Candidate Caste Effects in Uttar Pradesh Elections

Published date01 December 2015
Date01 December 2015
Subject MatterArticles
Candidate Caste Effects
in Uttar Pradesh Elections
Shikhar Singh1
This article assesses the impact of a candidate’s caste on the probability of voting for a party.
Candidate caste effects may be pronounced in multiethnic societies and patronage democracies. This
is demonstrated for four political parties in Uttar Pradesh (India) across three state elections. Using
data from fieldwork, a logistic regression is employed to test two hypotheses—voters are more likely
to vote for a party if it fields a candidate from their caste; and less likely to vote for that party if
other parties field co-caste candidates from that constituency. Results show that hypothesized effects
are statistically significant across parties and elections. Citing corroborative evidence, it is suggested
that caste parties employ candidate-centric strategies in some constituencies to widen their social base.
Indian elections, candidate caste effects, caste, Uttar Pradesh
In India, social relations and resource allocation have historically been shaped by caste, albeit in
different ways over time. In a closed village economy, caste operated as a hierarchy and structure of
domination. With the collapse of insulated local economies—brought about by abolition of zamindari
rights, land redistribution and economic integration—there was a democratization of the arena in which
social relations and resource allocation were negotiated. Armed with legal equality and freed from
the power structure of village communities, many backward castes formed political associations to
promote their interests (Alam, 1999b; Kothari, 1970; Michelutti, 2004). Though in competition, they
often formed strategic alliances with one another to secure greater benefits from the state (Alam, 1999a).
This way, castes developed into interest groups that were granted formal equality and basic freedoms
in a modern political system.2 After independence, the Indian state formally recognized group
interests and apportioned resources on that principle (Weiner, 2001). It ‘reserved’ seats in legislatures
(giving access to political resources), places at university (opportunities) and jobs in the public sector
1 Balliol College, Oxford OX1 3BJ, United Kingdom.
2 This is often referred to as the ‘horizontalization’ of castes (Gupta, 2005).
Studies in Indian Politics
3(2) 179–197
© 2015 Lokniti, Centre for the
Study of Developing Societies
SAGE Publications
DOI: 10.1177/2321023015601741
Corresponding author:
Shikhar Singh, Department of Political Science, Yale University, PO Box 208301, New Haven, CT 06520-8301,
180 Studies in Indian Politics 3(2)
(economic resources), initially for Dalits, and after the Mandal Commission, for a large number of
backward castes. Owing to this historical context, caste politics emerged as a salient feature of the
Indian polity.3
Both parties and candidates use caste to mobilize support and consequently, it affects vote choice in
complex ways. When a party system is based on caste cleavages, caste determines party preference.
Voters belonging to a caste are more likely than others to support their caste party. Such parties are
typically outgrowths of caste associations and reflect an evolution in their organizational structures,
political methods and objectives (Pai, 2002; Palmer, 1975). Instead of exerting pressure on the state and
petitioning it, castes participate in elections to capture public offices. For this purpose, they form parties
that exclusively pursue the interests of that particular group (or set of groups) and devote their entire
political capital to this end.4 Of course, there are multi-caste parties that do not claim to exclusively
represent any one group but champion the cause of many groups. For such parties, caste effects are
noticeably weaker. In Uttar Pradesh, India’s most populous and demographically diverse state, two
regional caste parties—Bahujan Samaj Party (BSP) and Samajwadi Party (SP)—and the Hindu nationalist
Bharatiya Janata Party (BJP) have dominated the political scene. BSP mobilizes Dalits (particularly the
Jatav caste), SP appeals to Yadavs and Muslims, while BJP receives support from upper castes
(particularly Brahmins and Rajputs) (Vora, 2004). The fourth political player is the ‘catch-all’ Congress
Party. Between 1991 and 2002, caste-based differences in party preference were more pronounced than
before.5 Parties made sectional appeals to voters, the electorate was polarized along caste lines and this
produced a fragmented polity, one in which no political party secured an overall majority at the elections.6
Caste voting of this kind has been extensively studied and is not discussed here in greater detail.
Caste is also used by candidates to mobilize local support and this influences vote choice. This
dimension of caste voting consistently escapes academic attention and is the main subject of this article.
Candidate caste affects vote choice in a positive and negative way. Voters are more likely to support a
party if it fields a candidate of their caste and less likely if other parties field co-caste candidates. Even
though these are commonly held assumptions about caste voting, they remain empirically unverified.
This is due to inaccessibility of data and excessive focus on macro-politics. Neither the Election
Commission of India nor political parties maintain any official record of candidate caste. Since this data
is only informally available, accessibility issues have deterred fieldwork and any empirical evaluation of
hypothesized effects. Moreover, academics downplay the significance of local politics and concentrate
on state-level trends. This is surprising given the local nature of caste identity.7 Combining original
3 Of the several overlapping group identities that are politically salient in Uttar Pradesh, caste and religion are the strongest deter-
minants of vote choice. When religious polarization is weak, caste-based differences shape political cleavages. This may result in
cross-community alliances—rival Hindu castes may ally with other religious groups to dominate local politics. Because religious
groups may be part of caste-based mobilization strategies, any study of caste politics cannot exclude religion. For this reason,
Muslims are part of our analysis on caste politics and any reference to ‘caste’ is typically an abbreviation for ‘caste–community’.
4 Kanchan Chandra calls them ‘ethnic parties’—any party that ‘overtly represents itself as a champion of the cause of one particular
ethnic category or set of categories to the exclusion of others, and that makes such a representation central to its strategy of mobiliz-
ing voters’ (Chandra, 2004, p. 3).
5 Pradeep Chhibber found caste to be a statistically significant determinant of party preference in the assembly elections of 1993
and concluded that the ‘party system [was] rooted in caste cleavages’ (Chhibber, 2004, p. 157). At the assembly elections of 2002,
84.7 per cent Dalits voted for Bahujan Samaj Party (BSP); 62.5 per cent Muslims and 75.2 per cent Yadavs voted for Samajwadi
Party (SP); while 64.3 per cent upper castes (Brahmins and Rajputs) voted for Bharatiya Janata Party (BJP) (Verma, 2002).
6 In 2002, vote share of major parties was as follows: SP, 25.4 per cent; BSP, 23.2 per cent; BJP, 20.7 per cent; and Congress, 8.9
per cent. The seats tally in the legislative assembly (403) was: SP, 143; BSP, 98; BJP, 88; and Congress, 25 (Shastri, Suri & Yadav,
7 Local caste rivalries and the geographical distribution of castes across regions are detailed in Schwartzberg (1965).

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT