Exit polls 2019: The predictions can get the winner right but not the seats

• In 2014, exit polls predicted a Modi wave but missed the number of seats by a wide margin.
  
  
• Pollsters faced much embarrassment in 2004 when they predicted an NDA win after a successful ‘India Shining’ campaign.
  
  
• In 2009 too, pollsters predicted a Congress win, but missed out the number of seats by a large margin.

Exit polls have gone wrong many times. They have been right too, but historically they have been playing safe, even when they were right.

In 2014, exit polls had predicted a Modi wave. They were right, yet they missed the number by around 40 more seats as Bharatiya Janata Party (BJP) came into power.

Exit polls arrive at their predictions after visiting constituencies, interviewing voters and collating data from samples. Currently, exit polls are conducted in India by India Today/CSDS (Centre for the Study of Developing Societies), AC Nielsen/Outlook etc.

Hit and miss

Pollsters faced much embarrassment in 2004 when they predicted an NDA win after the BJP’s ‘India Shining’ campaign. But, the country chose UPA instead.

In 2009 too, pollsters got it right in pegging a win for the Congress but missed out the number of seats by a large margin.

In 2019, the odds are once again stacked in the incumbent NDA’s favour. Even the stock markets rallied hard on the last trading before the exit polls.

After 2004 Election results were announced, Indian stock markets crashed to hit their first ever lower circuit, worried about the systemic change. Back then, UPA had not even announced the Prime Ministerial candidate.

When Manmohan Singh retained power in 2009, it was a pleasant surprise for the stock market. The thumping majority and stability was cheered by traders so much that the market hit upper circuit not once, but twice. And, trading was suspended for the day!

Fractures versus landslides

Exit pollsters were spot on in 1996, when they predicted a fractured mandate, and it was to be, leading to extended political crisis.

As polling veteran Prannoy Roy says in his book, ‘The Verdict’, it is much easier for pollsters to predict landslides, than close contests. But theoretically, only.

To publicly foretell a landslide victory would mean that the agency or a media house will have to stick their neck out and risk embarrassment in case they go wrong, according to an excerpt of his book.

This is probably why pollsters played it safe last elections and the one before and never uttered the forbidden word ‘landslide’ even if it was to be.

A landslide or a clean sweep is where the winning party wins twice as many seats than the second largest party.

Things probably would change if AI (Artificial Intelligence) were the one to poll and tell like in the US.

SEE ALSO:
Is India's 2019 election really a foregone conclusion?