| In Part I of this study, I rely on Kahneman and Tversky's prospect theory to propose a different perspective on the ideal speed of economic liberalization. On this topic, the neoclassical benchmark is Mussa: if agents are rational and there are no adjustment distortions, shock-therapy is the efficient strategy even when there are mobility costs. Justifications for gradualism have relied on improbable irrationality premises and distortions in the adjustment process or on specialized political economy arguments. I argue that the neoclassical analysis is not necessarily accurate, and prove that if individuals behave according to prospect theory, gradualism is the optimal policy when liberalization involves a large enough loss to workers employed in the inefficient sector. I use this result to analyze some recent evidence from the Eastern European transition. In particular, I argue that shock-therapy is the efficient policy for higher-income countries only. Empirical findings are consistent with this hypothesis, contradicting previous interpretation of the data. Finally, I find that the model correctly predicts most countries' liberalization strategy, suggesting that prospect theory provides an efficiency reason why some countries have preferred to liberalize gradually.;In Part II of this thesis, I discuss the relation between income distribution and growth in developing countries. On this matter, most studies have found a negative relation between inequality and growth, but have consistently ignored the particular features of developing countries. I prove that in these countries the link between inequality and growth has a U-shaped form. To model this, I build on Galor and Zeira's paper on income distribution and individuals' investments in human capital given imperfect credit markets. I suggest, however, that there are congestion costs in the education process, and show that when these congestion costs are significant enough, a decrease in the number of investors caused by more inequality is eventually offset by an increase in productivity growth. Hence, at high enough levels of inequality the relation between this variable and the rate of growth is positive. Empirical findings are consistent with the implications of the model. |
