In March 1977 there were 560297 registered voters in Mangaldoi Lok Sabha constituency of Assam which was among the rare seats that the JP movement managed to wrest from Indira Gandhi’s Congress Party that had won 10 of the 14 LS seats in the state contrary to the JP wave breezing across India. Almost exactly a year later when Hari Lal Tiwary, the sitting Janata MP, passed away, Election Commission of India was forced to conduct a by-poll to Mangaldoi LS seat. Within the span of 1 year, defying all demographic logic, the total number of electorate in Mangaldoi had increased by a whopping 80000 voters – essentially some 15% new people had miraculously sprung-up almost overnight from thin air through immaculate conception (pun unintended).

Congress party, having lost the seat in 1977, had virtually “imported” some 70000 Bangladeshi Muslim refugees into Mangaldoi for purely electoral gains. This was indeed the logical progression of vote-bank politics practiced unabashedly by the Congress party. More than 70000 cases were filed against the authenticity of these “imported voters” by various citizen groups and the court sustained 26900 of those cases. This is how “illegal Bangladeshis” entered the political lexicon of India and some 4 decades later it still remains a festering demographic problem only having become much larger in scale.

In fact, the scale of the problem has reached such humungous proportions that a political party catering almost exclusively to Bangla speaking Muslim refugees of Assam is now the third biggest electoral entity in the state. Yet, Badruddin Ajmal, the perfume baron is seldom described as “communal” whereas reams and reams of editorial space is being used up to paint BJP’s communal agenda against “illegal immigrants”; this is how the left-liberal media ecosystem manipulates narrative constructs in India.


When we asked to identify the issues that matter to the voters most (beyond regular suspects like corruption and price-rise which are often electorally insignificant), the migrant issue was by far the biggest problem that most voters identified with (37%). In fact, when narrowed down to purely Hindu demography an overwhelming 59% of the voters thought that “illegal migrants” was the most important electoral issue for them. No wonder then that BJP has made this as its core electoral plank. At last count some 20000 RSS workers are fanning out across the state and constantly educating people about the “illegals” and that effort is having its impact on the voters. So much so that the migrant issue may well transform into a wave by the time the state goes into polls next week.

Among other major issues, apart from a host of tea-garden economic woes and agrarian distress (18%), the one other big issue is the status of various smaller tribes which has found a lot of resonance in hinterland Assam (15%). BJP has on the one hand tied up with BPF (the often militant Bodo Peoples Front) while at the same time the Modi government has set up a committee to grant ST status to 6 OBC groups like Moran, Ahom, Tai etc. This is indeed a tightrope balancing act in the complex demographic maze of Assam where the existing tribe’s interests have to be safeguarded while ensuring fare reservations to newer OBCs. It is this complex arithmetic that may well decide the course of Assam elections. As of today, both Congress and BJP are perceived to be equal on this point, but then BJP is a late entrant in this maze and it has gained momentum in the last few months.

Broadly on all other parameters of the development agenda most voters believe that BJP is better placed to solve their problems of employment, jobs, floods et al. This is indeed indicative of the 15-year anti-incumbency of the Congress party which has created a saffron window of opportunity wherein people of Assam want to give a chance to a fresh new political leadership.


Here again Sarbananda Sonowal, the chief-ministerial face of the BJP, has narrowed the gap with the sitting CM, Tarun Gogoi. Past polling experience has shown us that incumbent Chief Ministers usually enjoy much higher recognition among the voters of a geography simply because of their sheer visibility and in such a backdrop if Sonowal is enjoying such high trust levels among the voters it simply shows that Assam is ready for change. In fact, there are two important trends that emerge out of this pre-poll survey which give us a broader hint of how Assamese voters are likely to vote this April.

  1. Majority of the Indian masses have an inherent bias towards a leading brand which translates into disproportionate market-share for market leaders (for example Maruti in the car segment or Colgate in the oral care segment etc.). This same principle also works in Indian elections wherein most voters prefer to cast their votes to the winning party rather than waste their democratic franchise. Thus in the past 15 years in Assam, Congress had always been ahead in the perception battle in the state which had enabled it to win three elections in a row. Today, voters have begun to believe that BJP led by Sarbananda Sonowal in the state and Modi nationally has an upper hand in the state which is reflected in both the leadership race (between Gogoi and Sonowal) as well as in solution matrix to Assam’s problems. This perception of BJP alliance being in a position to defeat Congress itself is enough to propel a large number of fence-sitting voters on the polling day.
  2. Contrary to media narrative, Assamese voters perceive BJP as a party of “development” along with an undercurrent of anti-immigrant stance. In fact, this new Modi led BJP which is overtly a “development agenda” party that has Hindutva as its unstated ideological core is what makes it such an attractive choice to a large number of voters not only in Assam but also in rest of India (whenever BJP has wavered from this political binary, as it did unnecessarily in Bihar, voters have moved away from it).

Ethnic v/s religious faultlines and USHV

Deciphering the electoral trends in eastern Indian states is slightly different from the rest of India where most of our voter-weightage sampling is based on caste lines. Here in Assam, caste is at best in peripheral existence and most of the opinion expressions come through ethnic identities. For instance, the constant friction between majority Assamese Hindus and Bengali speaking Hindus who are about 25% of the populace is mostly on ethnic lines and the fact that the Bengali speaking Hindus are majorly from the Dalit community is simply incidental to the whole demographic dynamic.

Thus, as a pollster one has to be extremely careful while designing surveys for a state like Assam using VWISM (Voter Weightage Index Sampling Methodology), lest the caste expressions of the heartland creep into the predictive arithmetic. Yet, the template of vote pattern is similar to the rest of India. Mostly Assam till now has been voting along ethnic faultlines which is why Congress has been traditionally strong here – the one time those ethnic faultlines got blurred, Prafulla Kumar Mohanta won a historic mandate in 1985. Just as in rest of India where we are seeing a unique trend of United Spectrum of Hindu Votes (USHV) coming together by overcoming caste boundaries, Assam too is poised today to express itself more on religious faultlines where ethnic splits are blurred.

The development agenda of BJP built on the edifice of the anti-illegal-migrant stance of the RSS is managing to wave together major ethnic groups like Tai Ahoms, Sooteas, Morans and Rajbongshis along with tribal groups like Bodos and Sonowals under one broad spectrum of Hindu vote. In fact, in most districts that we conducted our survey we found a typical geographic pattern that clearly demonstrates this newly developing religious faultline. In many larger towns of various districts (including Muslim dominated districts like Dhubri), we found that typically BJP’s support is much higher in the central parts of the town whereas the party weakens in the outskirts and peripheries – on many instances the contrast is at extremes. This difference underlines the fact that central parts of the towns in Assam are still home to original Hindu Assamese populace whereas outskirts are dens of illegal Bangladeshi migrants.


This differing vote pattern was also clearly visible sub-regionally when team 5Forty3 gathered people’s opinions. For instance, in the Muslim dominated Barak Valley, BJP enjoys upper hand in assembly segment pockets with higher Hindu population in districts like Cachar and Karimganj (represented by saffron dots in the map) whereas it has near zero traction in a Muslim dominated district like Hailakandi. Thus essentially, the religious faultline is active everywhere whereas intra-Hindu ethnic faultlines are mostly blurring.

Electorally, we have divided the state into 4 distinct geographies;

  • Upper Assam making up 34 assembly segments – this is where BJP is virtually sweeping in the pre-poll numbers. The saffron alliance is getting a major chunk of Brahmin and Ahom voters as of today while Tea-Tribe and Sooteas seem to be almost evenly split among Congress and BJP (most of the larger tea-tribes here have been traditional Congress voters, but BJP has made huge inroads in the last 2 years). In such a scenario, smaller tribal voters could prove to be decisive swing factors, but as we found in our survey, most of these smaller tribes are not supporting Congress as they did traditionally. For instance, the Mishing tribal vote which traditionally went to Congress overwhelmingly seems to be shifting at least partially towards Sanmalita Ganashakti which has put up some 4-5 candidates in this region.
  • North-Central Assam constituting 30 assembly segments – this is where the battle seems to be evenly matched but what is hurting BJP here is the bad ticket distribution and also some heart burn among its grass-root workers who have questioned the logic of an alliance with the AGP. On the other hand, CM Gogoi’s “go it alone” policy may hurt Congress’s prospects here as AIUDF has got some traction among Muslim voters.
  • Barak Valley has 15 seats – here the main contest is between Congress and AIUDF, but BJP is expected to win a handful of seats especially after a widely successful Modi rally
  • Lower Assam is home to the largest chunk of 47 assembly segments – this could be the crucial swing geography which may decide the eventual winner in the state. A large number of Muslim voters play an important role here and our survey numbers are showing surprising underperformance of Badruddin Ajmal’s AIUDF, but it must be stressed here that Muslim voters usually tend to under-respond during pre-poll surveys when it comes to choose Muslim parties. BJP’s strength is limited to urban geographies like Guwahati and Kamrup districts so it’s performance is mostly dependent on how well its alliance partners can perform, especially BPF which seems to be facing a tough fight in 6 out of 12 Bodo dominated seats as of today.


In the pre-poll scenario, just days before the elections, the BJP alliance seems to be enjoying a healthy lead of 5% over its nearest rival which under normal circumstances should be enough for a sweep in a multi-cornered election. As compared to the previous assembly election of 2011, BJP alliance is set to gain a positive swing of 6% votes whereas the Congress party is suffering a negative swing of 4%. Essentially, BJP seems to be getting the advantage of the sheer arithmetic of its alliance with AGP and BPF along with a shift of anti-incumbency vote from the Congress party. As compared to the 2014 LS Polls, the BJP alliance seems to be losing some votes due to 3% negative swing whereas the Congress party is gaining smartly at about 6% positive swing. These findings are consistent with BJP’s historically exaggerated performances in national elections as compared to state polls and also the declining saffron trend of the last 2 years after the historic 2014 Modi mandate. There are of course 5 important X factors that must be considered before reaching any conclusions based on these pre-poll findings;

  1. Generally, in Indian ‘winner-takes-it-all’ elections the frontrunner party or alliance enjoys a strong advantage on the day of polling wherein more voters tend to vote for the winning party than in pre-poll surveys which could give the BJP alliance a further push towards a larger sweep.
  2. In our pre-poll survey, we found that as many as 5% of the Assamese voters were still undecided (or unwilling) fence-sitters just days before the elections which could play a crucial role in deciding the eventual winner. Interestingly, the lead gap that BJP alliance enjoys over Congress also stands at 5% and off late BJP seems to have developed a unique disease of underperforming on polling days despite enjoying a clear advantage in the run-up to the election – what we have termed as the “last 72 hour syndrome” which played a crucial role in some recent setbacks including the one in Bihar. Indeed, that is one of the reasons why we put greater emphasis on exit/post polls than pre-polls because Indian elections are swinging so wildly in this information overload era.
  3. BJP’s ticket distribution errors and the heartburn of karyakartas is also a crucial X factor, especially because the party has developed a unique ability to internally sabotage elections in the last 1 year or so. But here in the case of Assam, the strong presence of the Sangh on the ground to tide away any internal differences may stem the problems to some extent
  4. In our interactions with ground-workers and cadre of different parties we found that generally an average Congress worker was nearly 40-60% less upbeat than an average BJP worker which is indicating a state of very low morale for the ruling party and typically it is a pre-cursor for a big defeat. In fact, the state of the Congress party is best summed up in the words of a hardcore Congress worker who had this to say about the coming assembly elections, “since 2013, Congress party has lost 9 state elections to BJP-NDA apart from the national election… we need a miracle to win Assam”.
  5. As of today, AIUDF seems to be underperforming in this election, surprisingly even among Bengali speaking Muslims of the state, and this could be a crucial factor on the day of polling especially as some local BJP leaders seem to be suggesting a “fixed match” between Ajmal and Gogoi. The only way that Congress can possibly win Assam today is if there is largescale Muslim vote transfer from AIUDF.

Despite all the pre-poll caveats of 3% error margin, as things stand today, BJP seems to be winning Assam. Now how does one convert this vote-share projection into actual seats? That is indeed the most difficult arithmetic of Indian elections since there is no clear statistical correlation from votes to seats, especially more so in a state like Assam with completely contrasting sub-regional dynamics. Thus a 5% gap in a multi-cornered fight may produce a complete sweep or even a nearly hung assembly! Yet, by deriving statistical averages of different sub-regions, we have tried to give a rough seat projection for the state of Assam.


In conclusion, the north eastern sunrise of Hindu polity is a new chapter in India’s evolution. A victory in Assam would not only have far reaching impact on other north-eastern states like Manipur and Arunachal Pradesh but also act as a cushion for the saffron party in the 2019 LS polls by giving a comfort zone of some one-dozen seats. For the Congress party a loss in Assam would mean the party’s last fortresses of the north-east and hill states are also crumbling and India is well on its course of a “Congress Mukt Bharat”.