T N Ninan: Polls apart

WEEKEND RUMINATIONS

Time was when everyone used to turn to journalists for wisdom on what was happening in the elections. But the fact is that ever since Indian elections ceased being a one-horse race, the journalists have usually got things wrong: for evidence, read the newspapers of 1971 and 1977.

But the journalists got badly exposed only in 1980, when none that I know or read predicted a return for Indira Gandhi with a two-thirds majority, whereas Prannoy Roy and ORG-MARG did, and confirmed their reputation as expert psephologists when they forecast Rajiv Gandhi's landslide win in 1984.

Soon the standard journalistic ploys of asking taxi-drivers and people gathered at roadside tea-shops became discredited as truth serums, and psephology became a growth industry.

But now the pollsters, with bigger and bigger budgets, armies of researchers and sophisticated tools of analysis, are beginning to get things wrong too. Their record in the last round of state elections was poor; why, even the exit polls were off the mark. And the record this time round is no better.

There were at least five opinion polls for these elections, and as many exit polls "" with four rounds each in each case. Yet no one forecast that the Congress would be the single largest party, or that the alliance it leads would do better than the NDA.

It's not just the TV companies and newspapers that get the wrong numbers on polls; Chandrababu Naidu is said to have commissioned a truly massive poll in Andhra Pradesh, with a 5-lakh sample, and he was told he would win.

Instead, he got a drubbing. Whether the BJP was similarly misled by its pollsters isn't known, because Pramod Mahajan is said to have forecast 260 for the NDA. So we must begin to wonder whether psephology will now go slightly out of fashion.

In truth, the politicians seem no wiser even without pollsters. A month or so ago, Shiela Dikshit didn't think her party would win more than two seats in Delhi; it has won six.

S M Krishna was assuring his party high command till the last minute that he would get 105 seats in the assembly (he got 75) and eight in the Lok Sabha (he was right on this one). The exception that I know of was Jairam Ramesh, who said two weeks ago that the Congress (sans allies) would win 145 seats. It has.

Given the problem that the experts (politicians, pollsters, journalists) have in divining the popular mood, it is odd how ordinary people are often able to put their finger on the pulse with seeming ease. I happened to be in Kerala on a whirlwind visit last month, and asked a cousin (who is a non-political tea planter) who he thought would win in the state.

He was quite clear that the Congress would get wiped out. In Bangalore, the owner of a printing press was equally clear: "Prannoy Roy and NDTV have got it completely wrong," he insisted.

"Deve Gowda is going to do very well in the state elections." He was right. And, back in Delhi, a government official invited to an editor's birthday bash last week berated the media for missing what seemed obvious to him: the NDA, he said, would not cross 200 seats.

Asked to give state-wise details, he gave his numbers; the total reached 184, and as things have turned out the NDA has got to 189. None of the assembled journalists (including your columnist!) would believe him, with one honourable exception: Bharat Bhushan of The Telegraph.

My point is that, if the experts could not divine the electoral mood before the votes were counted, how are they so sure immediately afterwards that the vote carries precisely x or y message?

Read the different analyses and it quickly becomes clear that most analysts are simply projecting their own biases. If you don't like Modi's handling of the Gujarat riots, the state's voting pattern is interpreted as a negation of Modi "" when all it might mean is that the BJP in the state is self-destructing because of internal squabbles.

Those who don't approve of economic reforms interpret the NDA's and the Telugu Desam's ouster as a message from Bharat to Shining India, even though the voting does not show an urban-rural divide, or even a rich-poor dichotomy. And so on. In short, treat the experts with the scepticism they deserve.