Delhi elections: police the pollsters, weed out the dubious
Most pollsters got the Delhi results wrong. none of the exit polls predicted a tantalisingly hung assembly, or that the AAP would be snapping at the BJP’s heels. Ajaz Ashraf writes.ht view Updated: Dec 29, 2013 23:38 IST
The Aam Aadmi Party’s (AAP) robust performance in the Delhi assembly elections should have turned the honchos of polling agencies scarlet, ignominious failures as they were in capturing the mood of the city’s electorate.
But for one pollster, none of the exit polls predicted a tantalisingly hung assembly, or that the AAP would be snapping at the BJP’s heels, or that the Congress would come a shockingly distant third. In their earlier rounds of polls, almost all furnished empirical evidence of the AAP’s campaign gathering momentum, but still lacking the steam to gather more than 20 seats.
Their abject failure assumes tremendous significance against the backdrop of the furious debate that had preceded the publication of these surveys.
Critics, mostly from the Congress, dubbed the polling agencies as opinion-makers, as against their own description of being opinion-seekers, who invoke the term scientific to describe their surveys, but which are, in reality, deeply flawed, or tailored to suit one party or the other, or satisfy the media-owners’ political preference.
Scientific or dubious, opinion polls, even their practitioners agree, can and do influence voters. Indeed, those who have no strong ideological inclination or are not steadfast in their tilt for or against a party tend to cast their votes for one widely expected to be a winner, which is what surveys portray.
This is why critics want them banned. Those who believe opinion polls are fraudulently manufactured ask: are such polls any better than distributing liquor and money to voters to influence them?
However, those opposing a ban claim such surveys inform and assist the voter in judging the competing political parties and decide on his or her vote. A ban, therefore, would violate the fundamental right to free speech.
Yet, it might not be rhetorical to ask: since opinion polls rarely predict exact electoral results, and often go egregiously wrong, are they not misinforming the public, deliberately so in the eyes of some? Isn’t misinformation encouraged because of the protection under the right to free speech?
But view the issue from another perspective. In journalism, a wrong story doesn’t lead to a ban on the publication which featured it. Nevertheless, checks and balances have been introduced — retraction and apology, and threats of defamation suits. Might it not be time to police the pollsters, to weed out the poorly skilled and the dubious?
Perhaps a way out to counter misinformation, deliberate or otherwise, is to insist on pollsters evolving an index of success and publishing it every time they release a new survey.
Such an index should convey the degree to which they were right in predicting past election results, say, over the last 10 years, both in terms of vote-share and projected seat tally. This would provide a sense to voters whether or not they can rely on a given survey on deciding on their vote.
Perhaps the polling agencies should also be asked to emulate psephologist Yogendra Yadav who published the raw data on the AAP website every time the party made public its internal surveys.
The raw data should not only disclose the methodology followed in a survey, the sample-size and its diversity, but also the questionnaire the field-workers took to respondents and their responses, as also the method adopted to numerically grade them. This is vital as poll experts claim it is easy for them to tell whether a given raw data is genuine or not, thus ensuring wannabes don’t venture into the complex business of predicting human behaviour.
Ajaz Ashraf is a Delhi-based journalist
The views expressed by the author are personal