Top Pollster YouGov: How We Got the British Election So Wrong
One of Britain’s leading polling firms apologizes for its awful performance in the British election and welcomes a full industry-wide inquiry into their failings.
LONDON — Britain woke up to its very own “Dewey Defeats Truman” moment Friday when Prime Minister David Cameron did what the pollsters said was impossible and won an outright parliamentary majority.
The shock result had been virtually discounted by polling companies who carried out surveys in Britain on an unprecedented scale in the leadup to the general election.
The CEO of YouGov, which published daily polls in the weeks before Thursday’s vote, apologized and said the entire polling industry must now hold a joint inquiry into how they had called the result so drastically wrong. Stephan Shakespeare said it was too soon to say for sure why they had underestimated the Conservatives’ decisive share of the vote but he has offered a first detailed explanation. “The first thing we have to do is say we got it wrong. There’s no point in saying it wasn’t that bad,” he said.
Obama campaign veteran David Axelrod expressed his frustration with the polling. "In all my years as journalist & strategist, I've never seen as stark a failure of polling as in UK. Huge project ahead to unravel that," he wrote on Twitter.
Those going to bed before the exit polls stunned Britain at 10 p.m. BST on Thursday night would have fallen asleep in the knowledge that the outcome of the election would not be confirmed for days. There was a clear consensus that the country was heading for a hung parliament and maybe even weeks of brutal negotiations. Just like the Chicago Daily Tribune’s infamous declaration that Governor Thomas Dewey of New York would beat Harry Truman in the 1948 presidential election, the pollsters had apparently settled on an outcome far too soon.
The rise of a multiparty system has created a far more complex electoral landscape and Shakespeare told The Daily Beast that the first fully tactical British election had given rise to a new category of “shy” voters. The notoriously inaccurate British opinion polls of 1992 predicted that John Major, the Conservative prime minister, would lose to the Labour Party. Analysts would later blame “shy Tories,” who didn’t want to tell the pollsters that they were voting Conservative, for their mistaken belief that Neil Kinnock was about to enter office.
Ingrained tactical voting may now have given rise to a new breed of shy voter and it’s not as simple as respondents being reluctant to admit they would vote for a fringe party such as the U.K. Independence Party (UKIP), which politicians and pollsters had already factored into their calculations.
“Imagine that you are an ex-Conservative, UKIP voter and that’s who you want to win, but you’re sitting in a constituency where there’s no chance of UKIP winning—let’s say a Conservative-Labour marginal,” said Shakespeare. “You absolutely hate the idea of an SNP-supported Labour Party getting rid of your [European in/out] referendum and all these bad things. You’re highly conflicted, so you’re going to vote Conservative but you don’t like the Conservatives so you’re telling the pollster that you’re going to vote UKIP because it’s an emotional truth.”
That may explain unexpected votes swinging in behind the Conservatives, but in many seats there is no accompanying collapse of UKIP votes. “What’s interesting is that the UKIP vote held up, didn’t it?” Shakespeare said. “It’s very possible that there are shy UKIPers on the Labour side—the opposite effect.”
Shy votes trading in all sorts of directions based on the tactical considerations within a particular local constituency would make it difficult to accurately extrapolate national vote shares.
It then becomes even more difficult to forecast a seat projection in a first-past-the-post system.
So, has the explosion of multiparty politics and tactical voting made it impossible to accurately predict the number of seats each party will win? “It’s very, very, very hard as you’ve learnt,” said Shakespeare. “But usually when we know what the problem is and we prepare ourselves for it we can get better at that.”
Katherine Peacock, the managing director of ComRes, offered a different explanation to the BBC. She said ComRes and other phone polling companies—YouGov gathers data online—had shown a Tory lead all year and had been within the margin of error on the final vote share. She said things had gone wrong when that data was converted into seat projections. “I think we are very clever as pollsters at ComRes because we actually have academics who translate our vote share into seats for us—we pass that on to the experts to do that,” she said. “When we have been looking at specific seats in terms of constituency polls, we’ve been incredibly accurate.”
Shakespeare rejected the notion that YouGov’s online polling had been markedly different than the phone poll companies. “I think the whole industry is going to agree to have an inquest—that’s what I think we’re going to do because everybody got essentially the same result,” he said. “Even though actually the average error of us and the other polls was under 2 percent, it was in the wrong place; it gave the wrong story and that’s bad.”
Polling analysts and political pundits warned throughout the campaign that there could be a late swing to the Tories because of the public’s underlying feelings about the leadership of Cameron over Edward Miliband and a strong preference for the Conservatives on the economy. These factors may well have come into play, and certainly that was the Tory election strategy.
Professor Tony Travers, of the London School of Economics, told The Daily Beast: “I wasn’t convinced by shy Tories, but I guess I was wrong. It looks like voters may have shied away from change at the last minute.”
Unless that swing kicked in as millions of voters were walking to the polling stations, however, it should still have been picked up in the late polls.
In an election that was so nakedly tactical from the start, the polls themselves would have been crucial in informing people about which party to lend their vote to in a given constituency. The Conservatives also focused on dire warnings about the fallout from a Labour government supported by the Scottish National Party. It is thought that may have helped move undecided Liberal Democrats into the Conservative camp.
Shakespeare conceded that the polls, which may or may not have been accurate, could have affected late swings themselves. “It may be that the polls actually have an effect on how people vote. I don’t want to give the impression that we were sort of right. I’m not saying that at all. But you could easily think that maybe it was actually very close and people didn’t want the SNP and that swung people toward the Conservatives and especially those UKIPers,” he said.