Irrational exuberance: ban opinion polls

Written by Surjit S Bhalla | Updated: Mar 2 2014, 04:45am hrs
Why is the knee-jerk reaction among Indians to indulge in bans What did the burning and banning (figuratively; in reality it was withdrawing from circulation and pulping) of Wendy Donigers book on Hindus achieve or is likely to achieve How many peoples views were affected, and of those whose views were changed, how does it matter Assume the worst case scenariosome people in India, and only in India, actually believed what Doniger said about Hinduism. So what

This is 2014, and such nonsense prevails because of the Indians irrational exuberance for authority. Back in 2004, the Election Commission (EC) of India came up with an anti-constitutional, anti-liberal view of banning opinion polls a full 45 days before the first date of polling! This was then, as expected, rejected by the Supreme Court. The UPA government, not to have such minor matters as the Constitution come in the way of its interventionist ideology, amended the Representation of the People Act to support a ban of 2 days for opinion polls and no exit polls before all polls were completed. Possibly because of this victory over the Indian Constitution, the 45-day ban is again making the rounds. Isnt it time that we question the ECs view as stupid, irrational, and totally devoid of any legal or moral or practical content

It all started quite sensibly with MS Gills Election Commission imposing restrictions on opinion and exit polls in February 1998. This policy was for no prohibition or restriction of opinion polls until just two days prior to polling; and for no exit polls until 48 hours before the last polling day. As can be seen from the accompanying table, this was a reasonably enlightened policy and well within best practices in other democracies. Many countries have zero restrictions on opinion polls. If there is a ban, it is of 1-2 days duration. The median is 0 days and the average ban is of 2-days duration. Italy is a clear outlier with a ban for a full 15 days before polling. The extreme outlier is the ECs proposed 45-days ban.

Did the 2004 election commission look at any data before coming out with their learned prescription Their conclusion, which they keep reiterating presumably every time they get prodded by an interested political party, has not changed all these years, especially years in which internet and social media penetration has expanded manifold. The effects of irrational bans, like in Italy, are predictable.

Responding to the restriction, bloggers in Italy found imaginative ways to make the ban look stupid. One site termed elections as horse races and candidates were, well, horses with thinly disguised names. Another site had candidates as cardinals selecting a Pope.

Contrast the Italian behaviour (disrespect for irrational authority) with that of an Indian (fawning respect for the same). Our norm is to respect bad authority and worse laws. A true-life example is revealing. A leading TV channel was scheduled to broadcast the findings of my research on the effects of economic performance and incumbency on voting behaviour. By no stretch of imagination could my research be described as an opinion poll and even by less imagination as an exit poll. Yet, the conservative Indian lawyers felt obligated to be safe and postponed the broadcast from December 3 (one day before polling date) to December 6; thus proving the adage that when the law is an ass, there are donkeys who feel compelled to follow.

Let us go a bit further into the implications of the proposed EC ban, and something for the Supreme Court to consider if the proposal reaches their enlightened shores for consideration and judgment. If the SC were to agree with ECs ultra-regressive views (all the liberal parties are in support and the non-liberal BJP is opposed to the ban) then how will the ban operate Just yesterday, a respected international polling organisation, Pew Research, published its findings of a nationwide opinion poll (part of its Global Attitudes Project). Over a period of one month (December 7, 2013, to January 12, 2014), they conducted a random, all-India sample of only 2,464 respondents. It found that 60% of the sample had a very favourable view of Narendra Modi compared to only 23% for Rahul Gandhi and a lesser 19% for Sonia Gandhi. Interestingly, and possibly unconstrained by domestic manipulation, Pew Research found it informative to poll the same question for Anna Hazare and not Arvind Kejriwal. It could also be that the sample found too few respondents reacting very favourably to the AAP representative.

Now many of our learned politicians, and perhaps even the EC and the Supreme Court, would want to ban the Pew opinion poll for its small size. How can one individuals opinion possibly reflect the opinion of 3,25,000 potential voters Off with the poll, and let us ban it after we burn it. The poll findings cannot possibly be right, for unlike the goras (whites), we Indians know that we are a very heterogeneous population. And such a small sample size must have been undertaken because Pew is either cheap or has taken money from some politician/political party/industrialist.

Will the ECs proposed ban mean that no survey data can be gathered and no one can go around asking questions about voting preferences Of course not. So, what happens when Pew gathers such information and publishes it on the internet The whole world will know, and as the whole world is watching, the ban will be totally ineffective. This is the key point that all our policy makers must be made aware of. Duniya bahut badal gayee hai. Banning just does not work any more, there are too many legal leakages in the system, and too many potholes on the way to effective enforcement. Just grin and bear itand dont make a fool of yourselves.

The author is chairman, Oxus Investments, an emerging market advisory firm, and contributing editor, Indian Express. Twitter: @surjitbhalla