Why did the pollsters get the EU referendum so wrong?

Britain was left stunned last month as the Leave campaign romped to a surprise victory in the EU referendum. Indications had been that Remain would prevail, with most polls projecting a narrow lead for staying in Europe. Indeed, a poll on the day suggested a four-point lead. The reverse was true

The EU referendum was another blow to the integrity of pollsters, who were heavily criticised for missing the mark in the 2015 general election. It wasn’t just the pollsters who got it wrong this time, though. Bookmakers were offering odds of up to 80% for a Remain vote. The financial markets were equally confident.

The collective failure points to error-prone methods and a fundamental misinterpretation of the changing political landscape.    

Phone polling unreliable

One of the biggest flaws seems to have been with polling conducted over the phone. When given a straight in or out option, undecided interviewees often felt pressured, and hastily answered with the status quo. As a result, phone polls always recorded a lead for Remain.

However, online polls gave a much more accurate reading. It correctly showed that many of the undecideds would indeed vote Leave come election day.

High turnout

It was previously assumed that a high national turnout would benefit the Remain side. This was not the case.

The reason is that many pro-remain regions had a lacklustre turnout, such as London. In contrast, it was the Eurosceptic populace of smaller provincial towns that turned up in force. These were groups that had previously shown a poor turnout in general elections. Pollsters assumed that the pattern of previous elections would be repeated.

They misjudged the emotive nature of this referendum, and the Leave campaign’s ability to galvanise the politically apathetic.  

Outdated political assumptions

Many commentators assumed that the traditional Labour constituencies would follow the party line and vote Remain. In reality, Labour’s influence in these areas was shown to be in serious doubt.

This was particularly evident in the old industrial heartlands of the North East. Sunderland voted decisively for Brexit.

At the core of this is a fundamental disillusionment with mainstream politics. Working class voters outside of the big cosmopolitan centres feel disaffected and are increasingly attracted to UKIP’s nationalist, populist rhetoric. They were the ones who ultimately swayed this election.

Against the odds  

The bookies were even further off the mark than the pollsters. There were odds of up to 80% for a Remain vote. In fact, betting firm Unibet distributed a press release on the week of the referendum arguing that the betting markets were more accurate for calling election than polls.

That may have been true for the 2015 general election. However, the odds were off this time around. More individual bets were placed on Leave. However, more money overall was placed for Remain. This reflects the fact that wealthier punters, who could afford to wager more, were demographically more likely to support Remain themselves. Their money skewed the odds.

Hot air from the markets

The financial markets were also fairly confident of a victory for the Remain camp. The Pound Sterling surged before the results were announced. As the truth became evident through the night, the markets were thrown into a frenzy, with the Pound plummeting to a 30-year low.

Many large financial firms conducted their own private surveys and exit polls. They seemed to have replicated the same mistakes as the mainstream pollsters.  Furthermore, the markets tend to pick up on the general moods and consensus. With the media and pundits forecasting a vote for the status quo, speculators piled in for Sterling.

As we have seen many times, the markets are prone to a herd mentality, and overvaluing unreliable causes. In the post-referendum aftermath, the bubble truly burst.