This article explores the puzzle of the BJP Muslim supporter in Gujarat. Based on nearly two years of research in Ahmedabad city, Raheel Dhattiwala flags the crucial distinction between supporting a political party in public and voting for it. She examines the transformation in behaviour of the Sunni Muslim voter in Gujarat since 2009 and explores different motivations leading to their support of the BJP and Chief Minister Narendra Modi, despite many acknowledging the role of the party and its leaders in the massacres of Muslims in 2002.
The impressive performance of the Bharatiya Janata Party (BJP) in the recent assembly elections in four States has refreshed our memories of a question that lingered on in our minds after the Gujarat assembly elections in 2012: are Muslims voting for the BJP? Another way to ask the question is - Why would victims of brutal violence embrace their own perpetrators?
It was in Gujarat in 2009 that the BJP-Muslim question first took shape and holds more relevance now than ever before given the elevation of its Chief Minister Narendra Modi as the country’s potential prime minister. That was the year when the BJP, since they assumed power in the State in 1995, began a historic rapprochement with the Sunni Muslims of Gujarat providing them political representation in the State party: between 2009 and 2013 (with the exception of the assembly elections in 2012), the BJP nominated 297 Muslim candidates – many Sunnis – for various local body elections of which 142 (48 per cent) won.
Almost simultaneously, Sunni Muslims had seemed to mellow their antipathy towards the BJP. This was a remarkable development when seen in perspective of the brutal massacres of Muslims in 2002 during the BJP regime under Mr. Modi. Although until 2009, around 6-8 per cent of Gujarat Muslims had voted for the BJP in Gujarat, these voters belonged primarily to the mercantile Shia Muslim sect, Dawoodi Bohras in particular, who form less than 1 per cent of the Muslim population in Gujarat. They have traditionally lent support to the party in power, be it Congress or BJP, motivated by favourable state patronage. But for the Sunnis, a BJP supporter in the community was perceived as a defector and support for the BJP, if at all, was tacit.
The reasons behind the BJP’s volte-face in Gujarat were less evident at the time given that the Muslim vote at 9.1% continued to remain inconsequential for the Gujarat BJP. The most cogent explanation was Modi’s aspiration of an elevated position in the 2014 general elections, which could not be achieved without the support of Muslim electors in States outside Gujarat. Indeed, this makes much more sense now.
What continues to perplex is the behaviour of the Muslim voter. Whether one chooses to debate the role of the ruling BJP in supporting anti-Muslim violence in 2002, what is central is that a large section of Muslims continue to acknowledge the BJP and its affiliates in the Sangh Parivar as complicit in the riots.
From Perpetrator to Benefactor
Just how visible is this transformation in the Sunni Muslim voter? Not too long ago, in early 2009, a Sunni cleric from Ahmedabad, Mufti Shabbir Sidiqqui, had called the government appointment of a Dawoodi Bohra police officer as the head of Gujarat police a foil to display a secular ideology, “like appointing Muslims as Presidents of India in the past to keep the community happy”. Less than a year after the BJP’s purported reconciliation with Muslim electors, a tangible surge in BJP Muslim supporters became evident. Salim, a follower of Tabligh – an orthodox Deobandi revivalist movement -- told me a few weeks before the municipal corporation elections in Ahmedabad in October 2010: “There is no shame today in supporting Modi. BJP is Allah. Allah ke sivay aur kaun hai? (Who else do we have other than Allah?)” During elections in Ahmedabad there were many more like Salim. Dressed simultaneously in Islamic attire and saffron bandanas they professed their hope in the BJP, especially in Modi: “So what if he presided over the violence. Uski dahshat hai (he commands fear); he can get anything done,” said Salman, an auto rickshaw driver. Many others denied the presence of residential segregation of Hindus and Muslims in Ahmedabad, calling it a “matter of choice”. This was in contrast to my interviews as a journalist with Muslims in 2005 and 2006 who resented the discrimination of Muslims in residential ownership and tenancy of properties in upmarket areas of Ahmedabad.
The BJP’s unexpected denial of tickets to Muslim candidates in the 2012 assembly elections indeed came as a sign of ‘betrayal’ for a section of recent BJP Muslim supporters. But the discomfiture usually lasted less than two minutes for “the BJP is the only alternative despite the reality that the party could organise anti- Muslim riots again”.
Public support of the Sunni Muslim voter on the streets of Ahmedabad had increased substantively by 2012, matched by the party’s acceptance among Sunni Muslim clerics nationally. Recall the ouster of Maulana Vastanvi, the rector of India’s leading seminary in Deoband, in 2009 for his public support of Modi’s economic policies. In October this year, Mahmood Madani the general secretary of the Jamiat Ulema-e-Hind, instrumental in ousting Vastanvi, commended the Gujarat leadership for its inclusive treatment of Muslims in the State.
The Gujarat phenomenon is rare in conflict-ridden democracies. Sri Lankan Tamils, for example, have refused to bow down to the incumbent government during recent elections despite the latter’s claims of having provided inclusive economic development. Of course, one is tempted to draw an analogy with Sikhs of Delhi after the horrific massacres in 1984. Like the Muslims in Gujarat, the Sikhs did vote for the Congress in the 1990s despite having blamed the party earlier for the 1984 pogrom. But there are differences between the two cases. First, unlike the antipathy of Gujarat (Sunni) Muslims for the BJP, the Sikhs were never averse to voting for the Congress even before 1984. A majority of Sunni Muslims of Gujarat had always perceived the BJP as an ideologically non-pluralist party, particularly antagonistic towards Muslims – the violence in 2002 reinforced this perception for many. Second, no explicit apology has been tendered by the Gujarat BJP to Muslims unlike the Congress to the Sikhs – Modi’s “anguish” at the violence, expressed on his recent blog, does not replace an apology to the Muslims who bore the brunt of the massacres. So the question arises: can the BJP’s political strategy to provide representation to Muslims explain why a Sunni Muslim voter would support a fellow Muslim – or, in case of the 2012 assembly elections, a Hindu -- favouring a Hindu nationalist party that has refused to apologise for the 2002 violence?
Exploring Motivations of the BJP Muslim Party Member and Party Supporter
Two popular explanations for the BJP-Muslim question currently exist. First, the BJP’s relatively equitable economic development in Gujarat in the last decade. Second, its riot-free governance. As I show in an earlier article in The Hindu, the attacks on Muslims after the death of Hindus on the Sabarmati Express were not spontaneous but organised. It follows that within the current political circumstances, allowing a trigger to develop into a full-blown riot makes little electoral sense in the State. What is worth discussing here is the explanation of equitable economic development. For this overarching term to make analytical sense, it needs to be broken down to the level of the individual voter. Whereas economic benefits mattered for the individual respondents I interviewed, notably, there were differences in the motivation to support the BJP for those Muslims who had joined the party as members, from those who were supporters/campaigners for the party.
Suraiyya, a resident of the dismal Bombay Hotel slum in eastern Ahmedabad, was one such party member. She had moved to the safety of this neighbourhood in 2002 following more than one attack in her old neighbourhood of Bapunagar. In 2012, her reasons to fear for her life were no longer riots but the civic conditions of her infamous neighbourhood, situated contiguous to the city’s official municipal sewage farm. She had recently joined the BJP as a party member. “Do we have a choice? We don’t want to keep drinking yellow water and die of dengue…”
It was interesting that Suraiyya chose to join the party rather than simply vote in its favour, for, surely, voting by itself would have given her access to the facilities she desired. Suraiyya was candid: “I want to earn money by helping people.” Suraiyya fit the typical ‘intermediary’ or ‘broker’ – local individuals who facilitate access of patronage-based State resources between citizens and the State. In turn she earned not only monetary benefits from the residents but also a positive social identity within the community. This patron-client form of resource access can dictate political preferences, especially for the extremely poor whose daily survival is a never-ending ordeal and for whom direct access to State resources is nearly impossible.
Suraiyya’s sentiments were reflected in the words of other Muslim BJP party workers I met. Notably, although they praised the BJP for its economic policies, none claimed to have directly benefited from them. Aslam, a local BJP leader, spoke of “indirect” benefits – housing developed to rehabilitate all slum dwellers displaced by a city development project, in keeping with a Gujarat High Court order, had also benefited Muslims. These recent members of the BJP were aware that their aspirations to becoming a successful intermediary were possible only within the ambit of an incumbent party with the potential to remain in power for a long period of time.
“We Muslims first believe in nation, that’s what Islam also says… to get rid of our anti-national image we have to be with the BJP. What has the Congress done except breed goons and use Muslims to sell illicit liquor?”
Do these testimonies suffice to explain why the Muslim in Gujarat is supporting the BJP? Perhaps so, although they are not sufficient to show that those who support the BJP – be it as party members or public supporters – vote for the party as well. Anonymous referendum implies there is no possible way to determine who voted for whom. Meticulous survey methods or analyses of booth-level voting data are the best alternatives even though one can only draw tentative inferences from these methods.
Next I will take a closer look at the difficulties in drawing conclusions of voting behaviour from constituency-level data, how inferences drawn from a preliminary booth analysis suggests the possibility of higher Muslim BJP support in public than on the secret ballot, and the implications of this paradoxical behaviour within a democratic system. I have already explored the various motivations of Sunni Muslims in Ahmedabad for supporting/campaigning for the BJP or joining it as party members. Whereas access to State-based provisions was a primary reason for supporting and, indeed, joining the party, a keen desire for societal acceptance also motivated the supporters to voice their approval of the party in power. This raises the question whether voting for the BJP was at all essential if supporters, especially party members, were already being rewarded beforehand – monetarily or by elevation of social status.
In this section, I examine the difficulty in disentangling public liking for a party/candidate from voting for the same party/candidate. A preliminary analysis of polling booth data of 12 booths in Ahmedabad city highlights the paradox of Muslim support for the BJP. Findings from the booth analysis suggest that the proportion of Muslims voting for the BJP is likely to be much less than their apparent support in public for the party or its leaders. The figure of 8% to 9% is not much different from BJP Muslim voting figures cited by observers in previous elections. I explain this in detail below.
Public versus Electoral Support
When we use the term “support”, it could mean either proclaiming in public a liking for a political party (or its leaders) and/or voting for the party. It is possible that a Muslim who shows public liking for a party also votes for it but that may not always be true. Secrecy of the ballot ensures one does not know who voted whom. Indeed, anonymous referendum also means that testimonies collected during ethnographic fieldwork may often be insufficient to disentangle public support from electoral support. As Shazia, a Congress campaigner in Ahmedabad, told me discreetly in 2012: “I am representing the Congress but voted for the independent candidate (names him)… how would anyone know?”
Following the BJP’s victory in the 2010 municipal elections across Gujarat, Modi declared: “Over 30% Muslims have voted for us”. Then in 2012, the BJP claimed to have won in 24 constituencies where Muslim electorate was 15% or higher. How certain can one be of these figures? A simple numerical illustration will help.
Suppose that a polling booth has 100 voters. Of these, 60 are Hindus and 40 Muslims. Suppose further that the local BJP candidate has attained 50 votes in this booth. Secret ballot ensures we cannot identify who all among the voters voted for the BJP candidate.
Hypothetically therefore, two scenarios are possible:
1. All of the 40 Muslims voted for the BJP candidate. Even so, at least (50-40) i.e. 10 Hindus (16.66%) voted for the BJP candidate.
2. All 50 votes to the BJP candidate were cast by Hindus. This means at the most 50 out of the 60 Hindus (83.33%) voted for the BJP candidate.
What follows is that, hypothetically, it is possible that every single Muslim voted for the BJP candidate. But it is equally possible that none of the Muslims voted for him. The same way, a number of arrangements are possible, including the equiprobability of 25 Hindus and 25 Muslims voting for the BJP. Point is, one cannot be sure of the proportion. Indeed, using aggregate-level (constituency) data to make inferences about the behaviour of individuals can lead to spurious conclusions or an ‘ecological fallacy’. Even from booth-level analysis we can only obtain upper and lower bounds on voting patterns. In the above-illustrated case, we can infer that Muslim support for the BJP candidate was between 0 to 100% and Hindu support for the candidate was between 16.6% to 83.3%.
However, the more homogeneous the polling booth in terms of religious composition, the tighter the bounds. If the above polling booth had 90 Muslims and 10 Hindus, we can hypothetically draw the inference – using the same logic as above – that 44% to 55% Muslims voted for the BJP candidate.
Of course, the first and foremost challenge is to identify the religion of electors in each polling booth. These figures are not publicly released by the Indian government and one needs to come up with resourceful methods of estimation – usually from electoral rolls and name recognition software. All these factors show how difficult it is to estimate voting behaviour from constituency-level data. An analysis of a larger sample of polling booth data is currently underway as part of my research project at The Hindu Centre.
Now the crucial question arises: why this contradiction in behaviour? Plausibly, the explanation is linked to a rather everyday experience in our lives – anxiety (or dissonance) over choices we make or the values we cherish. For example, I know caste discrimination is wrong but I still hesitate to allow my daughter marrying into a lower caste. These are contrasting thoughts, which are likely to cause psychological discomfort leading to my attempts to reduce the discomfort either by concealing my true beliefs or acquiring a new piece of knowledge that might justify my hesitation to allow the inter-caste marriage. For instance, if a relative convinces me of the immense social ostracism I’d be preventing for my family by forbidding such a marriage, I would willingly adjust my original beliefs. Similarly, it is possible that for a Muslim, who believed that the BJP is an anti-Muslim party, the new knowledge of the BJP’s inclusive measure to represent Muslims in the party could be inconsistent with the awareness that the BJP is anti-Muslim. If this Muslim individual chooses to support the BJP for potential economic gains expressing his original private belief might now invite social disapproval from the majority Hindus for opposing an inclusive government and from his own community for opposing a government that seeks to make amends and provide benefits. The inability to express oneself truthfully alters our private preferences to avoid social disapproval and, subsequently, concealment of misgivings. Thus a Muslim would overtly express his nationalism or mute any contrary views (e.g. denial of segregation). This is more likely for religious Muslims – especially those with visible religious signs, which increase their vulnerability to being identified as orthodox Muslims. A rise in Muslims publicly supporting the BJP would increase the overall pressure among other Muslims to find social approval, thereby leading them to support the BJP. Therefore, it is possible that Muslims residing in intermixed neighbourhoods (than in a Muslim ghetto) are more likely to support the BJP in public and less electorally because of the high social pressure to conform to the public discourse. The concept of dissonance and its reduction have been discussed at length in the works of social psychologists like Leon Festinger and political scientists like Timur Kuran.
Anthropologist James Scott’s work on the language of power relationships is also a plausible explanation to the contradictory behaviour of Muslims. Scott speaks of the ‘hidden’ transcript’ –the language of the dominant and the subjugated which occurs in private. Contrary to the idea of hegemony the hidden transcript suggests a parallel nonconformist subculture, albeit outside the gaze of political power, that allows the subjugated to fulfil their own personal goals. In case of Muslims, the vocal support for the BJP is analogous to the public transcript and anonymous secret ballot to the hidden transcript.
My ongoing work at The Hindu Centre is to examine a larger sample of homogeneous polling booths with variation on class, exposure to violence and proportion of Muslims to explain this puzzling phenomenon.
Names of respondents are anonymised