Analysis

U.K. Election: Why Are Pollsters All Over the Place?

No one is predicting a Labour parliamentary majority – but there's one scenario in which Jeremy Corbyn could beat Theresa May and become prime minister

A worker prepares signs outside their polling station on general election day in London, Britain, June 8, 2017.
NEIL HALL/REUTERS

Pollsters in Britain aren’t enjoying particularly high regard lately. In the last three campaigns that took place in the United Kingdom, most of them missed. In 2014 they predicted a very tight race in the Scottish independence referendum. Those voting to remain a part of Britain won by ten percent. In the 2015 General Election they predicted that neither of the large parties would win a majority in parliament. The Conservative party won a small but decisive majority. Last summer they were saying that in the referendum on leaving the Europe Union, the Remain camp would win by a small majority. Fifty-two percent voted to leave

>> U.K.'s accidental election between accidental leaders >> 

In advance of this election, the polling companies have been working on their methodology. It seems that they have embarked on very different courses as their predictions on the results of Thursday’s General election widely differ. All the polls by the recognized companies released this week forecast the Conservatives as the largest party in the new parliament, but the size of their majority over Labour, ranges from thirteen percent to a single point

Traditionally, the pollsters have failed to correctly predict the Conservatives’ vote share. For over 40 years, in every election, whether the Conservatives won or lost, the pollsters underestimated their actual result. This chronic mistake has largely been explained by the existence of “Shy Tories,” too ashamed to admit to the pollsters that they are planning to vote for the “nasty party." Strangely, in this election there is near consensus among the pollsters that the Conservatives will have the largest share of the vote, with around forty-plus percent. The varying margins are due to their disagreements over the Labour Party’s share. 

Could the wide difference in the predictions of Labour’s vote be caused by “Shy Corbynites” – voters ashamed of admitting they actually support Labour’s radical and controversial leader Jeremy Corbyn? 

Another reason that it’s so difficult to predict Labour’s share is the high level of support the party is expected to receive among young voters, who according to the polls are three times as likely to vote Labour than Conservative. The problem with these voters is that they are not that likely to vote. In 2015 only 42 percent of the 18-24 age group actually made it to the polling stations. The over-65 group, where the Conservatives have the same levels of support Labour has among the young, are much more reliable and twice as likely to turn out. The polling companies which show the smallest margins between the two parties are “weighting” the young voters as much more likely to vote today than they ever have in the past. But there is no way of knowing in advance if they will indeed change their electoral behavior. 

Forecasting the outcome of British elections is made even more complicated by the fact that the national share figures can be misleading. In Britain’s regional first-past-the-post system, there isn’t one national election, but 650 separate ones for each seat in the parliament. Once the Conservatives have over 40 percent, their chances of gaining a relatively large number of seats currently held by Labour, by small majorities, even if the margin they have over Labour is of a few points, dramatically increases. 

Could a surge in young voters help Labour defeat such an outcome? Not necessarily, as many of the pollsters who have analyzed the polling data according to the constituency maps believe that most of the seats with large reservoirs of young voters are already safely in Labour’s hands, so they may not change the electoral calculus noticeably. On the other hand, many of Labour’s current seats in the north of England are vulnerable to a shift of a few thousand elderly voters or those who voted in 2015 for the far-right UKIP, toward the Conservatives. 

Theresa May’s government had only a slim majority of twelve seats in the parliament. She began this election campaign seven weeks ago with polls that predicted a margin of over twenty percent. That would have yielded the Conservatives over a hundred new seats, chiefly at Labour’s expense. Despite the disagreement between the pollsters, all of them have shown that margin drastically narrowing over the last few weeks. But the Conservatives still seem to be on the path to winning, with a somewhat enhanced margin. The average margin of the last polls carried out this week is around 7 percent and the projected Conservative majority is around the 50-seat range. On the other hand, if the handful of most optimistic polls for Labour, putting the margin at just 1-3 percent, materialize, the Conservatives could even lose a few seats and Theresa May will not have a majority. She could still remain prime minister in a “hung parliament” but every vote on a piece of legislation would mean haggling with opposition parties for their support and she would be vulnerable to a no-confidence motion. 

None of the pollsters are predicting an alternative scenario where the Labour Party gains a parliamentary majority, but in a hung parliament, he could conceivably become prime minister with the support of the Scottish National Party, the Liberal-Democrats, the Greens and some of the regional Welsh and Northern Irish parties. It would still be a massive surprise. 

Either way, the only thing that is certain, is that when the results are finally out, at least some of the pollsters will be the losers (again).