Exit Polls Were Invented To Make Astrology Look More Scientific

The results of the recent state assembly polls have shown how problematic exit polls have become. Practically no pollster got his seat numbers completely right, sometimes even the broad trajectory.

Let’s take the India Today-Axis My India poll, which has been more consistently right than almost any other pollster. It got its Madhya Pradesh prediction of a Bharatiya Janata Party (BJP) sweep right, and also Telangana, but in both Rajasthan and Chhattisgarh it got the BJP’s victory trajectory wrong. Not only was the Congress not marginally ahead, but the BJP's victory went far beyond what punters had expected.

Almost all pollsters got Telangana right, in that they predicted a Congress victory, but all of them got Chhattisgarh wrong. The latter amounts to a monumental failure.

In Rajasthan, two polls actually predicted a Congress victory (including News 24-Chanakya), and in Madhya Pradesh, three suggested a small Congress victory, with ABC C-Voter giving the Congress seats in the range of 113-137.

The Congress’s actual final seat count was 66. That’s a horrendous mis-prediction. Even astrologers and Rahul Gandhi’s palm readers could have gotten better results.

Jan ki Baat got Telangana and Rajasthan right directionally, but did not anticipate a BJP sweep in Madhya Pradesh or Chhattisgarh.

The late economist and former US ambassador to India John Kenneth Galbraith once said that economic forecasting was invented to make astrologers look good. Exit polls in India, it appears, seem to have been invented  for the same purpose.