Maharashtra, Jharkhand Assembly Elections: Pollsters Don’t Know Who’s Winning

A note of caution, exit polls have gone wrong in the past, including the general elections and Maharashtra. Having said that, exit polls generally are a better barometer of trends and results than opinion polls. It's quite natural because the survey team talks to voters after they have exercised their franchise. 

Opinion polls take place a few days before the day of polling and a voters’ mind could change in between. Over the years, exit pollsters have struggled to estimate the correct tally of seats.

Eight agencies have released their exit poll numbers. Five of them (P-Marq, People’s Pulse, Matrize, JVC and Chanakya) are predicting a Mahayuti win (based on the midpoint of their projections), one (Electoral Edge) is predicting an MVA victory, while two (Lokshahi Marathi Rudra and Dainik Bhaskar) predict a hung house. Most of the pollsters predict a Mahayuti victory. 

Two of the biggest agencies, Axis and C-Voter, have not released their numbers, citing very close contests and time required to collate and decipher their data. In a channel debate, Yashwant Deshmuk, founder of C-Voter, stated that there were 100, too-close-to-call seats, keeping the suspense alive. 

While MVA won 30 out of 48 Lok Sabha seats in first elections after political realignments in Maharashtra, Mahayuti could win only 17 out of 48. However, in terms of vote share, the contest was very close with MVA leading by less than 1%. In terms of assembly segments, the contest was much closer with MVA leading in 153, and Mahayuti in 126 seats, with 145 being the majority mark. 

The estimates are also quite wide ranging from 118 (low; Electoral Edge) to 195 (high; People’s Pulse) for Mahayuti. Now that’s quite a big difference of 77 seats (one-fourth of assembly strength). For MVA the range is quite big too; from 85–150. Others who are expected to play a key role in government formation as winners or vote cutters are expected to bag anywhere between two to 25 seats.

Eight agencies have released their exit poll numbers. Four of them (Matrize, JVC, People’s Pulse and Chanakya) are predicting a BJP-led NDA win (based on the midpoint of their projections), three (Electoral Edge, PMarq, Axis) are predicting a JMM-led INDIA bloc victory, while one (Dainik Bhaskar) is predicting a hung house. Majority pollsters predict an NDA victory, while the pollster who claims to have the best record—tarnished by general elections and wrong predictions during Haryana assembly elections—predicts an INDIA victory. 

C-Voter says 20 seats are too close to call, giving NDA 34-26 edge over INDIA. While NDA won 9 out of 14 Lok Sabha seats in general elections, INDIA won 5 out of 14, sweeping all the ST reserved seats due to sympathy factor generated after CM Hemant Soren’s arrest. NDA was leading in 52 out of 81 assembly leads in general elections. Both the alliances have been bolstered since the last Vidhan Sabha elections. JVM has merged with BJP, while it has finalised a seat sharing deal with AJSU. CPI-ML has joined the INDIA bloc alliance. 

The estimates are also quite wide ranging from 17 (low; Axis) to 53 (high; People’s Pulse) for NDA. Now that’s quite a big difference of 36 seats (almost half of assembly strength). For INDIA, also, the range is quite big too from 25-59. Others, including Jairam Mahato—who has emerged as a key player enjoying support of youth, are expected to play a key role in government formation as winners or vote cutters, with anywhere between one to nine seats expected to come to them.

Many exit poll agencies have sprung up in the last few years. Very little is known about their ownership and track record. Many times, news channels which carry out the surveys exploit the newcomers to do the polls for free or at significantly lesser price. This impacts the quality of the survey. Sampling is the key to getting good results, but is very difficult to perfect. A solid data collection, as well as methodology is key for higher accuracy. Most polls arrive at vote shares and then use a conversion ratio, based on historical data, to arrive at seats which leads to difference in results. 

Not only in India, pollsters across the world are facing a test of credibility after getting many calls wrong; Brexit, US elections 2020, even US elections 2024, were not as close as predicted.

In India, all pollsters got it wrong in Lok Sabha, while all again got it wrong in Haryana. Many channels have stopped showing exit polls. The industry has gone back to 2004 levels, when all pollsters predicted Atal Bihari Vajpayee’s victory, while he was handed a shocking defeat. 

Despite this, none of them seem to have learnt lessons and changed methodology. When asked about this, a top pollster said they have tightened the monitoring process. A lot more needs to be done. It's redemption time for most pollsters in Maharashtra and Jharkhand. If they get it wrong again, then it could be curtains for industry. 

Amitabh Tiwari is a political commentator, strategist, and consultant advising political parties and leaders. He was previously a corporate and investment banker.

Disclaimer: The views expressed here are those of the author, and do not necessarily represent the views of NDTV Profit or its editorial team.