In the land of astrologers and soothsayers, it is faintly amusing to see our politicians in a lather over opinion polls. The very netas who plan their lives according to the stars are now choosing to call the polls ‘unscientific’! Barely five months ago, when the polls were predicting a Congress win in Karnataka, the party which now wants polls banned was praising their accuracy. In 2004, the BJP, which now is pushing for the publication of polls, had joined the chorus in demanding that polls be banned once elections were notified. As times change, so, well, does ‘opinion’. Sadly, as always in the noise of surround sound, the facts are in danger of losing out to a mix of ignorance and partisanship.
Examine the primary arguments against polls. It is being argued that opinion polls must be banned because they are inaccurate and unscientific. Yes, opinion polls have gone spectacularly wrong on occasions in predicting election results. But there have also been several times when they have been remarkably accurate. The point is, psephology is not mathematics: to expect exact numbers is to misread the science of analysing voter behaviour. A good poll is, at best, a trend-spotter; the actual projections are difficult to measure in a highly localised and competitive election environment.
There’s also the counter-argument: if opinion polls are inaccurate, what indeed is more ‘accurate’? Are journalists who confidently predict which way the wind is blowing the ultimate barometers of the political mood? Are politicians who will proclaim their party is winning a landslide to be trusted? Or is the satta market the real guide to political fortunes? Does hard data assembled through a rigorous polling exercise not matter, or do we simply predict elections on the basis of intuition?
Those who argue that a sample size of a few thousand cannot measure a country as large and diverse as India again have little understanding of either statistics or sampling. The true test of a well done poll is not its size, but how representative it is, and the quality of the questionnaire. The real information gleaned from a poll goes well beyond projecting who is winning elections. For example, one of the more fascinating aspects in a recent poll we did was to measure the popularity of Narendra Modi versus Rahul Gandhi among those voters who had Facebook and Twitter accounts. The viewer may look at the big number of which party stands where; a student of politics will find a minefield of information beyond just victory and defeat.
{{/usCountry}}Those who argue that a sample size of a few thousand cannot measure a country as large and diverse as India again have little understanding of either statistics or sampling. The true test of a well done poll is not its size, but how representative it is, and the quality of the questionnaire. The real information gleaned from a poll goes well beyond projecting who is winning elections. For example, one of the more fascinating aspects in a recent poll we did was to measure the popularity of Narendra Modi versus Rahul Gandhi among those voters who had Facebook and Twitter accounts. The viewer may look at the big number of which party stands where; a student of politics will find a minefield of information beyond just victory and defeat.
{{/usCountry}}The second argument made is that polls influence voter behaviour and question the idea of a free and fair election. The evidence here is sketchy at best and, in most instances, defeated by anecdotal history. The 2004 elections when most pollsters failed to predict a UPA surge is perhaps the best example. Right through the campaign period, the polls were predicting a NDA victory. The results revealed both the limitations of polling and the autonomy of voter behaviour. Political workers maybe swayed by opinion polls, the average voter has shown no indication of casting a ballot based on poll findings. If anything, the voter may take vicarious delight in proving pollsters wrong!
The third argument and, in this age of declining ethical standards, perhaps the most persuasive, is that opinion polls can be bought and sold. There is little doubt that a few individuals of dubious integrity have entered the opinion poll ‘business’. We have had the unedifying spectacle in recent times of the same poll research agency giving different results to different channels within the space of a fortnight for a ‘national’ poll. While a highly competitive market may eventually sort out the serious pollster from a fly-by-night operator, there is no reason not to put in place transparent guidelines for election-related polls.
For example, in Britain, the original home of psephology, the British Polling Council (BPC) has been formed to ensure the highest standards of full disclosure so that consumers can judge the reliability and validity of the results. Why can’t we have a similar body, mandated by the Election Commission, to establish a code of conduct for polling in the country? Regulation might be one way forward to reduce the credibility deficit confronting opinion polls.
What is not a way forward though is the culture of seeking a ban on opinion polls simply because a particular political party may be discomfited with the findings. This is not simply about the constitutional guarantees provided by free expression. It is increasingly about a sarkari mindset which believes that the best way to control information flow is to enforce bans: don’t like a book, ban it; don’t like a cartoon, withdraw it. Today, its opinion polls, in future, if you don’t like a contrary opinion, you may muzzle that too.
Truth be told, this ‘fear’ of opinion polls primarily stems from those who can be best described as ‘drawing room’ or ‘television studio’ netas. The serious politicians know their future lies in the heat and dust of real India, in the slums, bastis and colonies where voters are increasingly influenced by the governance record of their rulers, not by colourful graphics that buzz on a television screen. Every opinion poll, for example, has tended to underestimate the silent voter of the Bahujan Samaj Party. And yet, Mayawati’s mass appeal is not diminished by the fact that pollsters tend to write her off all too easily. Instead of agonising over poll results, maybe netas would be better off focussing on rediscovering the common touch which can be decisive on the real day that matters: counting day.
Postscript: last year, during the Uttar Pradesh elections, Congress leader Digvijaya Singh publicly challenged our poll forecasts giving Samajwadi Party a thumping win, and the Congress doing poorly. He even offered to pay for dinner if he was proven wrong. I still await the dinner!
Rajdeep Sardesai is editor-in-chief, IBN 18 network
The views expressed by the author are personal