Essays on Multiple Exchange Rate Systems

This dissertation measures the impact of multiple exchange rate systems on economic performance and on net capital flows in developing countries. The literature on the effectiveness of capital controls has some problems. Two of them are that it often ignores the endogeneity of capital controls, and that most of the evidence is dominated by some country specific studies. This dissertation fills this gap. It uses a multicountry panel to quantify the effects of parallel rates in the economy, but in doing so it explicitly models the endogeneity of multiple exchange rates. The dissertation is structured as follows. Chapter 1 evaluates the relationship between parallel exchange rates and economic performance in the post Bretton Woods period (1974-2001). The main findings are not only that parallel exchange rates are more likely to be adopted when economic performance is bad, but also that they hurt economic performance, indicating the existence of a negative feedback mechanism linking economic performance and parallel markets. It also finds that liability dollarization and high debt service are possible determinants of the likelihood to segment the foreign exchange market. Chapter 2 evaluates the effectiveness of multiple exchange rates systems as a policy tool to stop capital outflows. Controlling for push and pull factors that drive capital flows, and using data from 46 developing countries for the 1980-2001 period, it cannot find empirical support for the claim that segmenting the foreign exchange market stops capital outflows. The evidence suggests that multiple exchange rates systems do not have any effect on capital outflows, at best. At worst, the evidence suggests that parallel exchange rate systems increase capital outflows rather than discouraging them. This last result can be rationalized with a policy signaling model for capital controls.