How accurate have voting intention polls been in past elections?
How accurate have voting intention polls been in past elections?
With the general elections scheduled to take place exactly six months from today, attention is focused on the polls.
Current YouGov figures show that if an election were to be held tomorrow, 32 per cent of those asked would vote Conservative, 33 per cent Labour, 8 per cent Lib Dem and 15 per cent UKIP. But how trustworthy are polls like this?
History has shown that people often depart from their intentions when the time comes to actually cast their vote, especially when there are big changes at stake.
When YouGov reported a 2-point lead for the YES campaign 12 days before the Scottish referendum it electrified the campaign. The poll results were a high for the SNP, the moment when they were finally taken seriously by Westminster. The SNP eventually ended up 10 points down, prompting the Mail to brand the poll ‘rogue’ – and a threat to taxpayers’ money.
Six months before the 2010 election, YouGov showed that 41 per cent of voters intended to vote Conservative, compared to 28 per cent Labour and 16 per cent Lib Dem. On the eve of election polls showed the Conservatives at 35 per cent, Labour sat 28, and the Lib Dems with a massive surge to 28 per cent. When the results came in, Conservatives had a 36.1 per cent lead, Labour were on 29 per cent and Lib Dem were on 23.
This suggests that people are potentially more likely to vote for an outside party in theory than in practice; despite the huge escalation in public interest leading up to the elections, the Lib Dems ultimately received less votes than the two traditionally leading parties.
To conduct voting intention polls, YouGov prompt participants with the names of the three main parties. If the participant says they would not vote for any of these, they are asked to choose from a second list that includes secondary parties like UKIP and Greens – although UKIP are currently being offered in the first round of options, a highly significant shift.
Six months before the 2005 election, 24 per cent of those asked by YouGov said they would vote Conservative and 26 per cent said Labour. 17 per cent said that they would vote Lib Dem. On election day, Labour got 36 per cent of the votes, the Conservatives 33 and the Lib Dems 22.
There are other factors which mean that even consistent intention polls may not correspond with results. The Conservatives have long claimed that the electoral system gives an unfair advantage to Labour. In 2005, the Conservatives actually won by a narrow majority in England – 35.7 per cent to Labour’s 35.4 per cent – but because Labour took 286 English seats and the Conservatives only 193, Labour were victorious.
The question of electoral reform is on the backburner at the moment, but the current system does mean that polling individuals cannot always provide an accurate reflection of a likely election outcome.
Ruby Stockham is a staff writer at Left Foot Forward. Follow her on Twitter
13 Responses to “Six months until the election: we look at the polls”
Leon Wolfeson
You’re making up nonsense…Labour are straight-line austerity pushing neoliberals.
Leon Wolfeson
You’re abusing polling. Single polls are worthless, you can only look at trends. And yes, there often IS systematic bias…hence looking at trends!
Sigh.
Leon Wolfeson
Not seeing what’s different – it’s very clear that there will be zero change to austerity, for instance, you’re just making that up.
It represents straight-line neoliberalism of the sort I and many on the left will not vote for.
Guest
No incumbency effect is, actually, statistically noticeable in UK Westminster voting.
Tom
Labour haven’t been a party for the working class in a long time.
Their policy of mass immigration has had a detrimental effect on the lives of all working class people.