Delhi elections 2013: How AAP has changed the polling game

New Delhi: As Delhiites press the button on the electronic voting machine today, there is at least one entity, apart from the Bharatiya Janata Party (BJP) and the Congress, which would be hoping, and praying, that the Aam Aadmi Party (AAP) comes a cropper in the Assembly election. The entity’s name: the Pollster, a moniker covering a range of agencies engaged in the business of predicting election results.

For decades now, they have at times read the voter’s mind right. On several occasions they have been egregiously wrong. Their method has been hailed scientific every time they have called the electoral dice correctly. When wide of the mark, critics have called them imposters who are in cahoots with one of the parties, or willing to tailor the results to please their paymasters – the media houses – who wish to demonstrate their political allegiance. Fortunately for the pollsters, such allegations haven’t stuck to them permanently.

But not any longer: the results of the Delhi election could damage their credibility irreparably.

This is because the challenge to psephology– the rather flatulent term to describe the science of predicting election result – has come from within: from Yogendra Yadav, an AAP leader who has been forecasting for well over a decade, and who is a ubiquitous presence on the CNN-IBN channel each time a State of India has gone to polls.

 Delhi elections 2013: How AAP has changed the polling game

How will polling data change after Delhi elections? AFP.

Yadav has been challenging the community of pollsters for more than two months now. He fired his most powerful volley only last week, through a video titled, Why elections surveys are missing that AAP is winning, which the party’s website uploaded on its website. In his characteristic soft, persuasive voice, he said some surveys seem equivalent to paid news. But then, he also added that not all polling agencies are dishonest, and even the most scrupulous of pollsters can go horribly wrong.

His tirade, though, did not end there. He asked: why do pollsters go wrong? His answer: it was because they either do not follow the scientific procedure of choosing randomly their respondents, or that their sample size doesn’t reflect the diversity of the electorate or is too small to capture the popular mood.

He then lit into the procedure of eliciting responses. It is possible, he argued, that fieldworkers ask the respondents a direct question: whom do you plan to vote? In such a scenario, the voter tends to name the party in power, in Delhi’s case, the Congress. Also, pollsters calculate swings for or against for the parties in the fray by taking into account the vote-percentage they polled in the last election. But how do you calculate the last vote-share of a party fighting the election for the first time, as is true of AAP? To put it down to zero would be skewing the picture.

Yadav went on to narrate the experience of Prannoy Roy, who was embarrassingly wrong in his prediction of the 1983 assembly election. Then the Telugu Desam Party was fighting the election for the first time, and despite a clutch of suveys claiming its leader, NT Rama Rao, was drawing crowds but not their votes, the party swept the poll. Yadav claimed in the video that this was what happening in Delhi, and the results on December 8 would see the AAP cross the majority-mark.

Call it the hyperbole of a party leader, or the psephologist’s supreme confidence in the method of his art, there is no denying Yadav has staked his reputation. Should AAP fail to reach anywhere near the result he has predicted, his reputation as pollster would be sullied. He would, for sure, become the butt of jokes among those who ply the same trade.

However, in staking his reputation, Yadav has challenged the community of pollsters on many counts. Are they all genuine, not succumbing to pressure and pelf to manufacture results? Is their method of conducting opinion polls scientific enough to pass the muster? Above all, to what extent do opinion surveys influence voters in deciding which party they should vote?

The last question is pertinent because the argument against conducting opinion polls in the election season is that sways the voter to cast his preference in favour of the party which is projected as the eventual winner. In fact, the AAP decided to out its third survey, conducted in August, because in the findings of other pollsters the party was seen as a distant third.

But what prompted the AAP to make public their internal survey was a tip-off from journalists working in TV channels. The AAP leaders were told that five channels were scheduled to beam two surveys conducted by two different agencies. In an interview to this writer, Yadav said, “This was an extraordinary thing. Nobody telecasts surveys which they haven’t sponsored, and no one gives the data of any survey to anyone who hasn’t been paid for it. It looked something like a design. The impression being created was that we were a party which could, at best, get four-six seats, that it is a vote-katua party.”

It was to counter this impression they went public with its third survey, in the process taking an unprecedented step: it uploaded on the party website the raw data, giving out not only the sample size, methodology etc, but also the answers the respondents had given to their questions. These answers are usually are numerically coded. “If I am given a raw data file, I can tell within a couple of hours whether it is fake or genuine,” he told me. In its willingness to open the raw data for scrutiny, the AAP was challenging polling agencies to do the same.

Nor did Yadav conceal that the outing of the AAP survey was a battle strategy. As he told me, “In the face of what appeared to us professionally flawed, technically erroneous information, we used what was our internal survey – because it was truthful, because we were technically on firm grounds –as a political weapon. I have no hesitation in saying that.”

This weapon could fell Yadav as well. His party’s poor showing would have others claim that no matter how honest or diligent a psephologist, he shouldn’t be interpreting survey results on joining a political party.

Should the AAP perform as well as Yadav has predicted, there would be a clamour for polling agencies to be transparent and place their raw data for the scrutiny of people in general and experts in particular. It may belatedly inspire the growing tribe of pollsters to police their fraternity, separate the wannabes from the real practitioners. Perhaps it could also induce them to evolve an index of success, and declare the number of times their survey has been proved right on every occasion they conduct an opinion poll.

Nevertheless, the professional risk that Yadav has taken is enormous. But then this is typically AAP style – whether or not it succeeds, the party has shown it is possible to fight elections on low budget, without musclemen, without experienced politicians, by adopting innovative methods, and largely eschewing the temptation to appeal for votes on the basis of caste and religion. To this list, Yadav has added one more item, which is a question to media houses: shouldn’t we ask polling agencies to provide a better account of themselves?

The author is a Delhi-based journalist, and can be reached at

Updated Date: Dec 04, 2013 16:45:22 IST