India’s leading news channels and their poll pundits had on June 1 predicted a clean sweep for the Bharatiya Janata Party-led National Democratic Alliance (NDA) in their exit polls. Some had given the NDA more than 400 seats out of 543, the target PM Narendra Modi had set for his alliance with the slogan 'ab ki baar 400 paar'.

On June 4, defying all predictions, the INDIA bloc - formed by Congress and several regional parties a year ago - emerged as a strong Opposition against NDA by winning 233 seats. This has left NDA with a narrow majority, having secured 291 seats, just 19 more than the majority mark of 272.

So what did the news channels predict?

India Today-MyAxisIndia Exit Poll: NDA 361-401 seats, INDIA Bloc 131- 166

ABP-CVoter exit poll: NDA 353 to 383. INDIA bloc 152 to 182

India TV-CNX Exit Polls: NDA 371-401; INDIA Bloc 109 to 139

News 24-Today's Chanakya: NDA 400; INDIA Bloc 104

After the misfires, some pollsters sought to defend themselves by claiming that they were at least “directionally correct” by predicting a third term for the NDA.

These claims are misplaced, say critics. “Poll agencies had predicted an increase in the BJP tally this time as the party settled with just 240, compared to 303 in 2019. No polls predicted that Congress could win 99 seats or the INDIA Bloc could get more seats than NDA in the BJP’s bastion Uttar Pradesh. Of course, exit polls were directionally wrong too,” remarked an advertiser.

https://results.eci.gov.in/PcResultGenJune2024/index.htm

Flawed approach?

Questions are being raised as to how all major exit polls conducted by leading news channels in collaboration with renowned pollsters have gone off the mark again.

Many industry veterans alleged exit polls of unreliability, unfairness, improper sample size or data collection, and being sponsored or politically motivated. Some analysts even called them “estimated polls”.

The exit poll results favouring BJP heavily also led to excitement in the stock market on Monday, leaving netizens even demanding that poll agencies apologize to people and disclose their investments in the stock market.

“Some pollsters were simultaneously doing surveys for political parties besides doing exit polls for the news channels,” remarked an X user and senior journalist Mahesh Langa. He also said that channels should also come out with such disclosures.

Yogendra Yadav, political activist and former psephologist, said on a TV channel, “All agencies which conduct exit or opinion polls should declare their affiliation with political parties.”

Yadav, whose prediction (Opposition alliance 205 to 235 seats; NDA 275 to 305), was the closest, said that he had stopped poll math long back and made his predictions public only when he found that most news channels were giving 400 plus seats to NDA contrary to the ground reality.

e4m reached out to TV channels to understand the reasons behind the erred numbers.

Amit Sinha, Chief Strategy Officer at INDIA TV, said the channel was unhappy with its research agency for such a bad show.

Sinha told e4m, “News channels' role is to be informative to its audience with all possible data and insights related to knowledge across topics whereas Exit Poll is the practice used to be followed by broadcasts. Independent research agencies do the analysis based on voters' behaviour and predict the estimated seats, voting percentage etc. These are 3rd-party information leveraged by broadcasters which are followed to trends voters talked about. As a broadcaster, we are unhappy with our research agency for such a bad show.”

Reasons aplenty

Exit polls are conducted immediately after voting ends by either physical interviews outside the polling booth or telephonic interviews.

The manner of conducting exit polls in the country is varied, its accuracy depends on factors like structured questionnaires, in-person or telephonic interviews of random people for fairness, sample size, demographic representation and data compilation.

“It is increasingly becoming estimated polls predicting seats instead of real exit polls as many show only the number of seats and not vote share or methodological details,” a professor of political science said.

Media veterans noted that pollsters often face challenges, leading to inaccuracies in their predictions.

“Pollsters often rely on sampling techniques to gauge public opinion. However, if the sample is not representative of the entire population or if certain demographics are underrepresented, it can lead to biased results. For instance, if you survey 5,000 voters in urban areas, you can't extrapolate the results to 20 lakh people in the parliamentary constituency which may consist of rural, forest or tribal areas,” a brand leader said.

“Besides, differences in survey methodologies, such as sampling techniques, questionnaire design, and data collection methods, can impact the accuracy of poll predictions. Variations in these methods across different polling agencies can lead to discrepancies in results,” he added.

Voter dynamics and secretive behaviour also play a big role in such surveys. “Indian elections are complex, with diverse regional, caste, and religious dynamics influencing voter behaviour. Pollsters may struggle in accurately capturing these nuances. Besides, voters may not always reveal their true preferences during surveys due to fear of judgment or social profiling. This can result in inaccurate responses and skew poll results,” another expert stated.

On the other hand, Indian elections are dynamic, and voter preferences can change rapidly in the final days leading up to the election. Polls conducted earlier in the campaign period may fail to capture these late swings in public opinion, industry experts noted.

“Pollsters must continuously refine their methodologies and adapt to the evolving political landscape to improve the accuracy of their predictions,” an executive said.

Patchy Record

According to industry veterans, it was in 1996 that satellite TV made exit polls popular when state-owned broadcaster Doordarshan asked the Centre for the Study of Developing Societies to conduct nationwide exit polls.

Exit polls have had a patchy record in India. Below are the four instances where exit polls proved inaccurate.

West Bengal Assembly Elections 2021: Despite significant private exit polls indicating BJP's lead, the actual results saw Mamata Banerjee's Trinamool Congress securing a substantial victory with 213 out of 294 seats, leaving the BJP with only 77 seats.

Bihar Assembly Elections 2020 & 2015: Predictions for the Bihar 2020 Assembly Elections heavily favoured the Tejashwi Yadav-led Mahagatbandhan. But the BJP-JD(U) alliance emerged victorious. In 2015, exit polls faltered in foreseeing the resounding victory of the RJD, JD(U), and Congress coalition which swept 178 out of 243 seats.

Uttar Pradesh Assembly Polls 2017: Initial projections for the 2017 UP Assembly elections hinted at a hung assembly, albeit with the BJP as the largest party. Contrary to these expectations, the BJP clinched a landslide victory, securing 312 out of 403 seats in the state.

2004 Lok Sabha Polls: Riding on Prime Minister Atal Bihari Vajpayee's unprecedented popularity, most exit polls predicted a comfortable win for the BJP-led NDA, with estimates ranging from 240 to 275 seats. However, the NDA secured only 187 seats, while the Congress and allies clinched 216 and later formed the government.

For more updates, be socially connected with us on
Instagram, LinkedIn, Twitter, Facebook Youtube & Google News