David Dimbleby, the famous English journalist who hosted the BBC Election Night coverage from 1979 to 2017, said a few days ago that exit polls are “the worst invention ever” because their increasing accuracy is “entirely detrimental to the excitement of election night”.
His fears were exemplified by the fact that the exit poll carried out by Ipsos UK on July 5 for Sky News, the BBC and ITV News predicted with 99.5% accuracy the number of seats Keir Starmer's Labour Party would win. The party won 412 seats, just two more than what the poll had estimated.
Even French exit pollsters got it right when they predicted a shock win for the leftwing New Popular Front (NFP) coalition. Unfortunately, this was not the case in India where almost all exit polls completely loused up this year.
The worst part was that the failure of predictive analytics was sought to be masked by explaining it away with apodictic certainty. Thus, when the BJP lost its parliamentary majority, pollsters unleashed a flurry of ex-post explanations to draw a veil over their irredeemably flawed pre-poll predictions.
It was much later that the significant drop in BJP's seats was attributed to joblessness, rising prices, surging inequality, and the party's polarising campaign.
But clear indicators for a reduced BJP vote share were already there in some pre-poll surveys, especially the one done by CSDS-Lokniti, which found that unemployment and price rise are “the key concerns of nearly half the electorate”.
The survey had warned that voters “were upset about economic distress”, and “those who were not convinced that achchhe din (good days) have come did not show the same enthusiasm in voting for the BJP”.
Yet, even the CSDS — the accuracy of whose predictions is taken for granted by many in India — said that “the BJP is leading not because of its economic performance but despite the lack of it”. No explanation was offered as to why this was the case. In fact, the CSDS went on to predict that BJP's vote share will reach 40% which (for the CSDS) showed that “the BJP seems to have held on its core bastions and warded off the challenge elsewhere”.
There was, of course, a caveat: “The margin of error for this calculation of vote shares is ±3.08.” The largeness of this margin suggested that the CSDS was not very confident of its forecast.
We now know that BJP's vote share in 2024 fell by 0.84% to 36.56% instead of going up by 2.6% to touch 40% as predicted. As a result, the BJP lost 63 seats.
How wildly the CSDS estimates could have fluctuated within the range of its error margin can be understood by asking: If the BJP lost 63 seats with a 0.84% decrease in vote share, how many seats would it have gained if the vote share had increased by 2.6%? Would the BJP have crossed 400 seats if the margin of error had been +3.08%?
Although the CSDS did not mention any specific seat number, a 40% vote share meant a huge haul of seats for the BJP on its own which was impossible statistically or otherwise.
Nonetheless, most predictions went wrong because several significant ground realities were ignored. For instance, one of CSDS's pre-poll surveys categorically stated that “contrary to emerging perceptions, the idea of religious coexistence and tolerance holds its ground firmly” because 87% of Indian Muslims and 77% of Indian Hindus believe that “India belongs to citizens of all religions equally, not just Hindus”.
And, days before the consecration ceremony of the Ram Temple, several Muslim residents of Ayodhya had said: “We do not fear the local Hindus. They live with us in peace and harmony, but outsiders may cause trouble.”
This was backed by the fact that only about 22% of the CSDS respondents mentioned the construction of the Ram Mandir as the "most liked action" of the government. Even the Pew Research Center found that religion is not what Indians are focused on.
In a March 2024 survey titled ‘What can improve democracy?’, over 30,000 people in 24 countries were asked: “What do you think would help improve the way democracy in your country is working?”
For Indians, the top priorities were: Economic reforms, better policies and legislation, better politicians, government reforms, and better citizens who care more about others. Religion — spiritual or politicised — was nowhere in the reckoning.
These findings were available well before the elections. If they were not relied upon it is because most psephologists appeared to be living in the narrative bubble created by a large section of the local and foreign media wherein it was portrayed that the whole of India had, in all but name, become an Islamophobic Hindu state where Muslims felt “orphaned”.
The truth is that anti-Muslim attacks were confined to only a few parts of India, especially in the north, and mostly committed by hired gangs, not local Hindus. The election results would have been totally different if local Hindus had been involved.
Needless to say, India cannot become a Hindu theocracy without the active participation of the Hindus. And the election result was so decisively in favour of a pluralistic society that Rashtriya Swayamsevak Sangh (RSS) chief Mohan Bhagwat was forced to criticise the manner in which “untruths were spread with the use of technology” to polarise the electorate. He asked, “Is this the way knowledge should be used? How will the country operate like this?”
Despite their belatedness, the questions are valid because India's historical greatness lies in the fact that its religious and cultural heterogeneity has always come in the way of attempts to homogenise the ethnicities, religions, castes, theistic schools of thought and atheistic philosophies under one identity.
Hence it is impossible to enforce, in the name of uniformity, an insular concept of ethnocentrism that is acceptable to all Indians.
This, in all its simplicity, was the verdict of the people of India in 2024 which most pollsters found so difficult to predict.
A Faizur Rahman is the Secretary-General of the Islamic Forum for the Promotion of Moderate Thought. Email: themoderates2020@gmail.com, Twitter: @FaizEngineer
.jpg)
