Exit polls: They’re often hit and miss affairs

What the exit polls have revealed recently simply makes a pickle of any sense one tries to derive out of the experiences of the past two years. And they just seem to bolster and magnify people’s perceptions formed over the past six months.

After seeing the findings, the picture that emerges is just a clash between two time frames.

Everything was placid on the Indian political firmament in 2012. Uttar Pradesh was won, expectedly, by the Samajwadi Party, with the BJP struggling so badly that it forgot to criticise the Congress. Punjab, not so expectedly, was retained by the Akali-BJP. The Congress won back Uttarakhand, Himachal Pradesh and Karnataka, and there are no surprises here. Narendra Modi retained Gujarat with a reduced tally, but who said he wouldn’t? Anna Hazare’s movement peaked before Arvind Kejriwal broke away.

Against this picture, what does one make of the exit poll forecast that the BJP will thunder away with 55 of the 80 seats in UP? Doesn’t it go against received wisdom?

There are other reasons also to be sceptical. Writing of opinion polls, a former chief election commissioner wrote recently in Hindustan Times: “They are often neither scientific nor transparent about their sampling methods.” Exit polls, however, are more inscrutable. Why is there such a wide variation in such a short time between what the ‘opinion polls’ and ‘exit polls’ say on the composition of the next Lok Sabha? Was the size of the samples different? Had different localities been chosen for questioning? Had different people been chosen in the same localities to answer the same or different questions? Do people change their minds so fast? Why aren’t all these spelled out by the forecasting agencies?

Several organisations — Indian and foreign — go wrong in their projections about India’s GDP growth or exports, and often revise their numbers. Still they do not cause much of a flap because we know the exercise is undertaken in the cold light of day, taking into account hard facts and statistics.

Exit polls seem to suggest that cold, dispassionate logic has simply given way to opacity. Hence, they could prove hit and miss affairs if past experience is anything to go by.