The project Ethnic Parties and Democratic Stability expand a constructivist dataset on ethnicity and institutions (CDEI), which currently covers 100 countries and 1346 parties for the year 1996, to include two additional years 1976 and 1986. CDEI is based on a classification of political parties around the world in a given year as ethnic, multiethnic and on-ethnic. The Principal Investigator uses this classification to generate over forty variables related to ethnicity and institutions. The variable that gets the highest priority in the time-series expansion is EVOTE -- the percentage of vote obtained by ethnic political parties in each country for the year 1996. The PI codes several other variables related to the implicit and explicit mobilization of ethnic identities by political parties. The purpose in building and expanding CDEI is to test competing propositions on the rise of ethnic parties and democratic stability. According to a classic proposition in empirical democratic theory, ethnically divided societies invariably give rise to ethnic political parties and ethnic political parties destabilize a democratic system. But other work argues or implies that no matter how they arise, ethnic parties may have the opposite effect they may actually stabilize democracy by institutionalizing discontent which might otherwise be expressed through violence. Because of the lack of cross-national data on ethnic political parties and the degree of support they obtain in democratic systems, neither proposition has been empirically verified. We do have cross-national data on ethnic groups, which, input into the Index of Ethnolinguistic Fractionalization (ELF), have been the basis for studies that explore the relationship between ethnic diversity and democratic stability. But these data, and the ELF index, are divorced from a theoretical understanding of the nature of ethnic groups. Although theories of the origin of ethnic groups are constructivist, the data on ethnic groups and the ELF measure remain resolutely primordialist. As a result, we do not know what the relationship between any aspect of ethnicity and democratic stability is despite the large body of work on the subject. CDEI, which incorporates constructivist insights, and is specifically designed to test propositions in democratic theory about the origins and effects of ethnic political parties, aims to fill one part of this gap. Although CDEI is motivated by my own interest in testing propositions linking ethnic party behavior and democratic consolidation, the data it generates can be used to answer a very broad range of questions in the social sciences about the origins and effects of politicized ethnic identities and therefore to inform public policy on the relationship between politicized ethnic identities and political and economic outcomes. A sample of such questions includes the following: What is the effect of the explicit politicization of ethnic divisions on some outcome of interest, including war, riots, economic growth, public policy, welfare spending and so on? Is the politicization of particular types of ethnic divisions associated (e.g. region or religion or language or tribe) associated with particular types of outcomes? What determines the size of the coalition that an ethnic party is likely to mobilize? Are we more likely to see the ethnification of politics in new democracies? Is there a link between colonial history and the degree of ethnic politicization? How rapidly do politically activated ethnic identities change over time and what explains such change? Is there a link between ethnic majoritarianism and conflict? And so on. Broader Social Value: The Principal Investigator facilitates the broad use of the data generated by CDEI in two ways: (1) By inviting the involvement of other scholars in the design of the dataset, making the data widely available and offering an intensive introduction in the design of the dataset to those who wish to use it. The researcher does this through a combination of web-based portals and face-toface workshops. (2) By integrating the data collection process into graduate student education. Graduate students are currently involved as collaborators and co-authors in all aspects of the project, and the researcher plans to introduce a new laboratory course that combines theoretical instruction with hands on experience in data collection at NYU.