Research & Data

Research Interests

Ethnic and Electoral Politics in India; Religion and Politics; South Asia; Role of the Media in Elections; Regime Types and Conflict

Dissertation Abstract:     Masjid versus Mandal: Ethnic Voting in India

“In an ethnic party system…there are virtually no uncommitted votes to be had on the other side of the ethnic boundary.” 
–Donald Horowitz

My dissertation challenges common scholarly assumptions that ethnic groups’ intense and incompatible ethnic preferences are the main driver of electoral politics. I show that this is not the case by examining how one ethnic group, Muslims in India, are pursued by parties affiliated with other ethnic groups and the conditions under which Muslims support ethnic or non-co-ethnic parties. My research relies on extensive fieldwork in India, including semi-structure interviews and an author-conducted survey, to argue that many voters are willing to cross ethnic lines in order to vote for the winner, and that parties choose candidates to win under these conditions. Unlike existing studies of voting in India that focus on differences in electoral outcomes between states, my main unit of analysis is the electoral district/constituency. I am able to do this by using particularly fine-grained data on the voting choices of individual Muslim voters across electoral districts, the ethnic composition of all electoral districts, and data on candidate identity, in one state, Uttar Pradesh. By focusing on electoral districts, comparing individuals from the same ethnic group, and holding constant institutional legacies, I am able to highlight when individuals will make pragmatic cross-ethnic choices in response to electoral dynamics rather than blindly voting along identity lines.

My work opens up new debates in ethnic politics by making a case for strong incentives for voters to engage in cross-ethnic voting even in a highly ethnicized polity. In countries like India, where citizens are indiscriminately characterized into ‘vote blocs,’ the question of how ethnicity mediates political choices is of critical importance. By focusing on India's large and marginalized Muslim religious minority, I also address an important literature on the rise of Hindu nationalist politics in India. My work both dispels myths about Muslim political machinations, and highlights how minorities in developing democracies respond to the political system to ensure their welfare.

Data Sources

1) Devasher 2012 Survey Data on Muslim Voters in Uttar Pradesh

Following the 2012 state legislature elections in Uttar Pradesh, India,I conducted what I believe to be an unprecedented survey of Muslims which provided data on individual vote choices in 2007, 2009 and 2012, relationships with political parties and candidates, views on contemporary political issues as well as demographic data such as caste, income, and education. The survey included 1,685 respondents across 45 randomly selected state legislature constituencies.  The survey was designed to study variation in the voting behavior of a single minority ethnic group across a state. It showed variations in vote choice as well variations in levels of vote cohesion within constituencies (electoral districts), thus highlighting how variations in electoral dynamics encourage or limit co-ethnic preference and vote cohesion within an ethnic group. Data will be made available upon publication of my research.

2) Ethnic Demography Data

Besides electoral data on each constituency, my argument required that I test the impact of the ethnic landscape within each constituency on the preferences of voters. Specifically, this entails fine-grained data on politically salient groups which include caste and religious groups in each constituency. The government of India publishes very little constituency-level data, preferring instead to aggregate social and demographic data using a different set of administrative boundaries that have very little overlap with electoral boundaries. This has been a challenge for most researchers in this area. Most political parties, however, regularly use their party cadres to collect this data and voters tend to have rough estimates of the ethnic demography of their local constituency. I obtained this data from one political party member whose organization was involved with the collection of this data. I then compared this data with estimates from Center for the Study of Developing Societies in India as well as available data from the government of India on the size of the Scheduled Caste population in each constituency to confirm its reliability.