The irrelevance of exit polls  


Exit polls have always been a waste of time, energy, and resources. Their utter uselessness is evident from the numbers which are not only wide off the mark but more often than not delinked from ground reality.

Bafflingly enough they continue to get a disproportionate share of attention in the national media. What is worse is that the release of data across different television channels is followed by mind numbing debates in which representatives of political parties slated to be on the losing side furiously deny the authenticity of the numbers along with partisan analysts regardless of their merits, if any.

At the outset it is mind boggling why any TV channel would hire the services of a pollster to guess the results of an election when the declaration of the actual results is just 2-3 days away. But such is the lure for publicity and urge to satisfy public curiosity that logic trumps reason.

Barring the odd seriously conducted exit poll like the India Today-Axis My India run by Pradip Gupta, the projected seat distribution of others is reflective of what their sponsors would like the outcome to be rather than what they are likely to be. The data is often tweaked in accordance with the political preferences of the hiring agency. Most pollsters do not mind obliging their clients because at the end of the day predicting election results is their bread and butter. Credibility does not matter so long as they the client’s interest is served. Only that can ensure their retention in future elections.

The exit poll numbers of the Bengal assembly poll brought out the latent insecurities, contradictions and utter farce of the exercise. Those seeking the TMC’s return to power cite the results handed out by C-Voter which was strangely hired by both Times Now as well as the ABP group. Quite apart from relatively unknown pollsters like P-Marq, Poll Strat, and ETG Research, all of whom give Mamata Banerjee’s party a comfortable majority, some going to give her close to 200 of the 294 seats despite open evidence of anger against the TMC’s corrupt and violent rule.

Those familiar with the ABP group’s pre-poll and post-poll surveys know they have always been reluctant to show the BJP winning. Why? Because the group’s leading newspapers like the Ananda Bazar Patrika and the Telegraph are inveterate Modi hating publications which scoff at the possibility of a cow belt formation ruling the land of bhadralok regardless of the shifting tide of sentiment.

Times Now is a strange kettle of fish. Though the channel is seen as part of the godi media by critics and enemies of the current dispensation, it is ironical they hired C-Voter, the shiftiest pollster, to do their exit polls. C-Voter is well-known for hedging its bets.

Unlike Axis India which has the guts to stick its neck out when sure of its data, C-Voter is famous for pegging BJP numbers  low even when the public mood suggests an overwhelming preference for the party. The only reason why Times Now may have hired Yashwant Deshmukh to do the job is either because there was no other competent pollster available or that the channel did not want the numbers to show it was batting for the BJP.

A good example of C-Voter’s disingenuity is their pre-poll survey of the 2013 Madhya Pradesh assembly poll. While most others forecasted a BJP landslide, C-Voter gave the party around 126 of the 230 seats, ie. just 10 above the half-way mark. The actual results had the BJP getting 160 plus. Technically, C-Voter can claim that it rightly predicted a BJP victory when the hard reality is that its numbers were wide off the mark.

India TV, a pro-BJP channel, hired the services of People’s Pulse, a Hyderabad based research organization run by a young research scholar, Dr Sajjan Kumar Singh, a JNU alumnus. The channel may have hired him for his numbers which predict a BJP landslide in Bengal, but Sajjan is no loyalist of the saffron brigade.

In fact, if there is someone whose field work has been truly awe inspiring, it is his. And this is regardless of the fact whether his seat prediction (BJP 172-185 versus 65-88 TMC) hits bulls eye. A signed article by him appeared in the Indian Express on March 21 in which he held his ground that poriborton hobey (change will occur). Cogent and intuitive reasons were cited in support.

Having conducted the pre-poll study in all the 294 constituencies where his reports outlined the basis of his conclusions amid the shifting political sands, Sajjan undertook another round of study to come up with his exit poll numbers. He is probably the sole pollster who did a phase wise mapping of the political mood rather than a sub-region analysis. Given the solidity of his study it is hardly surprising that his pre-poll and post-poll numbers are roughly the same.

Most pollsters are clueless on sampling techniques or how much weightage be given to population diversity. Some are plain lax and non-serious. It is perplexing why they at all take the trouble of going through the motions of an exit poll.

A classic example is that of Today’s Chanakya which confessed that its 2015 exit poll for Bihar was an unmitigated fiasco. “A simple computer template coding marking the alliances got interchanged.” The reason: the day the template was made, a senior staffer was on leave.


Comments are closed, but trackbacks and pingbacks are open.