Minds and the mandate

I was a cub reporter, new to the capital and even newer to the world of politics, when I covered my first parliamentary poll in 1971. The ca...

I was a cub reporter, new to the capital and even newer to the world of politics, when I covered my first parliamentary poll in 1971. The campaign manager of the Jana Sangh candidate from Delhi’s Karol Bagh explained very convincingly why his party, the precursor of the BJP, would romp home. The constituency was made up of x percentage of Jatavs, y percentage of south Indians, z percentage of traders, and so on. He had a ready explanation as to why each caste and ethnic group would be supporting the Jana Sangh. The candidate was a Jatav, traders were the backbone of the Jana Sangh, the popular local Jana Sangh councilor was a Punjabi Khatri, the party corporator was a south Indian, each of whom would generate a, b, c and d number of votes respectively, he assured me. After I had totted up all his figures, I was left in no doubt that the Jana Sangh had the election sewn up.

Halfway through the campaign I began to have grave doubts about the assumption that ethnic and caste breakdowns would decide the electoral outcome. Even the fact that the Congress candidate from the constituency was considered no more competent than a lamppost did not seem to be an issue. Every autorickshaw driver I met assured me that Indira Gandhi, fresh from the victory over Pakistan in the Bangladesh war, would win hands down. My experienced chief reporter brushed aside my misgivings, assuring me that rickshaw drivers were not representative of the Delhi voter. An objective poll assessment was mercifully not required since my newspaper was fully backing the Jana Sangh. But I was struck by the certainty of the veterans in the newspaper and the Jana Sangh party office that the Congress was going to be defeated, when every person on the street I talked to told me quite the opposite.

Mrs Gandhi’s sweep taught me my first lessons in election coverage. Trust your own instincts rather than follow what others say. Never get into the trap of predicting victory by armchair analysis of the constituency profile. The popularity of parties usually matters much more than the standing of the candidates. Most importantly, never let your own hopes cloud your objectivity. Every person you speak to is representative of the constituency, you can’t pick and choose.

I recall during the parliamentary poll in 1989 I travelled with two staunchly secular journalists in UP who refused to see the saffron storm brewing under their noses. On the highway from Kanpur to Agra, two out of every three persons we spoke to were confident that the lotus was in full bloom. But my companions insisted we had not met a representative sample. We rattled vertebrae, punctured car tyres and infuriated the driver by driving over dirt tracks into the deep interiors. The answers did not differ substantially from the highway respondents, but the correspondents remained convinced that the Janata Dal’s caste combination would see it through.

Today, the professional pollsters seem to have taken over the job of forecasting poll results from us mere journalists. But psephologists reeling off statistics and the armies of surveyors sent out into the hinterland are not necessarily less fallible than the rest of us. Those who collect the voter responses are accustomed to shooting straight questions, like do you prefer Surf or Nirma. An approach which does not work so well with the average voter. Which explains the bewildering and conflicting responses which have been thrown up by the rash of recent surveys. One poll found that over 60 per cent of the respondents thought politicians were doing a good job, another that the majority of politicians were crooks.

In last year’s assembly polls for Rajasthan and Chhattisgarh not a single pollster got the results correct and few predicted the extent of Uma Bharati’s sweep in Madhya Pradesh. At least one poll survey even failed to sniff Narendra Modi’s landslide victory in Gujarat a year earlier. Similarly, in the 1999 parliamentary polls, if I recall correctly, no major survey predicted the sharp fall in the BJP vote share in UP or foresaw the extent to which Laloo Yadav was in trouble in Bihar.

The 2004 campaign will be remembered for the largest number of mega poll surveys ever commissioned. Even political parties have increasingly come to rely on the professionals for advice rather than trusting their own gut instincts. It was on the basis of commissioned surveys that the BJP decided to break its alliance with Om Prakash Chautala’s party in Haryana and the AGP in Assam, as the inputs suggested that the BJP would be burdened by the unpopularity of its former allies. In retrospect, this professional diagnosis was faulty, the BJP is behind the Congress in both these states.

Story continues below this ad

The striking point about this year’s opinion polls is that though there were widely varying projections of the NDA tally in the initial stages of the campaign, in the last lap suddenly all the opinion polls have veered around to a fairly narrow range of differences. The forecast now is that the NDA will win or lose by a whisker. Which cynics might say is a smart way for the pollsters to cover their flanks. They can glibly explain that there was a slight last minute swing in vote which caused a major shift in the number of seats this way or that. The curious part is that while the final seat tallies are not very dissimilar, the forecasts for winners in states like Delhi, Maharashtra, Andhra Pradesh and Punjab are completely at variance, suggesting that in poll predictions the final total is more important than the sum of its parts.

The line-up of initial exit polls leaves us even more confused. The victors in states like Maharashtra and Bihar vary from poll to poll. Contrary to the simplistic initial impression, the first round of exit polls did not necessarily predict an NDA victory. If you read the fine print, most are not sticking their necks out but basically saying it could go either way. Which is why the mystery and suspense will remain until the ballot boxes are finally opened and the last vote counted.