Uttar Pradesh CM Yogi Adityanath said that the state will witness a “80% versus 20%” election and that the BJP will retain power in the state. The opposition has alleged that Yogi’s comments seek to create a polarisation between the majority Hindu and minority Muslim population in the state. This comment notwithstanding, it is a fact that the Bharatiya Janata Party (BJP) is actively seeking a consolidation of the (otherwise caste-splintered) Hindu vote. Existence of Muslim consolidation on the other side is often used to justify this appeal. Are polemics about a Muslim consolidation factually correct? Looking at data on past electoral representation of Muslims in Uttar Pradesh may answer this question.
Muslim representation in Uttar Pradesh has been volatile
Muslims’ representation in Uttar Pradesh has historically fluctuated. The rise of socialist parties in the 1970s and 1980s and the decline of the Congress saw the first post-Independence rise in Muslims’ representation in the Vidhan Sabha, from 6.6% in 1967 to 12% in 1985. The first rise of the BJP in the state in the late 1980s brought this percentage down to 5.5% in 1991. As the chart shows, the overall participation of Muslims in elections as candidates also decreased over the same period, although not to the same extent, due to fewer nominations by major parties (other than the BJP). The second phase of growth in representation starts after 1991 and culminates in 2012, when Muslim candidates won 17% of the assembly seats, achieving near-demographic proportion for the first time (the carving of Uttarakhand in 2000 also contributed to raise the percentage of Muslim’s representation). The BJP’s emphatic victory in 2017 reversed this trend back to the 1991 level: 23 Muslims were elected, against 68 in the previous polls.
See Chart 1: Share of Muslim MLAs and candidates in Uttar Pradesh (pre-2002 data includes Uttarakhand)
Did ‘provincialisation’ of UP politics increase Muslim representation?
The current dominance of the BJP in Uttar Pradesh politics has significantly diminished the political fortunes of the two regional heavyweights in the state’s politics: the Samajwadi Party (SP) and the Bahujan Samaj Party (BSP).
Between 1996 and 216, these two parties had an average seat share of 63% in the state’s assembly (down to 16.4% in 2017). When read with the fact that this was also the period when the share of Muslim candidates and MLAs increased consistently, the question to be asked is whether the restoration of a national party’s dominance – Muslim representation was low during the period of Congress dominance as well – has undone something that came with the rise of regional players. A party-wise break up of Muslim MLAs and candidates in Uttar Pradesh suggests that the answer to this question is a yes.
A breakdown of nomination and representation data shows that the two regional parties, the SP and the BSP, provide the most space to Muslim candidates and therefore representatives. Since the BJP hardly nominates any Muslims (only eight candidates since 1991 and none in 2017), and since the Muslim candidates nominated by the Congress tend to lose, the bulk of Muslim MLAs in Uttar Pradesh belong to the two regional parties. It would therefore be tempting to see Muslims’ representation merely as a reflection of regional parties’ performance: increasing when they perform, and decreasing when the BJP does well. While this is by and large true, the data, however, reveals that the relationship between nominations and representation is not as straightforward as it may seem.
See Chart 2: Party-wise Muslim MLAs
See Chart 3: Party-wise Muslim candidates
The rise of SP and BSP in Uttar Pradesh split the Muslim vote rather than create a vote bank
The puzzle is the relationship between nominations and strike rates.
The BSP got more Muslims elected in 2007 even though it reduced the number of tickets it gave to Muslim candidates (62 against 83 in 2002). Even though the number of SP Muslim MLAs grew in 2012 (from 21 to 43), half of its Muslim candidates still lost their races, including in seats with large Muslim populations.
One explanation for the fuzzy connection between nominations and representation is that multiple Muslim candidates in seats with significant share of Muslim voters can actually harm their representation, through split voting. My definition of split voting is when the cumulative vote share of losing Muslim candidates among the first four exceeds the non-Muslim winner’s vote share. This is a crude measure that is merely indicative of the possibility of split voting, since other factors, such as overall party support across religious groups, also matter.
By this measure, since 1991, Muslims “lose” on average 18 seats in every election, with significant fluctuations (there were eight and seven cases in 1996 and 2007, against 26 and 27 cases in 2012 and 2017). As expected, the BJP is the main beneficiary of this phenomenon. Since 1991, it has won 104 of the 126 seats that saw split voting between Muslim candidates. Twenty-seven of these came to it in 2017. Split voting also does not necessarily occur across the same constituencies over time, which indicates that this is not just a question of demography, but of cohesiveness of votes at the constituency level.
See Chart 4: Party-wise winners in ACs where a Muslim candidate lost due to split voting
In fact, Muslims’ representation started to increase in Uttar Pradesh in the early 1990s before regional parties started nominating more of them. One could say that regional parties followed the trend and amplified it, rather than created it.
What the discussion above shows is that Muslim voters do not vote blindly for regional parties or simply for the most probable winner in their constituency. Their electoral behaviour varies between elections. Muslim voters, like any other, evaluate their choices ahead of voting, weighing what both party and candidate have to offer them. Crucially, their voting decisions vary from one seat to another, contradicting the claim that Muslims in Uttar Pradesh vote en bloc.
This also means, therefore, that parties cannot mobilise Muslim voters only by distributing tickets to members of their communities and staying silent on issues that concern them. Instead of taking Muslim voters for granted, parties should give cogent reasons to vote for them – beyond an opposition to BJP.
(The author is Assistant Professor of Political Science and Co-Director of the Trivedi Centre for Political Data)