Gujarat Elections: Exit polls give BJP an edge, but past goof ups, loopholes in methodology don't inspire confidence

If exit polls in Gujarat are mystifying and Indian pollsters are groping in the dark for a peg to hang their interpretations on, it's not much different even in NRI enclaves, which also see equal amounts of confusion. For some reason, the elongated voting-counting window has taken the tension out of prediction.

The large Gujarati-Sindhi population that lives in the Middle East isn't particularly bothered about whether the BJP or Congress will win in Gujarat, although the general feeling is that the former will sneak past. But there will not be much to celebrate either way, with exit polls suggesting the BJP's seat tally will drop from its current 115 seats, while the Congress, by adding to its current tally of 61, would have claimed a significant moral victory.

Representational image. PTI

Representational image. PTI

But what if the BJP loses and Congress actually wins? Can it even happen? And why not, for Indian exit polls are profound in their packaging, with suitably impressive acronyms and even more amazingly institutional backings, but so off the wall that the arrows often overshoot or fall short of the target.

Back up a bit and see how 3D the range has been over the past few elections, with all these fancy pronouncements that pepper the run-up to results day. In Uttar Pradesh elections in March, no one predicted a BJP landslide of 325 seats out of 403, and almost all pollsters were over 100 seats behind in their most sunny assessments.

And even if you're off your radar and are nowhere near the facts, you don't have to apologise. You just act more profound and wipe the egg off your face. Add large doses of clever sounding phrases like "voter apathy", "anti-incumbency" "pro-incumbency", "youth impact" and "female turnout", dress it up in technical finery, and no poll group is ever asked why they got things so horribly wrong in every election.

It is a sobering thought that in 2014, nobody even came close to predicting the eclipse of the almighty Congress, which was left with a mere 44 seats in the House. Later on, after the dust settled, there was much pretence of how this collapse was on the cards. Yeah, right!

In 2015, the JD(U)-RJD alliance made a dog's breakfast of the pollsters in Bihar, when the alliance creamed the BJP and by a handsome majority though all exit polls predicted a BJP win.

One can go on and on, but why do they get it so very wrong? Is it possible that exit polls like those 100-people surveys are arbitrary and do not represent the whole spectrum of the vote? It also depends on the time of the day the poll was taken, how much deceit has gone into it, how valid were the poll's markers and the depth and relevance of the questions. Into this mix is also the human factor: Do voters lie? Are the ones asking questions really listening and are they trained for the job? Furthermore, is the tabulation accurate or simply a bit of guesswork? Also, if you limit your exit strategy to a vote bank or a denomination of people or even a religious section, you are certain to get it wrong.

Who is the weakest link in the drivel that is then shoveled to the public? Take your pick. Including fudging the data to hedge your bets.

And it's not just an Indian flaw. In the US, they went crimson with embarrassment when Donald Trump kicked the apple cart.
Despite all the scientific claptrap, the mood of the people is fluid and amorphous and difficult to ascertain unless there is an open revolution. People are more private and cagey now and don't like to share confidences, which includes saying who they voted for. It goes against the secret ballot concept.

And the pollsters don't ask the voters in a more relaxed surrounding, after they have all cooled off, but they are accosted right outside the polling station, after they have just stood in a lengthy queue and exercised their franchise. This is an intrusion and they are more likely to fib now.

There is also the sycophancy element with polls being twisted and tweaked to support a specific media's favourites because nobody wants to be the harbinger of bad news. Candidates might also be feeding the pollsters or hiring them to add to the cacophony.

Remember how poorly pollsters did in Britain during the Brexit vote, with a near unanimous slew of them saying 'Remain' would win in the end?

No pollster has ever been sacked for getting it totally wrong. Like with Gujarat, the fact is that no one knows the result and all we do is flay in the wind.

That, in itself, is not good news for the BJP.

Firstpost is now on WhatsApp. For the latest analysis, commentary and news updates, sign up for our WhatsApp services. Just go to and hit the Subscribe button.

Updated Date: Dec 16, 2017 17:11:00 IST

Also See