The 'other' polls
The chaos and cacophony of elections is the surest sign of a vibrant democracy. What role do opinion and exit polls play in this democratic process? A non-partisan debate on this issue must underscore the qualitative difference between opinion polls and exit polls and the differing impact each has on the democratic process.
Opinion polls are sample surveys, reporting statistical results of potential future behaviour by voters who answer certain questions. An opinion poll is always conducted before a voter has exercised his franchise, usually several days or weeks prior to voting. It usually asks more than one question. An exit poll is conducted upon a voter who has already voted. It involves no question and seeks to report how a voter has voted.
Keeping this difference in mind, there are several reasons for banning exit polls and an arguable, though weaker, case for banning opinion polls. Democracy is the bedrock of our nation and of the Constitution. A republican secular democratic nation has been held by the apex court to be part of the basic structure of the Constitution. Free and fair elections have also been held judicially to be part of this basic structure. This means free and fair polls are an unamendable, inalienable feature of our Constitution.
In turn, free and fair elections necessitate a level playing field where the contestants and voters are not subjected to unfair external influences which vitiate the process of elections. The Representation of People Act (ROPA) and the Model Code of Conduct, along with several other rules, are all designed to ensure this essential fairness in the electoral process.
In a multi-phased electoral contest — and the exit polls ban argument is relevant only in multi-phased polls — the exit poll results of a prior poll have a direct impact on the next poll. For example, the present polls are scheduled for April 20, 26, May 5 and 10. An exit poll publication for April 20 suggesting heavy voting for Party A is bound to create an overall ambiance of victory in favour of that party. Such pre-declaration of exit poll results often leads to a self-fulfilling prophecy. Party A may have strong support in the April 20 polls but may normally be a miserable loser in all the next three phases. However, an ambiance in its favour qua the April 20 polls may generate a ‘herd instinct’ or a ‘bandwagon effect’ amid voters to follow the likely ‘winner’ of the first phase.
This may itself not be pernicious if the April 20 exit poll results reflected the complete reality. But exit poll results are based on fractional samples, generalising from ‘miniaturised reality’. Since they do not even claim to represent the whole truth, their impact on future polls is bound to distort the electoral process. Vitiation of democracy and violation of the basic structure is a likely consequence.
Opinion polls are worse because they generalise not from actual behaviour but from presumed and predicted behaviour. The fractional size of the sample (e.g. 40,000 in an Indian electorate of 60 crore) underlines its inherent inadequacy. Many surveys thrive on much smaller sample sizes! Second, the sample can’t be truly representative, especially in a humungous and heterogeneous society like India. Third, opinion polls are conducted well before the polls and their remoteness in time affects their predictive utility. Several recent polls have quoted widely divergent figures and this reflects their untenability. Nevertheless, they inexorably influence the voter and create an uneven playing field.
The strongest argument in favour of polls is the ‘freedom of speech’ argument under Article 19(1) (a) of the Constitution. It is argued that the voter is astute enough to discount such generalisations and that the 19(1) (a) right can only be restricted by law made by Parliament relating only to subjects itemised in Article 19 (2). The best answer is provided by Part XV of the Constitution and especially Article 324.
It vests all powers of superintendence, control, direction and conduct of all elections in an independent constitutional authority, viz. the EC. In the Gujarat case, a five-judge bench of the apex court read these words very widely and directed them to be given the widest possible amplitude. It was held that the EC’s discretion would be subject to specific parliamentary legislation on the subject. Since none presently exists, the EC would be free to exercise a very broad swathe of discretion if it were to
conclude that such polls affect the purity of the electoral process. The EC, arguably, could be held to be acting not in the realm of free speech at all but in the dialectics of conducting an election.
The Gujarat judgment builds upon and reiterates established earlier jurisprudence of the apex court on this issue. Statistical studies do cast doubt on the ‘bandwagon effect’, but it can’t be categorically asserted that there is no significant or material effect of such polls on elections. A level electoral playing field is too precious to be risked on unsubstantiated data.
For far too long has this issue been addressed by default. In 1998, the EC issued guidelines banning polls. In a challenge by two magazines, the oral observations of the apex court suggest that they were disinclined to uphold the EC stand. But the guidelines were withdrawn by the EC and the matter was never constitutionally tested. Oral observations in a withdrawn proceeding do not constitute the law of the land under Article 141. The Gujarat judgment is of recent vintage. There is no legislation on the subject.
Foreign legislation lies on both sides — some countries legislatively ban such polls, others don’t. The matter is thus ripe for EC intervention in an unoccupied field followed, undoubtedly, by a final word by the apex court.