Essays on the Political Economy of Debt in Emerging Countries: 1. Strategic Debt and Political Frictions in Small Open Economies. 2. Political Cost of Default and Business Cycle in Emerging Countries

This thesis is composed by two articles. In the first paper, co-authored with Roberto Pancrazi, we study the relationship between sovereign debt and political frictions. We model political frictions as a disagreement among parties about distribution of resources. When analyzing a small-open economy framework we find two important results. First, when considering standard utility function (CRRA with risk aversion parameter greater or equal to one) political frictions induce saving (not borrowing) incentives. Second, when introducing retrospective voting, for which electoral outcomes are affected by recent economic performance, we find that more severe political frictions indeed lead to stronger borrowing incentives. Then, we use the theoretical predictions of our model to structurally estimate the country-specific degree of retrospective voting using data on debt, quality of institutions, and election probability in 56 emerging and transition economies. We find that retrospective voting is strongly related to corruption indices. In the second paper I study the effect of political frictions in a model where repayment of sovereign debt is not enforceable. Sovereign default models that study how income fluctuations and the level of debt affect default risk when sovereign debt is non contingent, are successful in explaining business cycle in emerging economies by matching the stylized facts of main economic aggregates in normal and default periods but they fail in reproducing jointly the large levels of debt and spread observed in the data. I introduce political uncertainty in the standard default model of Arellano (2008): the incumbent has an exogenous probability of not being reappointed in the next period, but in the case she decides to default, there is a larger probability of losing power. Calibrating political uncertainty on Argentinian polls data, the model generates realistic levels of debt to gdp and spread without affecting the performance on the other business cycle statistics.