Essays on corruption, political instability, and military spending

The first chapter of this dissertation re-examines the relationship between corruption and economic growth. Theoretically, I build a model in which corrupt government officials have access to a common pool of resources. However their probability of being ousted depend;, not only on their rent-seeking activities, but also on their provision of public goods, and the (subjective) perception of the level of corruption. The main prediction of the model is that an increase in (subjective) corruption leads to an increase in political instability, which in turns reduces growth. Bringing the theory to the data, I find that corruption indices, which are subjective by construction, do have an adverse effect on economic growth through political instability. This result is robust with respect to different conditioning sets. Overall, my results have policy implications; in particular achieving political stability should be a priority.; In the second chapter, I build a model which helps explain a strong belief regarding economic growth, political instability and the corruption level. As a matter of fact, there seems to be a consensus that countries with high (low) levels of corruption also have low (high) levels of economic growth, and high (low) levels of political instability. To explain this, I build a neo-classical growth model with an endogenously determined political dimension. I assume strategic complementarity among corrupt individuals, that is the more individuals engaged in bribery, the less likely they can be detected. This strategic complementarity is at the heart of the result of the paper.; The third chapter is an attempt to understand the roles of total government spending and military spending in developing countries where ethno-linguistic and religious division is an important problem. Theoretically, I show that the "stick" of military expenditure and the "carrot" of non-military government consumption are often used as an insurance against political instability. Empirically, I show that the theory is borne out by the data.