Ukraura Mela is one of the many events of the Hindu festive calendar in eastern Uttar Pradesh during the Dussehra-Diwali season and is often seen as the platform to showcase the political power of various Bahubalis of Azamgarh. This year, during that Mela, there was a massive gun battle that resulted in at least a dozen people being injured. It was a mini-war between two huge Bhandaras organized by cousins Vijay Yadav and Pramod Yadav.

Local eyewitnesses spoke of how the shooting incident virtually resembled a gangster Bollywood movie and everybody believes it was a miracle that the casualties were not very high considering the sheer size of crowd present in the Mela. “It was a Yadav battle for Azamgarh between cousins turned enemies”, described a local shopkeeper, further adding, “the next few months will be quite tense here”.

Azamgarh is Mulayam Singh Yadav’s Karma-Bhoomi in eastern Uttar Pradesh for it represents that cleavage of Muslim-Yadav demographic coalition which has defined his politics for 3 decades. Consequently, Mulayam’s protégé, the state’s Forest Minister and local Bahubali, Durga Prasad Yadav has almost never lost Azamgarh since 1985 (except for 1993 when he was outside the SP-BSP coalition) winning it for a record 7 terms.

It is widely believed in Azamgarh that a big reason for Durga Prasad Yadav’s success has been his large family base mostly managed by his nephew Pramod Yadav who along with his other Bahubali brother, Runnu Yadav have both been Block Pramukhs. It is also widely believed here that ever since Durga Prasad’s son, Vijay Yadav blocked Pramod Yadav’s election to ZP (Zilla Panchayat) Chairman’s position there has been a war brewing within the family which culminated in the Ukraura shootout of October 22nd.

Just 2 days later, some 300 kilo-meters away at the Samajwadi headquarters in the state’s capital of Lucknow, the Chief Minister of Uttar Pradesh, Akhilesh Yadav and his uncle, party president Shivpal Yadav, virtually came to fisticuffs on the stage in the full view of hundreds of party leaders and workers who in turn started a street battle amongst themselves!

From Azamgarh to Lucknow to every other district in between, the Yadav clans are fighting pitched political battles among themselves. This is perhaps that stage of evolution of political empowerment of a caste where power becomes mutually exclusive within a family to the extent that mutual destruction becomes the sole survival option. The impact of this Yadav clash is potentially going to be very high on the upcoming assembly elections.

In about 3 months, the approval ratings of the state government have gone down by a huge 9 points and it is now in that deeply anti-incumbent territory of the sub 40% range which historically produces a complete opposition sweep in the ensuing elections – for instance, in the recently concluded Assam elections, the state government’s approval ratings never fell below 48% and yet it lost the election pretty badly.

Akhilesh Yadav must be a worried man, for the voter perception which was already negative due to the law & order situation has further spiralled out of control due to apparent lack of governance over the last 2 months. “First they all (Mulayam family) fought among themselves like wolves and dogs over meat, now he (the CM) has embarked on election campaign in a Mercedes van” says 24-year-old unemployed Mahesh Yadav with doles of sarcasm. His mother was suffering from a bout of Dengue-like fever throughout this festive season and there were not enough beds in the civil hospital, so his anger sums up the frustration of a state that is in a constant self-imposed turmoil.

A section of the media, including some big English language newspapers are heavily promoting Akhilesh as the young face of Uttar Pradesh and are even portraying that he has huge mass support, but most of these are exercises in futility, for the anger is quite palpable everywhere. Yes, there is a vocal section of the Samajwadi youth brigade that is creating a lot of noise in favour of the CM, but that can hardly hide the anti-incumbency anger of ordinary people, nor can it gloss over the inherent divisions at every level of the party. In every district, there are at the bare minimum, at least 2 different groups of Samajwadi leaders nurtured by CM Akhilesh Yadav and party president Shivpal Yadav respectively.

Akhilesh may be popular among the youth brigade of the party as many young leaders have benefited from his largesse in various contracts and Theka-permits, but Shivpal has also nurtured those long-time party workers who haven’t got much power or money in the last 5 years of SP rule. In fact, Mulayam wasn’t very wrong when he admonished his son openly in that party meet on October 24th, “What have you done? You have just nurtured Juwaris (gamblers) and Sharabis (drunkards) in your regime!” This is indeed the common refrain of many old Samajwadi warhorses in various districts where they see the young brigade as brazenly indulging in their vices after achieving power and making money. Such a vertical division in the party has left many leaders wondering how they would face the electorate this time.

Take the case of Azamgarh assembly seat as an example. Pramod Yadav, the nephew, has now been vying for a BSP ticket to challenge his uncle and Forest Minister, Durga Prasad Yadav who was otherwise seen as virtually undefeatable there. Now it has all the portends of turning into a battle royal with guns and gangsters thrown into the mix. If BJP nominates the right candidate, there is every chance that the saffron party may end up as a straight beneficiary in this uncle-nephew Yadav battle.

Our polling also shows that both the sitting MPs and MLAs are also facing a big bout of anti-incumbency, but the latter were significantly more unpopular. Riding on the surgical strikes success, the central government seems to be enjoying a massively high positive rating of 86% which is a historically significant number that suggests a tsunami of support building in favour of Modi. The contrast of how the voters see the state govt. vis-à-vis the central govt. could not have been starker – it is also testimonial to the intelligence of the average voter in the state who makes a clear distinction no matter how much the media narrative tries to show a state of false equivalence.

How is Akhilesh dealing with all these dangerous prospects for the ruling party? He seems to believe that elections can be won purely by hiring good political consultants. In fact, he has gone ahead and hired Harvard University’s Kennedy School professor and leading Democratic party consultant, Steve Jarding to manage his re-election campaign. One of the first things that Jarding has created is Akhilesh as the ‘family man’, so we see the CM playing cricket with his kids and sitting with his (photogenic) wife all over the adverts and images. This is the cornerstone of American political campaigns where candidates have to be seen as “doting on their families” to win the hearts of ordinary voters. In India, where Bhai-Bhatijawad (especially in parties like SP) is rampant in the political sphere how will voters react to such messaging is anybody’s guess. Imagine the pain and anger of ordinary people suffering from Dengue & Chikangunya waiting for a hospital bed when they see the CM fighting with his uncle one moment and then playing cricket with his kids the next moment!

Hubris has been the political currency all over the left-liberal world recently and Steve Jarding and his ilk of Democratic political consultants have been at the forefront of Hillary Clinton’s “shocking” defeat in the US, but here in India he is using those same mantras of training 4000 volunteers to spread brand Akhilesh across the state and terming “Uttar Pradesh is my family” etc. If only these political consultants understood even 10% of the angst of today’s world, they would probably not make such blatant mistakes.

In a straightforward swap election, Mayawati’s BSP should have been the direct beneficiary of all the anti-incumbency. Indeed, she is still the most popular face for leadership choice in the state, while Akhilesh has remained static at around 14% (just a 1 point drop from our previous survey). Elections are heavily personality centric these days and Maya has every chance of creating a campaign built around her as the most viable alternative. Yet, there seems to be one obstacle in Maya’s path and that obstacle is named as Narendra Bhai Damodardas Modi.

The recent Bihar defeat of BJP and the Assam victory in contrast have created a clear political narrative about the need to anoint a state-level leadership face to represent the saffron vote surge. Most BJP local leaders too agree that the party needs a mascot to create a successful campaign in this most populous state. Yet, data suggests a somewhat convoluted path.

Throughout the 2014 election cycle, Modi’s popularity & approval ratings across India and in specific election bound states like Maharashtra and Haryana remained above the 60th percentile in general when BJP also managed to win a host of elections. In the 2015 Bihar election, however, the Prime Minister’s popularity rating had just fallen below 50% which also correlated with BJP’s defeat along with various other reasons. Today, after the surgical strikes against Pakistan (this poll was done before demonetization of 1000 & 500 Rupee notes), the Prime Minister’s popularity rating is once again hitting its peak which may override any local political vagaries to create a favourable atmosphere for the national ruling party. In fact, there are 5 distinct reasons why BJP may benefit from this move of not declaring a CM candidate;

  • Unlike Bihar, Prime Minister Modi represents Uttar Pradesh in parliament and is virtually seen as a mascot of the state cutting across caste lines which augurs well for the USHV base of BJP
  • BJP has to adapt a three-layered demographic progression model in UP as it did in 2014 – A] create a subtle Hindu polarization, B] enable all caste groups, especially non-Yadav OBCs, to enthusiastically participate in the elections and C] articulate a clear development agenda for the party based on anti-corruption and good economics modules. As of today, no BJP leader can symbolize this triple layered model as best as Modi can
  • Among all the states of India, Uttar Pradesh demographic is the most inclined to weigh their votes with a ‘national political perspective’ by foregoing regional considerations. Essentially what this means is that UP voters want to be indulged by underlying their importance to national politics which they have rediscovered fully in 2014. (This national political obsession of UP voters is because of 2 reasons – 1] the state itself is not a homogenous entity like a Bihar or a Gujarat or a Karnataka to have a high degree of regional identity appeal and 2] its closeness to the power of Delhi both geographically & culturally as well as due its inherent weightage of 80 parliamentarians)
  • PM Modi is best placed to exploit the two big national issues of India’s aggressive response to Pakistan and the central government’s revolutionary new steps at curbing black money which will both find resonance among the UP voters if packaged creatively
  • There are no pan-UP leaders with a broad development and nationalist appeal in state BJP today, especially in the post-Kalyan Singh era. For example, Yogi Adityanath was emerging as the consensus choice in July, but since the party did not make any announcement, his popularity seems to be dimming purely on its own weight.

Although employment still remains the single biggest concern of people in UP, agrarian crisis seems to have relatively eased due to a better monsoon as compared to July. Interestingly, law & order problems have gained further mind-space among the voters, possibly due to the constant bickering and fights among the Yadav clans which is really hurting the state government much more than anything else. For the first time, we conducted a new case study on the impact of various central govt. schemes like PMJDY and the Ujjwala Yojana to assess the tangible changes to voter behaviour which has provided us with some very curious numbers.

A whopping 34 lakh BPL families of UP have benefited from the PM Ujjwala Yojana which provides for free LPG cooking gas with the savings made from voluntary surrendering of government subsidized cooking gas cylinders. This one single measure by the central govt. has the potential to upturn the UP election as it can directly influence a voter base of nearly 1.2-1.8 crore people by February next year!

To put things in perspective; BJP received totally 3.43 crore votes in 2014 while SP and BSP received 1.79 and 1.59 crore votes respectively and just this 1 scheme of providing cooking gas to families that heavily depended on unclean cooking fuels (like wood, kerosene, animal dung etc. which resulted in over 5 lakh death every year as per WHO statistics) can empower anywhere between 1.2-1.8 crore voters to think beyond caste silos and vote for real change. Our poll shows that 63% of rural housewives believe that this Yojana will make an impact on their voting decision. Going further, this single govt. measure may have a much bigger impact on voting decision than the surgical strike against Pakistan. In fact, in a decentralized trajectory, if this yojana manages to accrue even a basic average of 12-18k votes to BJP in each assembly seat (with a cumulative of 6 million votes), that could potentially impact the results in 70 to 90 seats of the state. Comparatively, the impact of PMJDY seems to be far lesser until now.

This is where BJP gets a big starting advantage while designing an election campaign. Voters tend to associate development agenda far better with tangible results when assessing the promises of a political party. This was one of the problem areas for BJP during the Bihar campaign last year where the voters simply had nothing substantial or realistic to relate to (despite the special package announcement which only sounded hollow) while opposition created a storm about reservations. Now, BJP and PM Modi can easily build a substantial campaign in UP about the party’s plans to generate employment and maintain law & order and solve health care problems. Voters would also have a greater degree of belief in the saffron campaign because of the success of Ujjwala Yojana & PMJDY etc.

Ruling Samajwadi Party is seeing a declining trend in both of its core voters of Yadavs and Muslims. As of today, the Muslim vote is facing the greatest uncertainty which is reflecting in our poll where almost 1 out of 4 Muslim voters in the state is still in the ‘undecided’ category. Essentially, Muslim voter is likely to break out late into the election and we need to constantly keep track of them.

Jab voter jeetata hai tho kissi pe ehsaan nahin karta aur jab haraata hai tho kissi ka lihaaz nahin karta” was the cryptic one-liner by Bachchan Miyan, a 65 year old fruit vendor near Allahabad who in his own words “has seen it all” in his life in the state. That underlying Muslim frustration with the state government in particular and the state of politics in general is a whole new phenomenon in UP. The UP Muslim voter has never felt so powerless, not even after the demolition of the Babri Masjid in 1992, for their votes always made the difference in the end as even in the 1993 election they were able to prevent a saffron victory at the peak of Kalyan Singh’s popularity.

Mayawati is trying her best to woo the Muslim vote of Uttar Pradesh. She is actually quite desperate which is showing in her speeches where she is putting up a direct appeal to Muslims not to split their votes and lead to a BJP victory. Her party, BSP is also openly communalizing the election by reciting the verses of Quran in their election gatherings. And yet, despite all these efforts, Muslims at least till now are not showing any signs of changing their historic pattern as about 22% are voting for BSP as of today. Just why is this Muslim-Dalit coalition so difficult to pull off?

  • Mathematically, historic electoral-demographic data shows an upper resistance level at the 30th percentile which would be a difficult ceiling to break unless there is some path-breaking change. Maya wrongly believes that increasing Muslim candidates by BSP is enough on its own to woo the minority vote, but past data shows a contrarian trend
  • Muslims and Dalits vie for similar socio-economic space in the UP society wherein Muslims feel that they are ‘artificially’ repressed due to reservations that the latter are entitled to
  • Most ordinary Muslims have never seen the party as their own and there is a widespread notion among them that BSP is essentially a “Chamaron ki party”
  • BSP lacks structural midlevel Muslim leadership at the districts and divisions unlike SP where Neta Ji has nurtured a solid ecosystem for decades
  • Most Muslim opinion makers like Maulvis and local Urdu press is deeply incorporated into the Samajwadi political system and is unlikely to abandon that even in 2017
  • Politically Maya is quite unpredictable and many Muslims still remember that she is not averse to join hands with the BJP which essentially means their own vote might be wasted in the end
  • After the 2014 bitter experience where many Muslim voters ‘wasted’ their vote on BSP which won zero seats, many of them believe that Maya simply cannot stop saffron surge despite her Dalit vote-base

We also conducted an addendum polling exercise by indigenously adapting a probability questionnaire model developed by the Center for Economic and Social Research and the Jesse M. Unruh Institute for Politics, both at the USC Dornsife College of Letters, Arts and Sciences, wherein we gave the voters a chance to quantify their leanings by choosing a range on a scale of zero to ten. While most core voters of different political parties like Jatavs for BSP and Brahmins & Thakurs for BJP displayed a clearly stable choice, Yadav support for the ruling SP was much more vulnerable. In fact, Yadavs and non-Yadav OBC’s (leaning in bigger numbers towards BJP) displayed a clear contrast in their stability of the vote-choice. On the other hand, Jatavs showed greater stability that non-Jatav Dalits in their political choice.

This exercise gave us a very interesting insight into voters’ minds based on their own definition of political power. In India, elections always operate on the ‘winner-takes-it-all’ phenomenon which essentially works on the invisible collectivism hypothesis wherein many seemingly disparate demographic subsets join hands to create an ‘unexpected’ majority for a political party. It is almost as if majority of the voters suddenly come together by design to produce a mandate. This happens because mostly average Indians do not want to waste their votes and decide to exercise their franchise towards the leading party in a state. Our model suggests that Muslims (with large undecided segments), Yadavs and non-Jatav Dalits of UP are in that special zone today wherein they are likely to go with the leading party to create a mandate.

On the whole BJP enjoys a clear 9% lead in UP as of early November (before the demonetization drive of the central government) which should potentially lead to a massive saffron sweep in the state. In fact, this is consistent with not only the shift in electorate of Uttar Pradesh but also the electoral data of the last 2 years across India.

Ruling Samajwadi party is stuck around the 20% mark and is steadily losing support at about 1% every 3 odd months. Of course, it may gain a few more points once the undecided voters breakout, but it looks highly unlikely that it will come any closer to the 30th percentile on its own because of A] sheer anti-incumbency, B] internal clan wars which have percolated down to the last common denominator and C] uncertain core vote-base of both Muslims and Yadavs

BSP has gained about 5 points in terms of vote-share which amounts to 20% extra gain as compared to 2014, but is struggling to go beyond that. BSP’s main obstacles are that it is unable to attract votes beyond the core base of Jatavs as even other, non-Jatav, Dalits are not fully sure about backing the party as it is suffering large scale desertions and lacklustre Muslim support. Yet, as of today, BSP remains the biggest challenger to a saffron surge in the state.

Since the peak of 2014 LS elections, BJP’s vote-share has remained adamant for most parts. Even when the party lost Bihar last year, it only dropped about 5 percentage points. Throughout the last 2 years the party has lost anywhere between 0 to 9 percentage points in state elections against the 2014 benchmark. Our model suggests that even in UP the floor for party’s vote-share could not be lower. Indeed, our poll projection is at the lower end of the spectrum as BJP has lost about 10% votes from 2014 with 12% undecided (and others) voters.

Although a sub-regional vote analysis would be futile at this point when candidates haven’t yet been announced, it gives us a peek into the demographic behavioural patterns. BJP’s sub-regional hegemony of 2014 has been more-or-less sustained. The party’s best performance comes from the east which also has the largest share of assembly seats and is home to 29 districts out of 75. Even in West UP, the polarization of 2014 is still active even though at a reduced pitch, which is giving BJP enough traction to gain a lead. The only region where BJP is not leading as of today is in Awadh which is much smaller in terms of overall impact. Looking at the trajectory, it is clear that BJP has gained relative to its opposition in the west while it has remained static in the east as compared to our own poll in July. The upside potential for BJP is quite high in all regions, so ticket distribution could be a crucial exercise for the party.

As of today, the only way that a saffron surge can be challenged, let alone be stopped, is by an opposition alliance. Unlike Bihar where all the opposition parties came together to fight under a united umbrella, here such a possibility is next to nil simply because neither Maya nor Akhilesh would be willing to give up on the CMs post (also neither of them face any legal limitations like Lalu Yadav who could afford to accept Nitish Kumar’s leadership). Among other possibilities, the coming together of all the erstwhile Janata parties like RLD and SP in alliance with Congress is the most realistic one. The problem with such an arrangement is that their votes may not really add up in a neat structural form creating greater fissures for BJP to exploit. Consequently, there are 5 problem areas for such an alliance;

  1. In West-UP, where the Jat-Muslim (and the overall Hindu-Muslim faultline) divide is extremely well demarcated, an alliance between RLD and SP could be a death knell for Ajit Singh’s party which may have recovered some ground otherwise. In such a scenario, polarization could end up being sharper than in 2014 resulting in a virtual sweep by BJP in this region
  2. Dalits, especially non-Jatav Dalits who have shown a higher degree of uncertainty in our voter-scale index, may then find it more viable to shift to BJP in such a scenario – saffronization of Dalits would then be complete and Modi would go unchallenged for a very long time in Indian elections
  3. The Akhilesh-Shivpal fault-line and the other internecine Yadav battles could come to the fore leading to many powerful rebel candidates from the ruling party contesting as independents
  4. BJP can strategically exploit reverse Hindu polarization across the state and Muslim votes may yet get divided, especially if a desperate Maya ties up with smaller Muslim regional parties like MIM and Peace Party
  5. With the rare exception of Bihar, whenever Congress has allied with other parties, it has only managed to harm them – the 3 latest examples being Bengal, Tamil Nadu and Assam. We tend to overlook the fact that average Indian voters are still not ready to forgive the Congress party even after the punishment of 2014.

This is also the reason why we believe that BJP not announcing a CM candidate is a good decision as of now, for it gives the party that much leeway to take strategic decisions whenever a need arises. In case a SP-RLD-Congress alliance becomes a reality, BJP can always announce a Dalit CM candidate to reap rich benefits of a massive Dalit surge in its favour. Similarly, in the unlikely event of a Congress-BSP alliance, the saffron party has the option of tapping Yadavs who in combination with other OBCs would mean a total vote-base of a whopping 40%+ on its own.

Our job as a pollster and a psephologist comes to an end here, but Indian election analysis would not be complete without projecting seat-shares which itself is nothing but a dark art due to non-availability of any statistical or mathematical model to convert votes to seats. Therefore, the error margins of our polls do not hold true for our seat projections.

A Note on 5Forty3 Uttar Pradesh Poll survey of October

This UP poll survey was conducted between October 15th and October 28th spread across 190 polling station areas (including swing polling booths) covering 63 specially chosen assembly segments of all the different electoral zones of the state. Our survey covered 281 geographic locations of 26 districts and had a target sample size of 5800, giving adequately weighted representation to various castes and socio-economic groups of both the rural and urban populace (see the below tables)

This 5Forty3 survey has a statistical error margin of 3.1% and a historic error margin of 2.7% (based on our past record). Our polls are especially accurate because of our path-breaking swing polling booth surveys based on indigenously developed tool RSSI (Randomized Social Swing Impact) which gave us near 100% accurate results during the Bihar elections where every other pollster failed. Based on our long experience we can say that there are three main error zones for political poll surveys in India;

  • Selection bias: Our methodology has possibly a near perfect selection probability, therefore this poll survey has produced one of the most robust results.
  • Random Sampling Error: Since we have used a controlled random sampling methodology this problem has been almost completely negated; for instance, most pollsters simply use a randomized sample across the geography to arrive at findings, but our system has extensively used electoral rolls in conjecture with our own ethnic (caste) database as the sampling frame.
  • Non-Response evaluation error: Most poll surveys completely ignore “no responses” and simply concentrate on the positive responses, which artificially inflates findings. Our system actually treats “no responses” as positive findings and assigns value to them, which adds a whole new dimension to understanding people’s choices.

Clear and precise questionnaires containing 67 different questions based on 3 criteria – preferences, opinions and factual information – were prepared in Hindi and English which were used to elicit the opinion of the respondents by carefully trained fieldworkers. The entire survey was conducted by direct face-to-face interviews without employing any CATI surveys or phone surveys. We achieved an extremely frugal hit-rate of 125 rupees per response which is much lower even for the historically cost effective team 5Forty3 (industry standard rates in India are roughly 200 to 270 rupees per response).