Fabricated forecasts are possible though exit polls do go wrong on valid grounds
Exit polls can go wrong because of vocal vs silent vote, fear of retribution, a thin sample size, untrained surveyors, diversity and complexity and various other reasons. They can also be dishonest
I believe in miracles, said Australia’s Prime Minister Scott Morrison on Sunday. His coalition government has just been re-elected to power. It was viewed by everyone as a ‘miracle’ – because every single Exit Poll had predicted that the rival Labour party would win with a big majority.
It was the same day on which India’s election ended and Exit Polls predicted a thumping victory for the Modi government. Shashi Tharoor was among the many who referred to the ‘miracle’ in Australia. “There were 56 independent Exit Polls and all forecast the same result and all of them were wrong”. He added: “In India, voters don't always tell pollsters about who they voted for. So with very small samples the agencies invariably get it horribly wrong”.
What he did not say was that in India, polling agencies don’t always tell the truth either. The reasons can be sheer dishonesty or sheer incompetence. Whatever the reason the fact is that 80% of Exit Polls predictions in India cannot be trusted at all.
This is precisely what Mamata Banerjee had said Sunday morning, hours before the evening TV shows: “I don’t trust Exit Poll gossip. The game plan is to manipulate or replace thousands of EVMs through this gossip. I appeal to all Opposition parties to be united, strong and bold. We will fight this battle together”.
Each of the four sentences in her 7 a.m. twitter message is imbued with meaning and is worthy of closer evaluation.
By saying she does not trust “Exit Poll gossip”, she used the word that was most apt. Gossip. Anyone who sat before the TV screens on Sunday and saw the helicopters and swanky cars inside the studios and listened to the gossipy chitchat would know that it was not a serious election analysis but just a glamorous game of non-stop entertainment.
Mamata’s second sentence “the game plan is to manipulate or replace thousands of EVMs through this gossip” is a devastating expose of the intention behind the fraudulent exercise.
Those who view this as an unfounded allegation being made without a shred of evidence are either incurable optimists or blind believers of official denials about the vulnerability of electronic voting machines. There is enough evidence world-wide to prove that EVMs are not-tamper-proof. There have also been regular instances of voting machines going missing or malfunctioning. But, as the saying goes, “a man hears what he wants to hear and disregards the rest”.
Regarding the third sentence of the Trinamool supremo’s tweet, it is an astute political call to Opposition parties to remain united. The urgent relevance of his appeal can be seen by the impact that the “300 for BJP” forecast has had – and was meant to have.
In the three days before actual counting of votes takes place, it will indeed be a challenge for level-headed Opposition leaders to shore up the morale of some who have already mentally accepted the Exit Polls as reliable.
In her fourth sentence Mamata has made another political statement by saying: We will fight this battle together. If the Exit Polls have got it right, there is no battle left to fight – the BJP has won and it’s all over bar the oath-taking for Narendra Modi.
If, on the other hand, the Exit Polls are nothing more than an exercise in mass deception, there is a real possibility that the numbers thrown up after the Election Commission announces the final result will give the Opposition parties a fighting chance to stake their claim to form the next government. For this, as Mamata has said, the need of the hour is to “united, strong and bold”.
As far as polling agencies are concerned, it is not generally understood why Exit Polls get it wrong more often than they get it right. The reality is that exit polls have an innate tendency to overstate the case of vocal voters and entirely miss the silent vote. This is a well-documented worldwide flaw in the exit poll methodology.
But in India, there is another factor has become critical in the present political climate. Fear has crept in. Many false or misleading responses by individual voters is driven by fear of retribution.
Apart from that, simple arithmetic can go wrong when there is a higher voter turnout than had initially been anticipated. This can completely skew all the previously assumed weightages - leading inevitably to erroneous calls on trajectory and final tally.
As an expert put it: "Exit polls tend to get it right when there is a clear edge for one side at the outset of the election. But exit polls can go haywire in close contests, especially when the sample size is thin”. Moreover, collecting raw data accurately may be easy - but calculating vote percentages on the basis of samples and then converting it into seats is an altogether more difficult ballgame.
Glaring discrepancies in the Exit polls have been listed by Uttam Sengupta which show just how flawed the process is and which rip apart the last vestiges of credibility of the agencies and anchors involved in misleading the country:
a) The Exit polls show BJP losses in all states except West Bengal and Orissa. And yet the NDA is shown as winning comfortably. Illogical.
b) The gap between the highest and lowest predictions for NDA is nearly 100 seats! It is a 110 seats gap for UPA.
c) Unbelievably, Chanakya agency has the same numbers in 2019 as it did in 2014: NDA 340, UPA 70, Others 133. Shocking to say the least.
d) Axis has claimed that BJP would win in Anantnag in the Kashmir Valley. If that proves true one can believe anything.
e) Mulayam Singh is forecast to lose Mainpuri and the BSP will lose Ambedkar Nagar. Not impossible but certainly something to ponder over.
f) For Tamil Nadu, the VMR exit poll says: BJP: 29 Cong: 09 Others: 0-0. This has to be a blatant typographical mistake. Question is – how many more careless mistakes are there?
g) Likewise Axis forecast the Odisha tally as - BJP 15-19 BJD 2-6. That would have to be either an election miracle or just plain EVM rigging.
h) ABP News projects the ‘Gathbandhan’ of SP-BSP-RLD winning 56 seats in Uttar Pradesh but India Today TV predicts 10 to 16 seats. Obviously, both cannot be correct. Both claim to be trained professional psychologists. So can one agency predict 56 and another 10 – and yet ultimately arrive at roughly the same overall national outcome? Fake survey, very obviously.
i) Times Now projected a 2.9% vote share for Aam Aadmi Party in Uttarakhand - the AAP party has not fielded any candidate in Uttarakhand. Just oversight? Or more proof of fraudulent figures fabricated without sample feedback?
A valid question to ask is whether the fabricated forecasts were inadvertent errors and omissions or deliberate deceit with criminal intent.
A gigantic disinformation campaign to mislead citizens with falsified statistics is not a trivial matter. At stake is the future of the country and attempts to muddy public perceptions just prior to counting day could arguably be viewed as a form of anti-national activity.