The past few days have been a period of crisis for both pro-Europe Tory and Labour leaders in the United Kingdom in the run-up to the June 23 referendum on Britain’s exit from the European Union: The British electorate looks increasingly willing to cast off EU membership. Just this past week, six consecutive polls have showed the “Leave” campaign ahead of “Remain” by as little as three points and by as much as seven.
If there’s hope for the British establishment, it may be found in, of all places, the betting markets. While betting markets have tightened substantially in recent days, they’re still implying a roughly 60 percent chance that Britain votes to remain. But why should observers trust bookmakers more than pollsters?
Polling has had a checkered record in the U.K. over the last few years. On the day before the 2014 referendum on Scottish independence, the Financial Times had the “No” side—voting to remain in the U.K.—up four points in the polling average. “No” ended up winning by ten. Things were no better in the general election the following year, when pre-election polls predicted the Conservatives would lose much of their plurality to Labour, imperiling the coalition government between Conservatives and the Liberal Democrats. In fact Conservatives outperformed the polls by about six points, gaining twenty-four seats and an absolute majority in Parliament.
Although not quite astute enough to foresee the Conservative upset in 2015, the markets have been modestly more accurate than the polls over the last two big elections. In the Scottish referendum, the betting odds consistently predicted about a 75 percent chance that Scotland remained in the U.K, even as the polls predicted a tightening of the race. And in the 2015 general election, bookmakers estimated about a 9 percent chance of a Conservative majority, while many election prediction sites, like FiveThirtyEight or ElectionForecast, considered an outright majority for either party extremely unlikely given the polls.
The betting markets might be more accurate than the polls because they are aware of certain biases in British polling. One common explanation for the surprising margin in the Scottish independence vote was that the large number of undecided—relatively less informed and less enthusiastic—were collectively inclined to support the status quo. Given that about ten percent of the electorate reports being undecided in the current polls on the Brexit referendum, the betting markets might be accounting for the likelihood that these undecideds will disproportionately vote Remain.
But another theory of British polling, the “shy Tory” theory, points in the opposite direction. Since the 1992 general election, when opinion polls massively underestimated the Conservative share of the vote, there has been speculation that Conservative voters in the United Kingdom, reacting to the left-wing leanings of much of the British media, may be less likely to disclose their voting intentions to pollsters. This could help explain both the Scottish referendum (Tories generally opposed independence) and the 2015 general election.
But if the “shy Tory” theory holds true this year, it suggests the polls are actually underestimating the support for Brexit. After all, Conservative voters are substantially more likely to vote Leave than Labour voters. If Tories are disproportionately shy, then Leave voters are likely to be disproportionately shy as well, particularly given the level of vitriol directed at them by much of the British press.
On the whole, then, the polls suggest that the referendum is leaning towards Leave; the betting markets suggest that it is leaning towards Remain. Next week’s results will validate which indicator is the more useful.
Max Bloom is a student at the University of Chicago and an intern at THE WEEKLY STANDARD.
