As India’s General Elections approach and Opinion polls begin to dominate TV news primetime, it is time to put the practices of this industry under the scanner.
Here is CSDS, often touted as “most reliable” with its opinion poll for Uttar Pradesh elections 2017.
When confronted with the disastrously wrong predictions after results came out, CSDS’ Sanjay Kumar haughtily dismissed all criticism by comparing himself to Sachin Tendulkar scoring a duck.
The failure of CSDS is hardly uncommon among well known Indian polling agencies: CVoter, Axis, Chanakya, etc. And these results from Uttar Pradesh accompanied by Sanjay Kumar’s response are just an example of the kind of arrogance and impunity with which India’s opinion poll agencies operate.
Can you imagine any other industry dismissing its consumers with such arrogance?
The reason such impunity and arrogance exists is because of lack of accountability. After all, nobody follows up with the opinion poll agencies after the election results come out. There is no public demand to put the data and practices into the open, to conduct public investigations of what went wrong and if necessary, take action against those found lacking. Did the opinion polls fail due to incompetence or because they were compromised? Either case would be good enough reason to make several “experts” lose their jobs. But that never happens. And the same experts make it back to TV before every election.
We could dismiss these opinion polls as just another media circus. Except for the disturbing possibility thrown up by elections in 2018 that these opinion polls could be feeding into a deliberately fake narrative that is seriously subverting our democracy. This is where it stops being funny.
Let us think about the final results from the recent elections in Rajasthan and Madhya Pradesh. In Rajasthan, the Congress had a lead of less than 2 lakh votes over the BJP across the entire state. In Madhya Pradesh, it was actually the BJP that had a lead of about 50,000 votes over the Congress across the state.
This means less than 1 lakh votes in Rajasthan could have changed the outcome of the polls. In Madhya Pradesh, if the BJP had got just another 50,000 votes, who knows?
Now, look at the kind of headlines generated due to these so-called opinion polls.
This article itself has over 10,000 shares! Not only do these opinion polls reach hundreds of thousands of viewers through widely popular television channels, but their impact is also magnified by coverage in newspapers. In the information age, the news about these so called opinion polls spreads within hours from smartphone to smartphone.
In Rajasthan this year, voter turnout actually dipped from 75% in 2013 to 74% in 2018. Who were these voters and what put them off? What would have happened if these people had gone out and voted?
Given how close these elections were in Rajasthan and Madhya Pradesh, there is a very real chance that consistently negative opinion poll results being shown for the BJP in both states changed the outcome of the election.
Here are CVoter predictions for Rajasthan!
The BJP has been underestimated by 5% and the Congress overestimated by a whopping 11%, which is way outside the 3% margin of error! How does a mistake of this magnitude happen? Sheer incompetence? Or an agenda?
Where is the accountability?
If just 0.2% of voters in Rajasthan fell for this narrative of the Congress being far ahead, it means that surveys like this determined the outcome of the election!
And isn’t that a scary thought? The fate of literally crores of people could have been decided by a private news channel and a private opinion poll agency, one or both of which could have been incompetent or pushing an agenda.
Sure, many surveys have got it wrong in many states in the past. In December itself, a lot of opinion polls got it wrong in Telangana. I have written about Uttar Pradesh before. But, fortunately, those were wave elections and you could say that it ultimately did not matter if 0.5-1% of people were affected by a (deliberately faked?) opinion poll narrative.
But what about an election such as Rajasthan or Madhya Pradesh where 0.5-1% makes all the difference?
And let us not forget what happened in mid-2018 in Karnataka. Here is a snapshot of opinion polls for Karnataka.
Simply shocking! The Congress dominated the opinion polls, despite falling far behind the BJP in final results. In fact, the BJP fell just 6 seats short of a majority. Again, if the opinion poll results affected just 0.5-1% of voters, it is possible that opinion polls have decided who rules over the 6.5 crore people of Karnataka.
This is a worrying sign for democracy. This is no longer funny.
It is time for the public to demand higher standards from opinion polls and the media in general. The science behind surveys is grounded in solid principles of mathematics and statistics. Surveys cannot fail so badly and so often. Something else has to be going on. Will honest opinion poll agencies admit their mistakes, conduct a public investigation into what went wrong and punish the incompetent and/or the corrupt if they can be identified?
Or will they return to television screens without a hint of shame or apology?
It is okay if some self-admiring intellectuals working with opinion poll agencies see themselves as Sachin Tendulkar. They would do well to remember that Sachin Tendulkar would work long and hard analyzing each failure instead of going on as nothing had happened. And when there was a string of failures, he had to quit as India captain. Twice. How many of our opinion poll Tendulkars have quit in the last 10 years?