Reply To:
Name - Reply Comment
Narendra Modi takes oath for the third consecutive term as Prime Minister of India
The biggest casualty was probably the credibility of exit polls, all of which predicted more than 350 seats for the NDA
Observers and Opposition party activists in India heaved a sigh of relief that the election had finally saved democracy and pluralism in the world’s largest democracy. The biggest casualties, though, other than Prime Minister Modi’s hubris, are India’s pollsters whose predictions of a gargantuan Modi sweep fell way short of actual results; this wiped out US$ 400 billion in the stock market when the reality hit.
A historic feat
Indomitable Narendra Modi won and was sworn in last week for the third consecutive term as the prime minister; a historic feat since Jawaharlal Nehru in 1962. But not with the 400 seats he has asked for, which would have enabled the BJP to change the Indian Constitution.
The Bharatiya Janata Party won 240 seats, well below its previous tally of 303 seats in the last Parliamentary poll in 2019, with its allies in the National Democratic Alliance winning 293 seats, more than needed to secure the simple majority of 272 in the Lok Sabha, the Lower house of India’s Parliament. That was still a serious comedown from the mammoth 353 seats the NDA won in 2019.
INDIA Alliance, led by the Indian Congress Party and comprising patchwork of regional parties, won 232 seats, defying all predictions.
The biggest casualty was probably the credibility of exit polls, all of which predicted more than 350 seats for the NDA. For instance, India Today-Axis My India projected 361-401 seats for the NDA and 149 seats for the INDIA bloc, almost 80 seats short of the actual tally.
ABP C Voter survey gave the NDA 368 seats while the INDIA bloc was projected to win 167 seats. Tv9 Bharatvarsh projected 342 seats for the NDA, and the Times Now-ETG survey predicted that the NDA would get 358 seats, and Today’s Chankya gave the NDA a mammoth 400 seats.
As the reality hit and the credibility was in tatters, Pradeep Gupta, the founder of polling firm Axis My India, was seen crying on TV. An anchor was heard consoling him that at least he got the winner right.
Low-income and Dalit voters
Many theories have floated about why the exit polls went so wrong, including the pollsters misreading the sentiments of the low-income and Dalit voters in three big states, Uttar Pradesh, Maharashtra and Bengal, which sent 175 members to Parliament.
Exit polls are largely for media consumption, considering they are conducted during the voting day. The election law regulates their publication, they cannot be used to manipulate the elections – though the Opposition leaders have accused the pollsters of a “stock market scam”, which has whipped out US$ hundreds of billions since the false euphoria created by the exit polls dissipated.
Sri Lankan readers who might have been bamboozled by some of the local opinion polls might have empathy with their Indian peers.
Consider some of the polled results:
Institute of Health Policy’s Sri Lanka Opinion Tracker Survey for April 2024 shows Sajith Premadasa and JVP Leader Anura Kumara Dissanayake at neck to neck with 39%, while Ranil Wickremesinghe at 13%.
Intriguingly, when this survey was first launched Dissanayake was projected to be a one horse race, with all other contenders in single digits, though the numbers have mellowed since then.
In another poll, three out of four Sri Lankan adults (75%) said the country was heading in the wrong direction in April 2024, while almost none (3%) said it was on the right track.
In another poll, Mode of the Nation of the Verite Research, the government’s approval rate fell to 7% in February from 10% a year ago.
Economic outlook
Also, 91% of the polled, when asked about the economic conditions of the country, said it was poor. This is an increase in the negative rating from 73% in June 2023 and 81% in October 2023. Similarly, the percentage that thinks the economic outlook of the country is getting worse has increased from 66% in October 2023 to 90% in February 2024.
Now, there is a hitch. In February 2023, the rupee was still in free fall, trading at 368 rupees a dollar. The country was still reeling with fuel rationing, and the inflation was at 50%.
It is mind-boggling how an average nonpartisan respondent would perceive conditions a year ago as much better than they are now, though much needs to be done.
This leads to the central question. It all boils down to the sample and weightage of the sample.
A cursory glance at these opinion poll numbers gives the impression that the sample is picked right from the picket lines of the JVP demonstrations. However, the HIS says its survey estimates are based on 17,134 interviews conducted from 1 October 2021–19 May 2024, including 444 interviews conducted in April 2024.
When certain things get too good to be true, they could also become fishy. A Sri Lankan, with their feet firmly on the ground, could make some very basic rational judgments. One of these would be that the JVP, as a single party, winning anything closer to 20 per cent of the popular vote in an election is a non-event. That is only an eventuality in the cloud cuckoo land of the pollsters.
Considering the default Sri Lankan sense of nagging and winning, one might agree that most Sri Lankans might not acknowledge the present, though how hard it, is an improvement from a year back. But, at the same time, Sri Lankan voters are conservative and risk aversive to bet on the JVP for their economic survival.
There is another point. Exit polls may, by and large be innocuous and cannot manipulate the election, but the manipulated opinion polls could do so by swinging largely undecided voters. The JVP’s latest bout of importance has less to do with growing public support, but rather because some opinion polls, such as the one by IHP, have placed it in an imaginary lead, which saw a bevy of foreign diplomats making a beeline to the JVP headquarters.
At one point, though, the reality hits. Sri Lankan pollsters should learn from Indian peers.
Follow @RangaJayasuriya on X