Over the past few weeks, preelection opinion polls have been published widely for the election-bound states of Uttar Pradesh, Uttarakhand, Punjab, Manipur and Goa across print, television and digital media. Recently, the Akhilesh Yadav-led Samajwadi Party, the BJP’s principal opponent in UP elections, wrote to the Election Commission of India (ECI) asking it to ban opinion polls on the ground that these are merely ‘campaign techniques in disguise’ to ‘influence’ voters. The SP’s contention was that publishing opinion polls just a few weeks before the election was a violation of the ECI’s ‘model code of conduct’.
The SP’s objection to opinion polls could be dismissed as loser’s disdain for ‘scientific’ surveys, given that all opinion polls have projected a BJP victory in UP. But there is some merit in its request to the ECI to take cognizance of such polls and regulate them. For, opinion polls are reflective of the opinions of a limited number of voters which could have an influence on the voting behaviour of others on the basis of projection of a likely winner. It is common knowledge that the vulnerable and undecided voters tend to vote for whoever they think is likely to win. Therefore, it is quite likely that opinion polls could have a ‘bandwagon’ effect that could possibly make the ‘likely’ winner an actual one.
Objection to opinion polls by political parties which are projected as losers, is an old grievance that has cropped up every few years since the late 1990s. Although psephology is a statistical study of elections and trends in voting, the quantitative analysis of elections has often been found to be inaccurate and problematic, simply because it is perceived as an attempt to quantify the ‘wind factor’ that sometimes goes horribly wrong. At times, the margin of error is so huge that the predicted neck-toneck fight becomes a one-sided contest. There have also been occasions when opinion polls have failed to predict the victory of one party and the extent of defeat of its rival.
Legally, restricting opinion polls could be deemed ‘unconstitutional’ under Article 19(2), as opined by the then Attorney General Soli Sorabjee in 2004, because they do not fall within reasonable restrictions on free speech. In 2018, a three-judge Supreme Court bench had also declined to intervene in a public interest litigation which claimed that ‘unregulated opinion polls’ were ‘disseminating false information’. While exit polls are banned during the course of multi-phase elections, opinion polls are banned during the 48-hour silence period before voting. But extending the ban to a longer period like when the model code of conduct is in force, so goes the argument, will impact media freedom and the people’s right to information. But the freedom is not absolute and allows for certain restrictions under the Representation of People’s Act.
Given that almost all polls are nontransparent and provide little information on methodology, the important question is whether the information provided by these polls is correct or compromised. As the information disseminated by opinion polls is an extrapolation of the data on voting choices, it is quite likely that the data can be suspect if the sample size is not truly representative of the diversity of the electorate. The fact that even exit polls sometimes get the actual outcome wrong indicates that actual voting decisions are far too complex for pollsters to comprehend and predict. This is why opinion polls sometimes get only the direction of the actual outcome right but many a time, fail to predict the winner. For instance, in Uttar Pradesh in 2017 polls, most opinion polls not only failed to predict the winner but also the extent of the BJP’s victory and the Congress-SP combine’s defeat.
The inexact science of predicting elections has been in crisis for quite some time. One problem is that most polls make mistakes in the ‘same direction’. While one would expect some opinion polls to underestimate one party’s eventual result and others to overestimate it, what is problematic is that most polls are often skewed in one direction. Since opinion polls suffer from many anomalies like ‘shy voter’ and ‘lazy voter’ factors and non-response errors, predicting elections not only becomes a difficult task but sometimes the result could also be disastrous. This raises an important question about the reliability of opinion polls. Another question that begs an answer is: can opinion polls be trusted as truly independent, objective and scientific study of the voting preferences of an electorate?
Since there is no way to ascertain that a certain opinion poll is rigged or not, a compromised poll disguised as a ‘scientific’ study disseminated on a pliable platform could not only mislead people, but could also be weaponised by interested parties to influence voters. This is not to suggest that all opinion polls are compromised or they often fail to predict an eventual winner in an election. But the fact that many times they get the election results wrong suggests that these polls may be scientifically flawed. After all, being right in some elections may not be proof of their credibility or scientific robustness, given that the cost of conducting opinion polls for large states runs into several crores of rupees. The question is: how many news channels or survey agencies can afford to spend large sums to conduct opinion polls in a proper, scientific manner?
Political parties are not alone in doubting the integrity of opinion polls; even the Election Commission and Press Council of India have expressed concerns about the possibility of opinion polls likely to be exploited by interested individuals or groups to misguide and mislead unwary voters, if they are ‘sponsored’, ‘motivated’ and ‘biased’. It may not be right to deny the suggestive power of opinion polls on voters with an argument that in any case, such polls often get the actual result wrong. It also may not be right to assume that all polls are done in a rigorous scientific way. Therefore, the Election Commission should look into the business and impact of opinion polls on voters and regulate them.
The writer is an independent Mumbai-based senior journalist