Afternoon on Sunday exactly a week ago in India, suddenly eggs went into short supply mode. Didn’t they all land up on the faces of psephologists and many pollster turned anchors in Delhi and other places??? For more than 3 months, India had been hijacked by elections in one state – Bihar. At the end of the tiring, long drawn 5 phased elections and just before the actual counting on 8th Nov (last Sunday) – almost all channels went on an overdrive with telecast of their exit poll results. The channels had used different agencies to conduct the same, extrapolate the vote share into seats and projected the results. However, in the afternoon when the actual results started coming in, it was clear that all the channels and their agencies were way off the mark. While all the channels had predicted a very close election with a wafer thin majority either in favour of the NDA or the Mahaghatbandhan (MGB from here), the actual outcome was more than a landslide win for the Nitish led MGB and a massive blow to Modi led NDA and ofcourse to the pollsters!!! Ironically the one agency (AXIS) which had got it correct didn’t get the opportunity to air their poll results. The channel’s (CNN-IBN) management thought it outrageous to telecast a poll which gave 169-183 seats to the MGB we are told!!! Amidst all this mass Exit poll debacle there was only one individual – Dr. Surjit Bhalla who stuck his neck out and gave 175 seats for the MGB just by his own forecast model (not exit polls) in his Indian Express article a few days before the counting and got it right.
This massive blow to Modi may be a 1st since his historic 2014 win, but for the pollsters this blow is not the 1st. Before the Bihar elections when Delhi went into polls this year, the same thing happened. While most polls predicted a close fight between BJP and AAP, the final result was clean sweep by the broom party. None of the exit/opinion polls captured that kind of a sweep. In the 2014 Lok Sabha elections again it was only one agency – Today’s Chanakya which correctly called the outcome. Most others projected just a simple majority for the NDA.
Inspite of this way off their mark predictions what is the bet that the same agencies will not again carry out opinion polls/ exit polls the next time around?? These are competitive times for the channels and the polling agencies. As any election approaches (Central/State) these must be the first to salivate and glee at the prospects of increasing their TRPs with all the election related programming in which the opinion polls and exit polls are now an integral part. So we will continue to see the much hyped exit polls packaged as with new improved methodology, bigger and improved sampling next time over. Again.
Years ago when the idea of Opinion polls and Exit polls got imported to India from the US (I guess) it was fascinating. As mortals, there is this tendency in us to jump the gun. Those trips to palmists, astrologers and soothsayers of different hues to know our Aane wale kal explain this tendency. Similarly we feel excited to know who will be our rulers before the actual counting is done through exit polls though it is just a few days ahead. But US and other western democracies from where these concepts originated, are less complex and more homogenous as a society. So calling an election after an exit poll with a very small sample size and extrapolating the results there is a less arduous task I guess. (Even there Exit polls have gone awfully wrong) Ofcourse India is another kettle of beef (😁) altogether. We are repeatedly told that we have religion based voting, caste based voting, community based voting and what have you. But one thing which is missed in all this “______based voting” which the pollsters and analysts keep talking about is the New Indian voter’s “Brain based voting”!!!
As we saw now in the 2014 Lok Sabha elections, Delhi state elections and now in Bihar the difference in an election swinging one way or the other depends on the non-core party voters. Core party voters (members/supporters and sympathisers) vote for their party in any case. It is the Non-core party voters who bring in the swing in a particular party’s favour. They voted in a big way in favour of Narendra Modi in the 2014 elections and NDA got an overwhelming mandate. However in Delhi state elections the Non-core party voters who voted for BJP in the Lok Sabha elections voted for AAP and swept it to power. Similarly the non-core party voters pulled their might in favour of MGB in Bihar. So most of the non-core party voters in Bihar who wanted to see Narendra Modi as PM wanted to see Nitish as their CM. Perhaps. I didn’t get to see (or maybe I missed) this angle where the voters are getting smarter by the day in terms of their choices and do “Brain based voting”. So we get these results. Overwhelming mandate for Modi for PM. But Kejriwal for CM or Nitish for CM though they are entrenched firmly in different camps.
So the next time may be the pollsters should just identify if a respondent is a core or a non-core party voter and record to which party he/she voted instead of spending too much time if respondents are upper caste or Dalit or Maha Dalit,..,.. In which case, they will be in a better position to forecast how the actual swing is and save those eggs on their faces😜
Toon courtesy: http://www.manjul.com