28 October 2016
The European and British flags fly together in London.  Pollsters failed to predict the victory of the "leave" camp. Photo: Bloomberg
The European and British flags fly together in London. Pollsters failed to predict the victory of the "leave" camp. Photo: Bloomberg

Why Brexit pollsters got it wrong

Britain’s vote to leave the European Union has stunned the world. 

The catastrophic result of the referendum shows that the traditional ways of communication are badly in need of reform.

Brexit may even provide some lessons to the Legislative Council election in Hong Kong this September, as well as the presidential election in the United States in November.

The final opinion polls showed that the “stay” camp had 52 percent just before the vote began, and it gained even more in the last minute. Also, the odds from betting companies showed the likelihood of stay getting more than 60 percent.

The so-called smart money bet that many of those in the “leave” camp were just trying to exert pressure on the administration of Prime Minister David Cameron, that these people would not risk the nation’s future when the actual voting came.

In that sense, both investors and academics were confident that the United Kingdom would stay in the EU.

However, the result from the first major city Sunderland showed that the leave camp won 61 percent against the stay’s 39 percent, far above the forecast of 54 percent for leave and 46 percent for remains.

When the bitter truth finally hit the market, the sterling slumped by nearly 10 percent within just three minutes.

After the referendum, fast-rising pollster YouGov noted that the polling industry has increasingly pivoted towards online samples in recent years.

Certainly, online polls cost less, and it’s easier to access a large number of respondents online compared with traditional phone or face-to-face methodologies.

Also, these polling companies claimed that their online samples were more representative of the public opinion.

Generally speaking, the most credible polling organizations won’t depend solely on online samples. They would maintain a 30 to 50 percent share for online samples while relying substantially on phone and face-to-face interviews. Then they would adjust the weightings to more accurately reflect public opinion.

YouGov has earned its spurs from its forecasts in many major elections, and it’s regarded as the most reliable polling group. Its final poll showed stay winning 52 percent of the vote against 48 percent for leave.

The huge divergence among the different age groups has led to the surprising result. Up to 75 percent of young people between 18 and 24 voted to stay, while 56 percent of those aged 50-64 and 61 percent of those above 65 voted to leave.

As we know, the elderly are less reflected in online polls, and as such, their views were not represented in online samples.

In the meantime, most pollsters failed to include the turnout levels in their online polls.

During the actual voting, the turnout in leave areas was unexpectedly high, and this factor proved to be crucial to the outcome of the very tight race.

The outcome also revealed the arrogance of netizens. They had expressed their views and seemingly found consensus in social networking sites like Facebook and Twitter.

They had thought they represented the public opinion, and the voice of those who were less active online has been ignored.

It’s quite interesting to note that a large number of the leave voters now regret their decision. They had thought the UK was poised to stay in the EU, and their votes wouldn’t change that.

Some of the leave voters just wanted to express their opposition to the administration’s stance toward the EU.

Also, many voters had been misled by the leave camp and focused on the potential benefits of leaving the EU such as less foreign labor and refugees.

They failed to take into account the heavy price of leaving the union, such as the potential damage to the UK’s exports, financial sector and job markets.

In that sense, the voters had been ill-informed.

This article appeared in the Hong Kong Economic Journal on June 27.

Translation by Julie Zhu

[Chinese version 中文版]

– Contact us at [email protected]


Hong Kong Economic Journal columnist

EJI Weekly Newsletter