When Communism fell, each newly liberated state embarked on a path of creating a new regime from existing institutions. The successor states shared a common history of conquest and Russification (in the case of the former Soviet Union) and several generations as Soviet satellites (in the case of Eastern Europe). They also carried pre-existing legacies shaped by geography and demography which, unlike the similar experience of Communist ideology and administration, varied dramatically from country to country. Did these legacies constrain actors in crafting new institutions, condemning some to authoritarianism and backwardness while providing others with the raw materials for success? Or was this inheritance merely a minor stumbling block easily overcome by actors who were then able to shape institutions as they pleased? (1) Ethnic diversity, although obviously not the only product of historical legacy, is nonetheless thought to be a major factor in determining a variety of outcomes in political science, including economic growth, good governance, public goods distribution, and stability, in addition to democracy. (2) The twenty-eight states of the former socialist bloc, whose regimes range from robust and consolidated democracy to unbridled authoritarianism, provide a strong test of the relationship between ethnic demography and democracy because they were all subject to the same political regime and all underwent regime change at the same time. To test this proposition, I first lay out several theories pertaining to the problems of consolidating democracy in plural societies, identifying three distinct mechanisms in the literature--violence, majority domination, and trust--through which ethnicity is thought to directly influence democratization. Next, I test each mechanism using ordinary least squares (OLS) regression to examine the relationships between ethnic pluralism, level of democracy, and the intervening determinants of democracy. Overall, the results offer little evidence that ethnic pluralism influences democracy. Finally, several case studies illustrate the relative unimportance of demography in determining the level of democratic consolidation and identify alternative explanations for the range of outcomes. Although a test of the relationship between ethnic pluralism and democracy can be conducted for many regions of the world, there are several reasons to limit this study to Eastern Europe and the former Soviet Union (FSU). First, whereas using a data set composed of different regions of the world would pose a challenge in controlling for the many factors that differ between regions, the shared legacy of Soviet rule in the socialist bloc automatically controls for historical factors and allows for the isolation of structural factors inherent in a polity, such as ethnic fractionalization, wealth, and geography. Although the states of Eastern Europe and the FSU had experienced varying levels of democracy and statehood over their histories, for the most pan, the Soviet Union prevented internal conflict and the development of independent institutions and civil society throughout its empire. Between 1989 and 1992, each of the twenty-eight countries in the region began implementing independent policies from the Soviet Union for the first time in generations (with the partial exception of Yugoslavia and Albania, which had started earlier), permitting focused comparison of the natural constraints inherited by political actors and their effect on the choices made in the postindependence period. Second, using countries that as of eleven years ago all exhibited a common value oil the dependent variable (regime type) reduces the chances for endogeneity, which is important in interpreting the results of a statistical analysis. Since I am testing the effect of ethnic diversity on democratic consolidation, I can better establish causality if the former is exogenous to the hitter in the data set. The data set used in this analysis is advantageous in this regard because it consists of polities that all possessed a single regime type and have only been democratizing for a short time, providing less of a risk for reverse causality. …