In a recent press conference at which Mayawati explained the causes of the Bahujan Samaj Party’s (BSP) defeat in the Uttar Pradesh elections, she blamed her supporters for deserting her and her candidates. She also blamed the Samajwadi Party (SP) for attracting the BSP’s Muslim voters, which, according to her, led to a Hindu consolidation behind the Bharatiya Janata Party (BJP).
This is not the first time that Mayawati has blamed voters for her electoral setbacks. In 2012, she criticised Muslim voters for supposedly causing her defeat by supporting the SP. Beyond pointing out the obvious lack of grace in defeat and keeping in mind that in a democracy, voters are free to vote for whomsoever they choose, it is worth wondering why so many voters deserted the BSP in the 2022 UP election.
There is a larger message for UP electoral politics that can be learnt from the decline of the BSP in UP. This is the decline of a particular way of doing politics that the BSP has been practising so far: of consolidating a core electoral base, and relying on fragmentation and vote bank transferability to win seats. The BJP, which emulated the BSP and SP’s winning formulas of 2007 and 2012, has come to understand that a campaign simply based on the mobilisation of narrow segments of the electorate is doomed to fail.
BSP’s decline
The decline of the BSP in UP in 2022 is nothing short of a collapse (Chart 1). In 2007, it won a small majority in the assembly (206 out of 403 seats) with 30.4% of the votes. At the time, the fragmentation of the electorate and of the party system meant that winning thresholds were generally low at the constituency level (on average 35% of the vote from 2002 to 2012).
In 2012, the BSP captured 80 seats with 25.9% of the vote. In 2017, it garnered only 19 seats with a 22.2% vote share. After the 2017 elections, most commentators thought that the party had reached its floor. In 2022, however, the BSP has suffered from a massive vote loss, down to 12.9%. The party has now been reduced to a single seat, in Rasra, Ballia district, which elected Umashankar Singh, a first-time contestant who has a business background.
A common explanation for the BSP’s collapse is that since the SP, as the main opposition, was expected to do well in a polarised contest against the BJP, strategic voters and BJP opponents would flock to the SP in the hope of preventing the governing party from winning a second consecutive term.
To identify the root of the BSP’s electoral problems, one needs to examine the BSP’s model of social engineering…and see how the model… has turned against it..
This argument is supported empirically by the fact that most BSP voters in eastern and western Uttar Pradesh seem to have shifted more towards the SP rather than to the BJP. Political scientist Neelanjan Sircar has examined exit poll data and suggested that the decision of non-Jatav Dalit voters not to vote for the BSP stemmed more from individual strategic voting rather than coordinated decisions among members of various castes across constituencies.
However, this does not entirely explain why so many voters seemed to have decided that the SP was the most credible contender against the BJP and deserted the BSP accordingly. It also does not account for why the BSP has been on a long trajectory of decline, both in state and in general elections. To identify the root of the party’s electoral problems, one needs to examine the BSP’s model of social engineering, as it used to be called in the mid-2000s. The model upon which the BSP built its success, a mix of local elite co-optation and vote bank transferability, turned against it when the circumstances of UP politics changed.
BSP’s model of outsourcing votes
Ever since its inception in 1984, the BSP has devised a strategy of forging local and temporary caste-based alliances as a means of gaining political ground. As a Dalit party facing substantial prejudice from upper and intermediate castes, it could not reasonably expect to get many Dalits elected in general seats. This is particularly true given that the Dalit proportion of the population in UP is only around 20%.
The data bears this out. According to data analysed by the Trivedi Centre for Political Data, only 40 Scheduled Caste (SC) candidates have been elected in UP on general seats since 1989, and only eight of them on a BSP ticket (including Mayawati once, in the Srinagar constituency in what is now Uttarakhand, in 1996).
The BSP therefore had to find candidates outside its core electoral base and outside its own organisation, which was essentially composed of Dalit cadre. The strategy devised by the then-leader Kanshi Ram was to build local alliances with relevant dominant castes, and combine the strength of the BSP’s appeal among Dalits with the vote-getting capacity of candidates recruited externally.
Local party branches would send to the BSP headquarters information on electorally relevant groups, based on demographics and their socio-economic hold over constituencies, and proceed to identify suitable candidates within these identified groups. The criteria included money, muscle and, of course, caste. Over time, as the BSP’s prospects in state politics improved, suitable candidates meant individuals who had the resources, the capacity, and the incentives to invest in a political career. Most of these candidates came from a business background and used their wealth to enter politics by acquiring BSP tickets which, more than with other parties, happened to be for sale.
The BSP worked on the assumption that nearly any individual with some combination of the right caste identity, money and muscle “would do” as a BSP candidate.
This became the foundation of the BSP’s 'caste-arithmetic' strategy. It did not rely on broad horizontal mobilisation of various castes across the territory, but rather counted on the transferability of its Dalit vote bank to candidates identified and recruited according to local circumstances and by following specific criteria.
Since these circumstances would vary from one seat to another, it enabled the BSP to claim that it was promoting diversity. But considering the way the strategy worked, it also meant that the individuals thus recruited belonged to the elite segments of locally dominant groups. Thus, the BSP’s quest for diversity did not proceed from some principled position on equity, but from the externalisation of the business of winning seats to individual political entrepreneurs who chose to invest in politics by acquiring a BSP ticket.
Inclusion thus became a by-product of an elitist, pragmatic method of candidate recruitment that created the kind of predatory political class that made UP infamous through the 2000s. Candidate recruitment was precise in terms of selection — or winnability — criteria, but loose when it came to selecting the actual individual. The BSP worked on the assumption that nearly any individual with some combination of the right caste identity, money and muscle 'would do' as a BSP candidate.
As a result, candidate selection in the BSP became quite open to those willing to pay for a ticket. In the seats where several candidates were willing to contest, tickets were at times auctioned among these aspirants. Quite often, confirmed nominations got cancelled on the eve of the campaign, and were transferred to a higher bidder. This created a gateway for gangsters, shady businessmen, and various kinds of political mercenaries to get into local positions of power.
Local BSP branches remained quite cohesive at the exclusive service of the party, and thus were unlikely to be used to prop up the individual ambitions of BSP candidates...
This recruitment method provided two other advantages to the BSP. One, it could concentrate on retaining its core support base among the Jatav community by giving them control of the local party organisation. Local BSP branches remained quite cohesive and at the exclusive service of the party. They were unlikely to be used to prop up the individual ambitions of BSP candidates at the cost of the party. It helped the BSP build a dependable base of cadre devoted to party work, rather than to the advancement of their own political careers.
The second advantage was that the cohesiveness of the BSP organisation supported the goals of the party headquarters. As the BSP’s local branches were detached from the candidates and the elected representatives, this conferred autonomy on them and allowed a certain degree of reliability in terms of the information that the base was sending to the party high command. When I was doing fieldwork collecting data on candidates during the 2007, 2009 and 2012 elections, most workers and cadres from other parties admitted that the BSP’s organization had by far the most ‘scientific’, disciplined, and ultimately reliable system of local data collection on candidates and constituencies. When candidates belong to the party organisation, they have incentives to inflate numbers in their favour, deliberately overestimating the size of their own community, or minimising the size of their rival’s social base.
A model imitated by other parties
In 2012, the SP learned from the BSP’s strategy and sought to emulate it by leading its own campaign of 'inclusion'. By then, the SP had already started diluting the presence of core Other Backward Class (OBC) communities among its candidates, distributing more tickets to upper caste candidates and to a lesser extent non-Yadav OBCs. It adopted a similar localised diversification strategy by distributing tickets across a broader array of castes. Just like the BSP, it did so in the name of inclusion and it also selected its candidates on elite criteria of wealth and the capacity to capture votes from locally influential groups. The SP had a certain degree of success, as it won an outright majority in 2012.
The production of an elite class of politicians by both the BSP and SP helped the BJP to frame the terms of electoral competition in its favour five years later, in 2017. The BJP denounced the elitism of its opponents, called them (with reason) the new establishment of the state and criticised them for treating their core support bases and local bosses preferentially.
At the same time, the BJP also learned from its adversaries and integrated some aspects of their strategy into its own. In the name of inclusion, it started to include a greater diversity of caste among its candidates, while retaining a disproportionate number of upper caste candidates. The BJP candidates’ profile also changed. It included fewer candidates with preexisting ties with the Sangh Parivar and more candidates grounded in business. In 2017, it fielded 76 turncoat candidates, 37 coming from the BSP alone (30 of whom won). Winnability on conventional criteria rather than ideological commitments became the dominant factor in political recruitment.
The production of an elite class of politicians by both the BSP and SP helped the BJP [in 2017] to frame the terms of electoral competition in its favour ....
The BJP was more cautious in recruiting among backward groups that had been historically underrepresented. It expressly shunned Yadav OBCs and Jatav Dalits, though it did nominate a few candidates from both groups in constituencies where demography made it difficult to do otherwise. But even among the more marginalised groups from which it selects its candidates, the BJP still tends to recruit wealthy powerful individuals, who can sustain some of the competitive pressure of UP politics on their own terms. Data from the Association for Democratic Reforms (ADR) show (Chart 2) that the average value of assets of BJP candidates has gone from Rs. 61 lakhs, in 2007, to Rs. 7.5 crore in 2022. Only SP candidates come close to that level of average wealth, well above BSP and Congress candidates.
There are of course variations among the candidates within each party. But in 2022, 86% of all BJP candidates were crorepatis, against 85% for the SP, 74% for the BSP, and 47% for all Congress candidates.
The end of the BSP model
The BSP model of candidate recruitment worked as long as 30% of the votes was all that a party needed to get a majority of seats in the state assembly. The presence of several major parties mobilising specific segments of the electorate and of two national parties at the time in decline, contributed to maintaining low winning thresholds (on average 35% from 2002 to 2012).
In such a context, a strategy that relied on building minimal winning social coalitions paid off, as candidates could win on the basis of small numbers that had been gathered through the mobilisation of narrow segments of the electorate. That strategy was optimal too since it freed parties from the need to go after the large number of undecided voters, who at the time were floating in all directions. The fact that the BSP had a larger core support base than any other party — and one that was more cohesive and loyal — gave it a competitive advantage that accounted largely for its electoral successes at that time.
One issue with the auctioning or selling of tickets is that it creates very little sense of loyalty from the MLAs or MPs towards the party, since the relationship is basically transactional.
But once the BJP reemerged as a dominant force in UP, it raised those winning thresholds considerably — 43% in 2017 and 46% in 2022 — making the minimal winning social coalitions now less effective.
The BSP strategy had started showing its limits earlier. In 2012, cracks had already started appearing at the base of the edifice. Inducting non-Dalit outsiders into the party and the maintenance of a firewall between the party’s local organisation and the party’s elected representatives meant that many BSP MLAs felt little obligation to pursue an agenda other than their own. One issue with the auctioning or selling of tickets is that it creates very little sense of loyalty from the MLAs or MPs towards the party, since the relationship is transactional. There was then growing discontent within the party’s base, which may account for the slow erosion of support over time among Jatav Dalits, as shown by CSDS/Lokniti data. Some BSP Jatav cadres were even recruited by other parties, as candidates in reserved seats.
The new model
The collapse of the BSP model in UP does not make caste irrelevant as a political variable. It remains central to ticket allocation strategies and it remains central to the discourse on inclusion of political parties, including the BJP. Exit poll data on the 2022 elections from CSDS and from Axis MyIndia indicate that both the SP and BSP succeeded in mobilising their respective core support bases, but failed to mobilise much beyond these bases (or not at all, in the case of BSP). This explains why the SP got stuck at 37% in the seats it contested (32% for all seats)—the best score it could hope to achieve on that basis—and why the BSP’s vote share collapsed. The BSP lost support from nearly every segment of the electorate, especially from those outside its Jatav Dalit base.
We are now in a different electoral model in which leadership appeal matters more than the candidates’ ability to mobilise voters on the basis of their own narrow identities.
The electoral strategies of the 2000s have greatly contributed to producing a political class that is elitist in class terms, grounded in local business networks, self-serving, and quite incapable of satisfying the limitless needs of their constituents. This has led to the extraordinary turnover of MLAs observed in the UP assembly in which most fly-by-night MLAs are either discarded by voters after a single election, or are de-selected by their own parties in anticipation of anti-incumbent feelings .
More importantly, we are now in a different electoral model in which leadership appeal matters more than the candidates’ ability to mobilise voters on the basis of their own narrow identities. It is also a model in which voters seem to respond to what parties in power provide them. This might have created an opportunity for the BJP to rely less on money and muscle for the selection of its candidates, something that the Aam Admi Party has been stressing in Punjab.
But this is not how power works in UP. The fundamental incentives to join politics and the competitive pressure on parties push them to maintain the same form of elite recruitment as in the recent past, although with some important variations for the BJP in terms of caste composition. The cost of entry into politics keeps rising, to the disadvantage of most honest political aspirants, including women and minorities or members of other disadvantaged groups. Thus, the new model retains important features of the old one, notably the elitism of political recruitment and the reliance on self-funding candidates.
The turnover of UP’s political class has also facilitated the concentration of power within parties and by extension within the government. Centralisation, polarisation, personalisation of state action and broad forms of mobilisation across identity groups through provision strategies, combined with the localised caste arithmetic of the past and its elitist encumbrances, are some of the enduring features of the new model of politics in the state. These are the bases of the BJP’s success in Uttar Pradesh, woven around the projection of strong leadership. Regional parties will not be able to counter this new model by clinging on to the formula of their past successes.
Gilles Verniers is an assistant professor of political science, director of Trivedi Centre for Political Data at Ashoka University, and a senior visiting fellow at the Centre for Policy Research. Views are personal.
