Electing Women in Ethnically Divided Societies: The Case of Bihar, India

• In Event: Explaining Women's Inclusion and Exclusion

Abstract

In this article, we explore the interaction of gender, caste, and campaign strategy in India’s state-level elections, focusing on the state of Bihar. We are concerned specifically with elections for the position of MLA, member of the state legislative assembly, the Vidhan Sabha. Our goal is to identify the key factors that impact women’s political success. This information can hopefully be used to promote significant and inclusive gains in local socio-economic growth and improved gender parity.

We explore six distinct but interrelated questions. First, we draw on the literature in American politics to investigate whether women are likely to be more successful with stereotypically “feminine” electoral appeals to public goods rather than “masculine” appeals to security. Second, we expand on that question by asking whether voters who are more exposed to caste violence will be more interested in candidates making security appeals and less likely to support a woman. We anticipate that women will be less able to make security appeals and will be at a disadvantage among voters, and by extension among districts, with a history of caste and communal violence. In other words, we expect that reduced political opportunities for women can be added to the many doleful consequences of civil violence.

Third, we test whether voters who have lived under a female local leader will be more amenable to voting for a woman at the higher state level, where no quotas exist. These local leaders, the elected heads of village councils, are known variously as a pradhan, mukhia, or sarpanch. Bihari state law requires that women be selected as pradhans in 50% of villages, but it provides no gender reservation for state-level elections. We therefore explore the very important question of quota “spillover effects.” Do gender quotas at the local level, where they are easiest to implement, ease the path of women seeking electoral office at higher tiers? And, relatedly, are voters living in MLA constituencies already represented by a woman more likely to vote for a woman in future elections?

For our fourth question, we explore the dynamics of intersectionality in India, where gender and caste identities often overlap and may reinforce each other in unexpected ways. While these issues have received the attention of sociologists and anthropologists, we are among the first to examine systematically their implications for politics India’s states. Our expectation is that the burdens of lower caste status will fall disproportionally on women candidates.

Fifth, we explore the interactive effect of party on support for women candidates. We anticipate that identifying a woman candidate with a particular political party will dampen the effect of gender on vote choice. Party loyalty will take over and trump, we believe, the identity of the candidate. As a sub-question, we investigate the impact of affiliation with a right-wing communal party (the BJP) with affiliation with a more leftist caste-based party (the RJD).
Finally, we investigate how the success of women candidates is mediated by local and state party activists. These lower-tier political workers can play an outsized role in determining the fate of potential elected officials, and we expect that they will tend to dissuade women from running in constituencies that have experienced a high level of caste or communal violence. We also anticipate that they will steer women candidates towards more “feminine” campaign appeals.

To test our arguments, we accumulate original, primary data from a large survey experiment of voters in the Indian state of Bihar. The survey experiments will be conducted in sixteen state legislative (Vidhan Sabha) constituencies through a random sample of 2500 respondents. Our surveys will present respondents with hypothetical candidates distinguished by caste, gender, campaign appeal, and political party in order to isolate the additive and interactive effects of each characteristic on voter preferences.

We complement these survey experiments with interviews of 120 political party functionaries, NGO members, and current elected representatives in Bihar. These interviews will shed light on the mechanisms and context behind the causal inferences established by the survey experiments. They will also help us understand better the “supply” side of the political marketplace, allowing us to address not only the behavior of voters but also the incentives affecting those who select, support, and fund potential candidates.

Authors

• Sayan Banerjee, University of Virginia

• Charles R. Hankla, Georgia State University

• Anjali Thomas, Georgia Institute of Technology