Opinion polls don't just reflect but can shape a campaign
General election 2017 analysis: opinion polls show a swing in favour of Labour, but we should be aware of the poll effect
Opinion polls in Scotland before the general election 2017 - Image credit: PA Images
When Theresa May called an early general election on 18 April, it was clear she was confident of an easy and predicable landslide victory over a seemingly weak, divided Labour Party and an ‘unelectable’ opposition leader.
That was the whole basis of calling the snap election after all, regardless of what she actually said.
While Jeremy Corbyn’s popular support among Labour members – as opposed to the parliamentary party – was clear from his victories in two leadership contests in two years, gaining 59 per cent of the votes in 2015 and 62 per cent in 2016, the narrative has always been that no one could consider making him PM.
Worse still, many of the comments suggesting he is unelectable have come from within his own party.
- Theresa May’s campaign compared to Iain Gray's ‘subway moment’ by election expert
- Lies and smears have blighted this election
- General election: Lord Ashcroft discusses focus group results of previous SNP, Labour and Lib Dem voters in Edinburgh and Aberdeen
But the last couple of weeks have highlighted again, as if we needed reminding, that elections are not predictable, and one wrong step in the course of a campaign can swiftly turn the tables.
The key turning point was the publication of the Conservative manifesto and the announcement of an unexpected policy that the cost of social care for pensioners in England was to be met by selling their houses after their death, with only £100,000 of the value protected.
It was a policy seemingly not tested on the public or even members of the Conservative Party, and perhaps borne of complacency over the party’s unassailable position.
While the policy itself would not apply in Scotland, with social care being devolved, the fall-out from that, and the subsequent semi U-turn a few days later to promise an – unspecified – cap on how much people will have to pay, the damage may have already been done.
The U-turn allowed people to ask whether perhaps rather than being ‘strong and stable’, Theresa May was in fact ‘weak and wobbly’ and Jeremy Paxman to suggest in a TV interview that those in the EU May is to negotiate with over Brexit might conclude that she is “a blowhard who collapses at the first sign of gunfire”.
Former Tory chancellor George Osborne, seemingly enjoying his new-found role as a newspaper editor to get his own back on former colleagues, inflicted further damage by turning on his own party’s policies, including the suggestion that no senior member of the Cabinet supported the Conservatives’ migration target of tens of thousands and that it would harm the economy – despite the fact it was in his budget.
His newspaper described the manifesto as “the self-inflicted wound of the most disastrous manifesto in recent history”.
Tory candidates too have criticised the campaign. Speaking anonymously to the Huffington Post last week, one spoke of the “unforced error” of the dementia tax U-turn, while another described the whole campaign as a “clusterfuck”.
Savaging of one’s own party leader, it seems, is now not just the preserve of Labour.
Jeremy Corbyn, on the other hand, has been in his comfort zone on the campaign trail, addressing gatherings of supporters around the country and largely coming across well in TV appearances, despite some notable gaffes over costings from both himself and Diane Abbott in interviews.
Most notable was the tactical win of last week, with Corbyn announcing on the day of the BBC leaders’ debate that he would, after all, participate, leaving Theresa May in the impossible position of looking weak if she didn’t turn up – or even if she did.
The Labour manifesto, too, has included popular policies such as a £10 minimum wage, lifting the public sector pay cap and increasing free childcare hours in England, while addressing areas that have been the preserve of the Tories and UKIP – proposing to keep Trident, reduce migration and continue with Brexit.
The opinion polls suggest there has been a swing towards Labour, although it seems unlikely to be enough for the party to win the election.
The first poll after Theresa May called the snap general election in April, an ICM poll for the Guardian had the Conservatives on 46 per cent and Labour on 25 per cent, a gap of 21 percentage points.
Further polls during April saw similar results, with one outlier by ComRes on 19-20 April even giving the Tories a 25 percentage point lead.
However, after the Conservative manifesto launch on 18 May, there began to be a downturn for the Tories and an upturn for Labour in successive opinion polls.
Labour had already begun to increase in popularity throughout the first fortnight of May, with most polls putting Labour in the high 20s or low 30s, although the Conservatives continued to poll in the high 40s in the majority of polls.
But after 18 May, successive polls, with a few exceptions that were a percentage point or two higher, saw the Conservatives on 43 or 44 per cent and Labour closing in to around 10 percentage points behind at 34 per cent, less than half the gap that it had started on when the election was called a month earlier.
Two polls at the end of May, one by ORB for the Sunday Telegraph and one by Survation for Good Morning Britain, have put the gap at six percentage points and one, by YouGov for The Times, at five percentage points.
A YouGov poll taken on 30-31 May even found the gap had narrowed to just three percentage points, with the Conservatives on 42 per cent and Labour on 39 per cent.
However, other polls during the last couple of weeks of May showed a continued larger gap of 12-14 percentage points.
How that might play out in terms of seats is difficult to predict with the first past the post system. Theresa May’s hope of gaining a bigger majority still looks possible, but by no means as certain as it did when the election was announced.
Meanwhile in Scotland, an Ipsos MORI poll for STV published last week put the SNP on 43 per cent, down seven per cent on 2015, Labour and the Conservatives on 25 per cent each and the Lib Dems on five per cent among those who said they were certain to vote.
Running those percentages through Weber Shandwick’s ScotlandVotes predictor has the SNP holding onto 51 seats – down five from 2015 – with the Conservatives gaining three of those to increase their seats to four and Labour and the Lib Dems, gaining one more each, on two.
Mark Diffley of Ipsos MORI said: “Our poll for STV picks up some clear signals from voters. The SNP remains likely to win the most votes next week but is unlikely to repeat its unprecedented success of 2015.
“In vote number terms, the biggest beneficiaries appear to be the Conservatives, whose share of the vote looks likely to increase significantly from its performance in 2015.
“They will be hoping that this translates into a significant number of seats turning from yellow to blue on June 8th.
“For Labour, the poll confirms some recovery, as seen in GB-wide surveys, where it may return to something approaching the result from 2015 in Scotland.”
This election was always going to be a difficult one for the SNP, with any reduction from its landslide of 2015 being spun as a loss of support for independence, and Tories pitching for SNP voters opposed either to a second independence referendum or Scotland remaining in the EU.
However, opinion polls themselves can influence as well as project the result of an election.
In a series of tweets following a controversial YouGov prediction last week that there could be a hung parliament, former Labour minister Douglas Alexander, who ran Labour’s 2015 general election campaign, warned against being too influenced by them.
He tweeted: “Polls are the compass by which journalists and party strategists navigate the news agenda during the short campaign.
“In 2015 journalists and party strategists believed – on the basis of the polls – that we were in a 35/35 [per cent] election.
“The Tories’ Messina numbers were apparently more accurate – but didn’t shape any of the public debate.
“The public polls meant the media focus of the whole campaign was on a minority Labour Government & the issues that raised.
“These polls also meant there was almost no focus or scrutiny on what a majority Tory Govt would do – e.g. on Brexit.
“Then, at 10pm on polling day, everyone (the PM included) suddenly realised we had been in a 37/30 race. The compass had been wrong.
“For journalists & politicians it was like going to bed expecting to sail into New York and waking up to find yourself in Vladivostok.
“Today’s headlines scream about the possibility of another hung parliament. My best advice is to treat them with extreme caution.
“Journalists should be wary of doing what should be a party’s job for them – of motivating supporters and shaping debate.
“Polls don’t just reflect but shape a campaign. I offer this advice as the guy the 2015 polls said would be the next Foreign [Secretary].”
So with that in mind, will polls that predict the Conservatives doing less well than hoped across the UK mobilise Tory voters who might not otherwise have bothered to vote, thinking their party was a shoo-in for a landslide?
Or will it encourage Labour voters who might have concluded the party didn’t have a chance to turn out and support them? Can the party persuade its much-needed youth voters to the polling station?
And in Scotland, can Labour gain from Corbyn’s recent success across the UK, will the Conservatives revive their fortunes after a couple of decades in the wilderness or will predictions of peak-Nat turn out to be premature?
We’ll find out on Friday.
It’s a risky business to provide any sort of prediction of what will happen next in the twists and turns of Brexit
It is understood Michael Gove has rejected the role of Brexit Secretary
Sketch: The Scottish Tories are absolutely not planning to send children out to work on farms
Raab said he was unable to support the terms of the proposed Brexit deal