I was at work, but I saw a few folks on Twitter (especially Louise Mensch) debating the accuracy of the final exit poll predicting a big win for the Tories. Too tired, I didn't even blog Thursday night, but I did check the results at the British newspapers throughout the evening. David Cameron's party shocked the political establishment. Louise Mensch was imploring her partisans against gloating. Numbers guru
Nate Silver issued a major wake-up call on the state of elections surveys.
I have no idea what went wrong other than the pollsters got lousy samples, which underrepresented conservatives. Some of the articles I've read suggest that Tory voters were shy and refused to state their real voter preferences. I always doubt such stories, especially in this case, since there's no shame in voting for the incumbent party. It wasn't a race issue like that of the much-hyped "
Bradley effect." So I suspect that the "independent inquiry" that been proposed by the British Polling Council will basically be going back to the drawing board on basic methods. The group might do well to examine institutional left-wing biases among pollsters, a situation so serious that Survation, a market survey research firm, refused to publish a poll on the eve of the election showing the Tories holding
a 37-to-31 percent lead.
In any case, here's the New York Times from this morning's paper, "
British Election’s Other Losers: Pollsters":
LONDON — The Labour leader Ed Miliband may have stumbled badly in the British election, but there was another big loser on Thursday night: the pollsters who were far off the mark and failed to see the outright majority won by Prime Minister David Cameron and his Conservative Party.
Before the election, nearly every poll showed the race as a near-tie that would result in a hung Parliament and force complex negotiations to form a coalition government. In the last days of the campaign, a survey by Ipsos/MORI, a widely respected pollster, forecast that the Conservatives would win 36 percent of the vote and the Labour Party 35 percent. On Thursday, The Guardian reported a poll by ICM putting Labour at 35 percent and the Conservatives at 34 percent.
A consortium of researchers from the University of East Anglia, the London School of Economics and Durham University aggregated national polling and online surveys, and in its final projection on Thursday forecast that the most likely outcome would give the Conservatives 278 seats in Parliament and Labour 267.
The final result, with the Conservatives securing a majority and projected to win as many as 331 seats, only added to an intensifying debate in the United States, Britain and elsewhere about the accuracy of polling, the problems of getting accurate samples in the era of the iPhone when voters can no longer be reached as easily by traditional means like landlines and the fracturing of politics making it harder to predict voter behavior.
In Britain’s case, weeks of assumptions built around the consistency of pre-election polling gave way to a sense of shock among even veteran Westminster watchers when broadcasters unveiled the results of their exit poll right after the polls closed on Thursday night. The exit poll accurately predicted, within a few seats, the final outcome, but it was initially greeted with deep skepticism by party leaders and some voters.
Paddy Ashdown, a former leader of the Liberal Democrats, pledged to “publicly eat my hat” after the first exit polls suggested that his party would see its parliamentary ranks slashed by 10 members. “I have been offered 10 hats on Twitter tonight,” he told Andrew Neil of the BBC, “not all of them politely, I have to say.”
As the poor performance of the Liberal Democrats became more clear, Mr. Ashdown’s hat went viral on Twitter, with a fake account attracting more than 12,000 followers, and manipulated images of his eating a hat proliferated on social media.
The failure to accurately predict the result was reminiscent of recent elections in Israel, where Prime Minister Benjamin Netanyahu won a clear victory after pre-election polls showing the rival Zionist Union in the lead. Analysts attributed Mr. Netanyahu’s surprise comeback to an 11th-hour political offensive, including a pledge that there would be no Palestinian state.
Writing in The Guardian, Alberto Nardelli, the news organization’s data editor, said there was no simple explanation to what went wrong with the polling.
“It could be simply that people lied to the pollsters, that they were shy or that they genuinely had a change of heart on polling day,” he said. “Or there could be more complicated underlying challenges within the polling industry, due, for example, to the fact that a diminishing number of people use landlines or that Internet polls are ultimately based on a self-selected sample.”
Peter Kellner, the president of YouGov, a leading survey firm, told The Daily Telegraph that the pollsters had erred, attributing the results to the capriciousness of voters rather than to statistical lapses.
“What seems to have gone wrong is that people have said one thing and they did something else in the ballot box,” he was quoted as saying by The Telegraph. “We are not as far out as we were in 1992, not that that is a great commendation.”
He was referring to the British general election of 1992, when pollsters predicted a hung Parliament, only to see the Conservatives win an outright majority.
Later in the day, YouGov issued a mea culpa, though it declined to draw firm conclusions on what went wrong. “For any polling company, there inevitably comes a time when you get something wrong,” it said in a statement. “Every couple of decades a time comes along when all the companies get something wrong. Yesterday appears to have been one such day.”
Such was the concern over the failure of the pre-election polls to get it right that the British Polling Council announced Friday that it would set up an independent inquiry to determine what had gone wrong.
“The final opinion polls before the election were clearly not as accurate as we would like, and the fact that all the pollsters underestimated the Conservative lead over Labour suggests that the methods that were used should be subject to careful, independent investigation,” it said in a statement. It said the independent inquiry would “look into the possible causes of this apparent bias, and to make recommendations for future polling.”
More.
Also at the BBC, "
Election 2015: Inquiry into opinion poll failures," and "
Election results: How did pollsters get it so wrong?"
And at the Independent UK, "
Election results: What went so wrong for the pollsters – and how did the exit poll get it right?"
