in a local environment to the expansion of diet breadth among foragers (known as the broad spectrum revolution) and the eventual transition to agriculture are complementary rather than mutually exclusive.Chapter 2 examines the so-called middle-income trap, a tendency to experience prolonged economic stagnation said to exist among the countries with middle-income status, by proposing a new approach to detect the presence and mechanism of the trap. Based on the results of a structural break test on 173 economies' growth rate series during the 1951-2014 period, I show that economies in the 40-80 percentiles of the cross-country real GDP per capita distribution face both increasing risks of experiencing growth slowdowns and diminishing chances of experiencing growth accelerations as they transit to a higher per capita income level. Given that this relative income range falls in the middle part of the global cross-country income distribution, the observed unfavorable evolution of break probabilities can be interpreted as indicating the presence of a middle-income trap. However, the study also finds that the prominence of the trap observed in data disappears once some confounding factors are controlled for in the multinomial probit regressions, including lagged growth rates, human capital, the share of service value added in GDP, and quality of democratic institutions. This points to the possibility that it is the combination of these factors, rather than relative per capita income level per se, that are giving rise to stagnation among middle-income countries. Although the regression results in this study should be treated with caution due to the likely endogeneity of these control, this finding is still indicative of the need for reorientation of the way we think about the trap.