Monday, 11 May 2015

People versus Pollsters: The Unpredictability of Human Beings.

by David Danso-Amoako




The top 10 pollsters of the 2015 UK election all predicted that we would have a hung parliament. For a long time, David Cameron and Ed Miliband went over and under each other at the top of the opinion polls. Ipsos-Mori, Survation and YouGov among others showed us how tight the race to Number 10 was and the certainty of a Hung Parliament. This is summarized in the graph above. 

I for one believed them because of the complex formulae they use in predicting the results. But after the exit poll I began to seriously doubt them and I am sure David Cameron’s smile would have widened just as the gap widened. 

After the election took place an exit poll was produced to show a good representation of the seats. The exit poll asked people how they voted as they came out of the polling station. Their results were the closest to the final result of the election. When the 2015 exit poll was revealed at 10:00pm, indicating the Conservatives would be the largest party but would not have an overall majority, there was blue excitement in the air. The results were told to Theresa May MP Home Secretary. On hearing the result, she May said that she would be “very cautious about the results”. She was right, if political history is anything to go by.
In 1992, Labour was winning by 20 percentage points ahead of the Conservatives in opinion polls and the  country was entering recession. The increasingly unpopular Margaret Thatcher had been replaced by John Major as the Conservative leader. The polls suggested that Labour would be the largest party but without a majority; but the election results contradicted the pollsters like this election. The Conservatives won with a (small) majority.
  
In this election what the pollsters were not expecting was a record-breaking night for the parties. The Conservatives would go on to make a majority government with 331 seats whilst Labour party would go on to get their worst result since 1987 with 232 seats, losing some of their most prominent figures. After sleeping with the Tories, the Lib Dems may not have expected a nightmare result, losing 51 of their 58 seats making the party end up in shambles. The SNP got 56 out of a possible 59 in Scotland. UKIP kept one of their two seats after Conservatives gained Rochester and Stroud from Mark Reckless, a former Conservative who defected to UKIP. The results left the losing parties without leaders and pollsters mind-boggled.


The result will leave Labour, UKIP, the Lib Dems, and the pollsters thinking. The pollsters were very off, and when the results came, the MPs were right to be cautious about the result the pollsters gave.
It must be noted however that the pollsters managed to predict the pattern of the results even if they were not right. They predicted the SNP tsunami and the LIB Dem downfall. They managed to predict the Tory rise but not to the right degree.

The pollsters were right to a certain degree, which is debatable. However humans are very complex and cannot be easily predicted by a model. The pollsters got it wrong and in the future they have to refine their models and not be too certain about the outcome.

1 comment:

  1. Hi David, the pollsters are wrong again in America. Wonderful piece.

    ReplyDelete

Comments with names are more likely to be published.