With the recent request by the Congress Party to the Election Commission to ban or restrict pre-election opinion polls, the subject is now a matter of debate in the media and elsewhere. Naturally, TV media, indeed newspapers also, oppose the move in general, probably more from their own TRP/readership perspective.
First, looking at the opinion polls from a purely technical point of view, questions need to be answered relating to the number of samples, randomness of the subjects covered, existence or lack of bias or skew. These are extremely relevant issues from the scientific/mathematical point of view to establish ‘confidence’ limits, that is reliability. The tendency in many samples is to cover metropolitan or urban areas, as this is more convenient; only a small portion of the samples come from rural areas—and very few from really remote parts. Second, very rigorous identification of the subjects, based on mathematical randomness, is prescribed to enhance reliability—it is doubtful if many of the present polls meet these criteria.
The other technical aspect relates to whether the interviews are true proxies to how a person actually will vote in the polling booth. Without referring to change of mind between the interview and the polling date, it is a well-known phenomenon in India that a large number of respondents do not give accurate or true answers to questions which are posed to them—witness the frequency of perjury in our court system, much more than in other countries. In older, Western democracies, citizens are much more forthcoming on such issues. This is a significant factor affecting the reliability of opinion polls.
The other technical issues relate to the nature of questions asked, their potential bias, as well as the level of complication in the queries. All these contribute to the reliability of the answers received, and thus the credibility of the survey.
Apart from the technical issues involved, other significant potential distortions are possible, indeed are seen, in India. In the first place, many or most, if not all, media houses and newspapers are owned or controlled, however indirectly and camouflaged with complexity, by large business houses and strong commercial interests. Many political parties also have a large, if not total, say in the management of their media vehicle. It is utopian to expect that the owners do not influence directly the editorial position, even on individual items, of the relevant media. This translates to strong potential skewing of perspectives, procedures, indeed distorting the process of ascertaining opinion, influenced by parties who are directly interested.
Media houses and many political parties have vehemently argued that any check or control on dissemination of poll results will tantamount to interference with free speech, which is a fundamental right. This is certainly true; any circumscription of dissemination of opinion polls ought to be done only if absolutely essential and with great care. On the reverse side of the coin is the imperative necessity that the voter should not be influenced through false or coloured information; this will distort or adversely affect his right to vote freely, by his choice—this is somewhat akin to using false advertisement to influence purchasing decisions, which is illegal.
Many TV channels and newspapers have selectively publicised ‘expert’ opinions from foreign commentators, from US and Europe, to propagate the perspective that opinion polls, accurate or otherwise, do not influence voting patterns. The nature of the electorate in India is quite different from Western countries; we cannot translate Western experience in this regard to India. Besides, there is general craving for ‘stable’ governments—this could be a factor influencing voting patterns, if poll predictions, right or wrong, are made in advance. In a large number of constituencies, the contest is multi-pronged, with sometimes four or five serious contestants—even a 2 to 3 per cent shift in the vote due to false information could tilt election results. This is not a theoretical issue.
Once upon a time, the national elections used to be conducted within a period of a week or so. We now see that even this year’s elections in five states are spread over a month. In a large country like the US, the entire voting process is completed in a day, and the final results declared the same evening. It is a reflection of the immaturity of our democracy, and the passions generated in the polling process, along with the inordinate rewards associated with success, that the spectre of violence is ever present in Indian conditions—one can only hope that this situation will drastically improve soon. There is no question that exit polls should not be permitted; that is when the overall polling in a bunch of states, or for the national elections, once the process starts, till the final results are announced, there should be total ban on interim opinion polls, or exit polls.
On balance, it appears to be necessary to bring some control or regulation on the ‘opinion polls’. Apart from ensuring that the pollster publishes the full methodology, it also needs to be considered whether opinion polls should be permitted after the ‘election notification’ is issued.
tsrsubramanian@gmail.com