| Summary: | <p>This thesis consists of three distinct chapters examining how political power in India is shaped by social divisions along religion, gender, and caste.</p>
<p>In Chapter one, titled “Elections and protest in the wake of cow vigilantism,” I study vigilante attacks by Hindu “cow protection groups”—lynchings of mostly Muslim dairy farmers, cattle traders, carcass skinners, leather manufacturers, and beef eaters on allegations of slaughtering cows, animals considered holy by Hindus—which have proliferated in India since the Hindu-nationalist National Democratic Alliance (NDA) assumed power in 2014, led by Prime Minister Narendra Modi. I estimate the effect of such attacks on the performance of the NDA in 2019, when it was returned to power with an even larger mandate. I combine election results with data on cow vigilantism, demographics, and socio-economic indicators to construct a unique dataset for 542 parliamentary constituencies. In explaining the change in the NDA’s vote share from 2014 to 2019 in these constituencies, cow vigilante attacks are likely to be endogenous, and to address this problem, I derive a new instrument from the variation in cattle density across constituencies. After adjusting this effect for a potential underreporting of attacks, I find that for the average constituency, a casualty in a cow vigilante attack led to about a 9 percentage points greater increase in the vote share of the NDA from 2014 to 2019. This effect appears to operate primarily through a higher turnout, consistent with theories of polarisation, where communal violence increases the salience of religious identity and motivates people to turn out to vote for the party representing that identity.</p>
<p>In Chapter two, titled “Affirmative action-induced term limits,” I study an affirmative action program in the municipal councils of Delhi and Mumbai, where half of all wards are randomly selected before an election to be reserved for women, and randomly selected again before the next election five years later. A byproduct of these reservations is the imposition of effective term limits on many male incumbents (who may find that their ward is reserved for women for the subsequent electoral term). I assess how these effective term limits affect the performance of councillors, both at council meetings and at the provision of public goods such as schools, hospitals, trees, electricity, water, and sanitation. I combine election results with data on the provision of public goods, questions raised by councillors at meetings, and complaints lodged by citizens against councillors, collected by the nonprofit Praja Foundation to construct unique panel datasets for the 272 municipal wards in Delhi and the 227 wards in Mumbai from 2012 to 2022. Exploiting the natural experiment created by the randomness of reservations, I find that male councillors who find themselves to be term-limited (as well as female councillors who stand to lose their reservations) perform less well at councillor meetings, face more complaints from their constituents, and take longer to respond to those complaints, in the months after the subsequent reservation status of their wards is announced, as compared to their performance during the bulk of their term leading up to the announcement.</p>
<p>Chapter three, titled “Caste flight from public schools” analyses data from the second wave of the Indian Human Development Survey (2011-12) to examine parents’ preferences between public and private schools while deciding where to enrol their children. It is well-documented that private schools are disproportionately attended by upper-caste children, and the existing literature attributes this to the higher purchasing power of upper-caste parents and to the better quality of physical infrastructure in private schools. However, we find that even after controlling for parental income and the quality of infrastructure in schools, an upper-caste preference for private schools remains. This leads us to hypothesise than instead of merely proxying for class, caste may be producing independent social dynamics that feed into preferences across school types. Indeed, we find that in villages where at least one public school has segregated sources of drinking water for children belonging to different castes, upper-caste parents profess a significantly higher confidence in public schools. It is when segregation on the basis of caste is not achievable in public schools that upper-caste parents display a greater alacrity to send their children to private schools. Schools therefore appear to be at least in part mechanisms for upper-castes to mark themselves out from lower-castes, to signal a higher social status.</p>
|
|---|