Op-Ed: Why We Should Ignore Predictions on Polls?

Pre-election ‘surveys’ are the flavour of the season. And true to form, we have worked ourselves into a tizzy over the findings of two back to back surveys – the India Today ‘Political Stock Exchange’ survey and ABP C-Voter ‘Desh ka Mood’ poll – coming within a week of each other. While the India Today survey wisely refrained from talking seats, ABP stuck its neck out to predict that the BJP would win 13 seats in the state in the Lok Sabha polls, but with a huge rider: ‘if’ PM Modi contests from Puri! What if he doesn’t? The survey did not bother answering this question. Had a layman said BJP would win 13 seats, he would have been immediately dismissed as having gone crazy or being a ‘Bhakt’. But here we are, animatedly discussing the same ‘findings’ of a survey long after they were made public!

I have been covering elections since 1991 and I can’t remember one occasion when opinion polls have hit bull’s eye. Even exit polls, supposed to be an accurate reflection of the actual voting numbers, have proved wide off the mark. In more recent times, no opinion/exit polls told us that Vajpayee would be voted out in 2004, the Congress would return to power with a greater majority in 2009 or the BJP will emerge as the first party since 1984 to win a majority on its own in 2014. On the rare occasions when these surveys have predicted numbers that have turned out close to the actual results, they have been merely coincidental than the result of painstaking, scientific psephology. A case in point is Chanakya, which burst into the scene with its fairly accurate prediction of the 2014 general elections and then went on to eat crow in every election since then.

And yet, we simply can’t stop obsessing over the surveys. Even the standard disclaimer “if elections were held today” has not stopped us from endlessly debating the findings in street corner chai/paan shops, newspaper columns and TV studios. Assuming that the elections will be held when they are due, there are over six months still left for the Lok Sabha polls. And a lot can change between now and election time. We are unable to (or refuse to) see that in adding the rider “if elections were held today”, the agencies doing the surveys are essentially guarding themselves against the inevitable charge of hoodwinking the public.

I can think of only one reason for our obsession with this ‘game’. Every Indian with even a cursory interest in politics and plenty of time on hands is a closet psephologist. S/he can give the so-called psephologists a run for their money and reel off constituency-wise break up of who will lose to whom in which seat and why. So it is only natural for them to get animated when they see the findings of a ‘survey’, whether they are in conformity with or at complete variance with his/her own assessment.

In the case of the media, it provides them a heaven sent opportunity to fill up space/time, which otherwise would have taken a lot of hard work, secure in the knowledge that the reader/viewer would just lap it up. For the individual media person, it is a time to show his/her profound knowledge and understanding of politics. And I can vouch for the fact that most people do believe scribes know better than them about these things. “You must be knowing better. How many seats is the BJD going to win this time?” is the kind of question you are most likely to be asked if you are a journalist. If you tell them that you don’t know, they would look at you in disbelief and think you must be joking.

Time was when the opinions of journalists, especially the senior ones among them, carried great weight, both with the public and the politicians. But in my experience, I have found their assessment to be as untrustworthy as the dime-a-dozen opinion polls that gobble up so much media and public space/time these days. Here I can’t resist the temptation of narrating two anecdotes that I think would illustrate the point I am trying to make better. The year was 1998. The newly born BJD and the BJP were fighting the general elections together with the Congress, which was in power in the state, ranged against them. Like other newspapers, the Delhi based publication for which I worked at the time was conducting a nationwide opinion poll. Unlike others, however, they were serious about it and had sent two reporters from Delhi to assist me in doing the survey. Together, we toured 14 of the 21 Lok Sabha constituencies extensively over a period of nearly two weeks and came up with a report that gave 15 seats to the newly-formed alliance with the Kendrapara seat marked 50:50). A veteran journalist of the state, feared and respected in equal measure, exploded at me. “Where the hell do you think a party that was only formed yesterday and another that no one knows about will get the seats from? Mark my word, they would be lucky to get five.” When the results came (at Soochana Bhavan in Bhubaneswar), I could barely suppress a chuckle while the veteran was red on face. It was 16 in favour of the BJD-BJP! [This was the only time I allowed myself to be drawn into the numbers game and that too only because I didn’t really have a choice.]

During the campaign for the same election, I and my godd friend Dillip Satpathy had gone on day’s whistle-stop tour of the Puri Lok Sabha constituency. On our way back late in the evening, we decided to meet a senior journalist based in the holy town. He fished out a piece of paper and a pen and went on to explain to us – through some complex mathematical diagrams and calculations – why Braja Kishore Tripathy would lose ‘by at least one lakh votes’. Tripathy won the election with a margin of well over a lakh!

My own take on the issue has always been that journalists should never get into the business of predicting elections, numbers in particular. For, Indian elections are complex affairs, influenced by factors which are hard to decipher even for a local, let alone a journalist from another place who covers a constituency in a matter of a few hours or – worse still – makes his assessment sitting hundreds of kilometers away.

At best, journos should restrict themselves to outlining the broad trends, if any, and flagging off the issues that they think would influence the outcome. Anything more than that is plain bunkum!

(DISCLAIMER: This is an opinion piece. The views expressed are author’s own and have nothing to do with OTV’s charter or views. OTV does not assume any responsibility or liability for the same)