In Exit Polls Prediction, A Non-Existent AAP Candidate, 7 Seats For LJP In Bihar

On May 19, the seventh and final phase of Lok Sabha elections, most of the country was glued to their TV screens. After nearly two months of a raucous election campaign, began an even more jarring exercise spanning multiple hours: the ‘exit poll’ revelation.

All exit polls predicted a thumping majority for BJP-led NDA government and a second term for Prime Minister Narendra Modi. 

There was high drama in some TV studios, while others literally rode the drama into them. Most of the exit polls announced 300+ seats for BJP-led NDA government. However, some extrapolations were just, well… extra.

To begin with, the lack of transparency in terms of the methodology used for conducting ‘exit polls’ leaves much to the imagination. Most, as has been revealed later, were post-poll surveys, those conducted a day after the polling.

On May 20, Axis-MyIndia, which had partnered with the India Today Group, deleted their seat-by-seat projections from their website. The data, however, has reappeared again, but the seat-by-seat projections are now on the India Today website. Prior to that, there had been discrepancies which were pointed out.

The bigger problem seemed to be with a disclaimer that said: “Seat by seat indicated are based purely on the popularity of the political party during exit polls and not the individual candidate and hence we cannot be held responsible for any variation of the winning or losing of the individual candidate stated in the seat by seat in our exit polls.”

The ‘exit poll’ related queries are on a webpage on their website. The second question has been answered pretty straightforwardly: “Even though 'exit poll' is the commonly-used nomenclature, technically speaking, the survey was a 'post-poll' study. The key difference between the two lies in the location and the timing of the interviews.”

The differences between the two were spelled out by a piece in a Hindustan Times story where the author also raised questions about vote share being translated into seats. The author writes: “What the exit polls are expected to have are vote share numbers. Even if they are accurate, there is no guarantee that seat share projections will be accurate.”

That aside, there were also major discrepancies in how other pollsters went about their day jobs:

• The Lok Janshakti Party is contesting 6 seats in Bihar while News 18-IPSOS exit poll gave it between 5-7 seats.
• Times Now- VMR exit poll predicted that AAP would win 2.9% vote share in Uttarakhand while the party has not announced a candidate from the state.
• It also said that the BJP candidate would win the single seat of Chandigarh despite having a lesser vote share of about 37% as compared to the Congress candidate’s over 42% vote share.
• Aside from getting the names of five constituencies from Uttarakhand wrong, Axis predicted that the Congress would win Chennai Central- a constituency where the Congress did not field its candidate.
• It also apparently got confused between the Sikkim Democratic Front and the Sikkim Krantikari Morcha, exchanging names of winners and vote shares in the single-seat state.

What pollsters aside from ABP’s Nielsen and News 24’s Neta seemed to agree on, however, was a thumping victory for the NDA at the Centre. Even after the results on May 23 though, electioneering and pollstergiri will continue.