With a week to go until the British general election, experts are trawling through opinion poll data for some clues to a race that remains stubbornly deadlocked, and the result wide open.
Prime Minister David Cameron’s Conservatives and the opposition Labour party have been polling within a few percentage points of each other throughout the campaign, both hovering at about 34 percent support.
If neither party moves ahead in the final week, then neither will secure a parliamentary majority and will have to negotiate with smaller parties in order to govern.
The exact make-up of the House of Commons remains difficult to predict, however, due to Britain’s “winner takes all” system that means votes are decided not on the overall popular vote but on the individual outcomes in each of the 650 constituencies.
The fragmentation of British politics also makes life harder for the pollsters, as voters are increasingly ditching tribal loyalties and could change their minds on the day.
In the 2010 election, the polls were largely accurate in predicting the Conservatives’ share of the vote, but all underestimated Labour’s support and overestimated the third party, the Liberal Democrats.
This year one certainty is that the Scottish nationalists will win most, if not all, of the 59 seats in Scotland, up from just six in the 2010 vote. Beyond that, experts can only agree that the result is the most unpredictable for a generation.
Nate Silver, the US polling supremo who accurately predicted Barack Obama’s election as president in 2008, admits he got it wrong in the 2010 British vote because of the difficulty in translating the national vote to a seat count.
For the first time in this election, huge sums have been spent on polling in individual constituencies, in the hope of taking a more accurate pulse of the
It has been funded by Lord Michael Ashcroft, a billionaire businessman and former Conservative party donor who says national polls should only be viewed as “mood music”.
Mike Smithson, founder of politicalbetting.com, goes so far as to call national polls “irrelevant”, saying: ” The election is determined by individual battles in 650 seats.”
Ashcroft’s polling has reinforced predictions that the anti-European, anti-immigration UK Independence Party (UKIP) is unlikely to do as well as national polls predict.
UKIP has been polling in third place behind the Conservatives and Labour, but experts suggest it could only win between three and five seats.
Other factors are harder to predict, not least because of the growing voter tendency to ditch traditional party allegiances.
Gideon Skinner, head of political research at polling company Ipsos MORI, said that currently the Conservatives have the edge on the economy, and Labour on public services.
It “makes it difficult for either to have a clear breakthrough because nobody owns everything”, he said.
Another unknown is the impact of incumbency, a factor that could be crucial for the hopes of Cameron’s junior coalition partners, the centrist Liberal Democrats.
The Lib Dems are polling in fourth place and could lose half their 57 seats, but some believe they may do better than expected because of local support for constituency MPs.
“Normally you reckon incumbency is only worth a maximum of 1,500 votes,” said political commentator Iain Dale.
“I think that’s slightly different in Liberal Democrat seats—I think they can increase that quite substantially.”
Once the votes are cast, the situation will likely remain murky for days if not weeks, as the Conservatives and Labour scramble to form a parliamentary majority.
Some pollsters expect a small, last-minute surge for the Conservatives, but even if they win the most votes, Labour leader Ed Miliband could end up prime minister as he can count on the support of the Scottish nationalists.
Silver is again trying to analyse the British election but admits it remains a tough call.
“Sometimes you can gather all the data you like and the best forecast is ¯_(ツ)_/¯”, he wrote in a blog posting this week, using the Internet symbol for shrugging.