Election results jolt for pollsters, buck projections
India's 2024 election results defy exit poll projections, with BJP falling short of majority. Pollsters admit errors in predicting NDA's success.
The results of India’s 2024 general elections declared on Tuesday were in sharp contrast to the exit poll projections released on the final day of voting last week, prompting at least some analysts to admit that they needed to look at what went wrong.
As of 11.50pm on Tuesday, the BJP had won or was leading in 240 seats, far short of the 273 needed for a majority in the 543-seat Lok Sabha. The larger alliance it leads was on track to secure 293 seats. The opposition Congress party had won or was leading in 99 seats, with the broader INDIA bloc was on track to secure 233 seats, according to the Election Commission of India.
On June 4, most exit poll agencies predicted a thumping majority for Prime Minister Narendra Modi’s party. For instance, the India Today-Axis My India poll forecast 361-401 seats for the NDA and just 131-166 for the INDIA bloc. Axis My India chairman Pradeep Gupta broke down on India Today television as the results trickled in Tuesday, acknowledging that his agency needed to examine where it went wrong.
“In Uttar Pradesh, we predicted a lower range of about 67 seats, but at this point, the NDA has secured 38 seats,” Gupta said. Similar shortfalls, compared to projections, were in West Bengal and in Maharashtra.
Gupta said the exit polls failed to capture a last-minute shift away from the ruling party among Dalit voters, especially in Uttar Pradesh, West Bengal and Maharashtra. Issues like the BJP’s perceived attempts to alter India’s constitutional guarantees of affirmative action may have driven this change, he said.
Other major polling organizations, including CNX, P-Marq, Jan Ki Baat and CVoter, too overestimated the NDA’s tally.
Rahul Verma, a senior fellow at the Centre for Policy Research think tank in New Delhi, said: “They got three states completely wrong — Uttar Pradesh, Maharashtra and West Bengal. These three states are the largest states in terms of seats. Why did they get it wrong? At this point, it is hard to say. … As someone who uses polling data a lot, I don’t want to dismiss them altogether. They also got some states right.”
Gilles Verniers, another scholar at CPR, criticised the exit polls’ methodology, arguing that most prioritise large sample sizes over representative samples. “Besides, in the context of this election, the social desirability bias or the fear of voicing a contrary opinion might have affected many respondents’ answers,” Verniers said.
