Poll does not prove Brits want to quit the EU

“100,000 say no to EU: Poll proves Britain wants out”

Daily Express, 11 February 2016

“92% want to quit the EU”

Daily Express, 5 February 2016

An online poll hosted by the Express online has made the front page of its daily newspaper twice in the last week.

The survey isn’t representative, and so the fact that over 90% of those who completed it said they’d prefer to leave the EU doesn’t tell us much about what “Britain” or “Brits” want. Last week’s “92%” headline drew some criticism for this, and although the headline is now a bit more cautious in saying instead that 100,000 had voted to exit the EU, it still says the poll “proves” that Britain wants to leave.

The ‘remain’ side has been slightly ahead for the past four months, according to the “poll of polls” published by non-partisan information site What UK Thinks: EU. This is the average level of support across the six most recently published professional polls. At the time of writing, ‘remain’ has 51% and ‘leave’ has 49%.

Pollsters might not have the best reputation at the moment, after they consistently underestimated the Conservative’s lead over Labour in last May’s general election. An inquiry into the issue found that Conservative voters were underrepresented in poll samples.

Different methods produce different results—so far in the EU referendum campaign we’ve seen that telephone polls are more likely to find a lead for the ‘remain’ camp than online polls are.

So no poll, or poll of polls, should be taken as the definitive answer on this.

But professional pollsters do at least try to get a representative result, for example by choosing people at random. Online polls such as the Express’s, on the other hand, are completed by whoever wants to complete them. Where they are featured and what they ask will attract specific types of people, and so represent their views rather than that of the average voter. 

We are seeking a correction from the Express.