Hits and misses of exit polls in predicting elections

India News-jaan Ki Baat predicted the exit polls of the five Assembly elections correctly.

The results of the five state Assembly elections concluded earlier this week gave a clear mandate to the BJP in four out of the five states that went to polls in February and March, but the exit polls which predicted a win for the BJP in Uttar Pradesh and the Aam Aadmi Party (AAP) in Punjab, have failed to gauge the voters’ pulse in the smaller states like Goa and Uttarakhand. Most of the pollsters predicted a hung Assembly in both these states. Exit polls often have their fair share of hits and misses and The Sunday Guardian tracks the recent polls and how these exit polls have fared.

The exit polls for the recently concluded elections in Uttar Pradesh, Punjab, Goa, Uttarakhand and Manipur were correct to predict that there was a wave in favour of the AAP in Punjab and that the BJP was comfortably getting a majority in Uttar Pradesh, though with a reduced number of seats. But for Goa and Uttarakhand, pollsters differed in giving a clear majority to any party. Many pollsters predicted that both these states were to have a hung Assembly as the electors, according to them, were not able to give a clear majority to any party (in this case either the Congress or the BJP).

However, when the results of these Assembly elections started to pour in on Thursday morning, the prediction of the pollsters over Uttar Pradesh and Punjab brought cheer to their faces, but their prediction for Uttarakhand and Goa slightly brushed away from the reality. The electors of both these states gave a clear majority to the BJP. Pollster like P-marq, Veto, Axis My India, Ground Zero, Designboxed, predicted that Goa would be a hung Assembly, giving the BJP an average number of seats ranging between 15-18, while also giving the Congress an equal number of seats. However, the final results for the Goa Assembly which has 40 seats, gave the BJP an upper edge with 20 seats and the Congress 11.

Even for Uttarakhand, the above agencies predicted a hung Assembly. Many among the above-mentioned pollsters tried to stretch their prediction for this hill state suggesting that the BJP was slightly ahead of the Congress, but they were not sure if the saffron party would be able to pull off a simple majority on their own. Some suggested in their exit polls that the BJP in this state could be restricted between 25-29 seats. However, the actual results were different from the predictions of exit polls, where the BJP secured 47 of the 70 Assembly seats in the state, leaving behind the Congress with just 19 seats.

It is not the first time that exit polls have erred in their prediction of the results of elections. One of the biggest setbacks for pollsters came when they got the whole picture of West Bengal Assembly elections held last year completely wrong. Though most pollsters were giving an upper edge to the Trinamool Congress, they had predicted that the BJP was on an upswing in Bengal. However, when the results for Bengal were declared on 2 May last year, their prediction fell flat. The pollsters predicted that the BJP in Bengal was getting somewhere between 110 to 140 seats, and the TMC giving a close fight to the BJP with seats between 130-160 in the 294-seat Assembly. The results for Bengal came out to be completely opposite and the BJP could not even cross the double-digit mark in the state, while the TMC scored above 200 seats in Bengal.

Similarly, even for the 2017 Uttar Pradesh and Punjab elections, many pollsters got their prediction for the states wrong. Most of them predicted that there was a tough fight between the then ruling Samajwadi Party and the BJP in Uttar Pradesh and that the BJP would be unable to cross the halfway mark of 202 seats. These pollsters said that the BJP would get somewhere between 160-190 seats, while the SP could secure 120-140 seats, predicting a hung assembly.

However, the actual election result for Uttar Pradesh in 2017 gave BJP more than 300 seats in the 403-member assembly. Even for Punjab, many projected a AAP sweep in 2017 assembly elections, but the Congress was able to get much more than the majority required to form government in the Punjab Assembly.

The Bihar Assembly elections of 2020 was also a similar story where pollsters erred in their prediction on the pulse of the voter and had predicted a comfortable majority for the RJD led alliance in the state. Pollsters predicted that in Bihar 2020 assembly elections the RJD led alliance were all set to get somewhere between 120160 seats in the 243-member assembly and form the next government in Bihar, while giving the Jdu-led NDA alliance 55 to 91 seats.

But the results for Bihar Assembly elections turned out to be completely different, with the Jdu-led alliance securing 126 seats and the RJD alliance securing 91 seats. However, it’s not always that pollsters get their predictions on elections wrong. Many a time they do fairly well in their prediction. The most talked about “correct” exit poll by pollsters in the appropriate prediction of the 2014 Lok Sabha election results where all exit polls showed that the BJP was on its way to get its highest ever tally in terms of number of seats. Today’s Chanakya grabbed eyeballs when they predicted that the NDA was on its way to win 340 seats in the Lok Sabha and the NDA had actually won 334 seats in its final tally when the results came out in May 2014. In the same election, most pollsters predicted BJP to get between 230 to 300 seats.

Even the 2019 Lok Sabha elections were correctly predicted by most pollsters in the country. Exit polls are carried out after doing a survey with electors after they have casted their vote and is, therefore, often than not believed to be a more accurate prediction on the outcome of any election and is therefore more reliable than an opinion poll. However, pollsters conducting exit polls say that their predictions do miss sometimes because of various reasons including people not giving out the exact details of whom they voted for. They also say that exit polls only “give a sense of direction of the wind” and may not always be true to the numbers as they are most often extrapolated data.

However, the India Newsjaan Ki Baat predicted the exit polls of the five Assembly elections correctly which turned out to be almost similar to the actual results that were declared on Thursday. The India News-jaan ki Baat exit poll predicted that the BJP+ in Uttar Pradesh would get 260 seats, while the SP will end up getting about 135 seats, with a vote share of 40-43%. The actual result on Thursday also showed a similar result where the BJP+ received 267 seats, with a vote share of 41%. This pollster also predicted Punjab, Uttarakhand, Goa and Manipur correctly, where AAP had won Punjab and the rest was bagged by the BJP.

Exit polls in UP, Punjab, Goa, Uttarakhand and Manipur were correct to predict that there was a wave in favour of AAP in Punjab and that the BJP was comfortably getting a majority in UP. But for Goa and Uttarakhand, pollsters differed in giving a clear majority to any party.