Give 'em a break.

Photographer: Simon Dawson/Bloomberg

Early Elections for the U.K.? Don't Bet on It

Matt Singh runs Number Cruncher Politics, a nonpartisan polling and elections site that predicted the 2015 U.K. election polling failure.
Read More.
a | A

Within days of taking office, British Prime Minister Theresa May ruled out holding an early general election to seek a mandate of her own. But the strength of her Conservative Party’s polling -- with leads of between six and 15 points in surveys conducted during September -- has resulted in continued speculation. Indeed opposition parties have been preparing for (and even advocating) a fresh vote.

The attractions are obvious. The May government has a majority of 12 seats (effectively 16, counting abstentions) in the House of Commons, meaning that even a small rebellion among her party’s lawmakers is sufficient to derail legislation. An immediate general election would allow an increased majority, the argument goes, and remove any remaining risk of the Conservatives facing an opposition leader with stronger ratings than Labour Party leader Jeremy Corbyn.

The prime minister could also be forgiven for being haunted by the ghosts of predecessors who delayed going to the country for a mandate. James Callaghan had a poll lead in late 1978, but didn’t call an election and subsequently lost. Gordon Brown hinted at calling an election after the Labour’s Party conference in 2007, only for the Conservatives to retake the lead after their own conference a week later.

But there are difficulties and risks. The Fixed Term Parliament Act of 2011 prevents early dissolutions of parliament without a two-thirds supermajority in the house, or if a government loses a confidence motion. Corbyn had said that he would support an early poll, though it is less clear that he would be able to deliver the 100 or so non-Conservative votes needed. If he can’t, the government might need to vote no-confidence in itself -- a novel occurrence that would likely come with a political cost.

Assuming an election could actually be called, gauging the likely outcome is less straightforward than is widely assumed. There is no recent precedent for a government calling an election as early as this; every parliamentary term since 1974 has been at least three years 11 months. Earlier snap elections generally had limited success, the exception being Harold Wilson’s increased mandate in 1966, which was predominantly due to Labour being able to squeeze the third-placed Liberal Party.

This matters, because as the EU referendum showed, lack of historical precedent can make votes unpredictable, particularly when it comes to voter turnout. And turnout in a snap election at this time could be an issue, due to voter fatigue. A general election in the spring would be the third major vote in two years (the fourth in Scotland) in addition to well-attended regional elections earlier this year.

And against this backdrop, if an election is seen to be held for the purpose of increasing the Conservative Party’s majority rather than the national interest, the risk is that many voters would resent being dragged to the polls yet again. The latest YouGov poll supports this, with 46 percent of those polled saying they are against an early election to 36 percent in favor.

Theresa May’s Conservatives do have comfortable poll leads at the moment, and all the indications are that they would win an election in 2019 or 2020. However, most of the poll leads significantly larger than their 6.6 point winning margin in 2015 can be explained by those polls using favorable assumptions about likely voters. Such assumptions will probably hold in a “normal” general election, but in a snap general election under such circumstances, the danger is that they do not.

If a lack of enthusiasm among May’s voters -- or increased enthusiasm among her opponents’ voters -- leads to a more neutral turnout differential, then May might not increase her majority at all. To see an extreme example of how strong this turnout effect can be, consider midterm elections, where governments almost always do badly and oppositions usually gain, as swing voters and the less enthused stay home.

Even oppositions that lost by a landslide have won midterm contests, as Labour did at this year’s local elections. Indeed spring 2017 would be two years into the term, which is often a government’s weakest point in the electoral cycle. And there’s another reason the government may chose to wait: A proposed redistricting favorable to the Conservatives, that is scheduled for the next election, would not have time to be implemented, further denting the potential rewards.

Although an early election could have the desired result, the risks are greater than they look. Don’t be surprised if “no election” really does mean “no election.”

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

To contact the author of this story:
Matt Singh at matt@ncpolitics.uk

To contact the editor responsible for this story:
Therese Raphael at traphael4@bloomberg.net