!-- consent -->
Even for a site which wishes the Conservatives the very best, it is surely hard to dispute that the party has to date had a poor general election campaign.
The u-turn over the social care reforms – a surprising move that shifted the focus of the election away from Brexit, the Tories’ strongest issue – is the most high-profile sign of this, but it fed into a longer-running story about Labour’s climb in the polls.
For the titanic leads enjoyed by the Conservatives when the election was called – at one point they led Labour by 50 per cent to 25 – have been whittled down to much more narrow leads.
To an extent, Theresa May is the victim of expectations. Even just two years ago any leader who could get the Party bumping along in the mid-40s in the polls would have looked like little less than a saviour. But with the stars seemingly aligned for a 1997-style landslide in April, a solid two-figure majority is going to look like a serious setback.
So the crucial question is: are the polls right? And if they are, how much will it affect the result?
First, we should recall the school of thought which believes that polls are snapshots of the electorate’s mood at a particular moment, not a prediction of how they’ll actually vote. Less partisan voters may well feel better about one party after a string of good stories about it without it being enough to shift their allegiance on polling day.
Nonetheless, it’s generally accepted that if the polls are substantially wide of the actual vote on polling day, they’ve gone adrift somewhere.
Another long-standing theory, which Tories must surely cherish at the minute, holds that campaigns don’t tend to change very much. The fundamentals – perceptions of the leaders, economic competence, etc. – are ‘locked in’ well before the short campaign. Matt Singh writes in the Financial Times that June the 8th will put this notion to the ultimate stress-test.
But if this were true, how would we explain (at least the majority of) Labour’s polling surge?
Some psephologists have pointed out that there may still be ways that polling companies are over-estimating the Labour share. Will Jennings and Patrick Sturgis point out, for example, that the ‘snap’ nature of the election means that some pollsters may not have finished implementing the changes to their methodology after 2015.
Several others believe that so-called ‘risky’ voters – i.e. those less likely to actually vote, such as young people – may be being over-sampled. For example, Singh thinks that if you exclude people who didn’t vote in the Brexit referendum (a relatively high-turnout, high-interest contest) the Conservative lead “doubles to 16 points“. There is also the possibility that Labour may be piling up votes in unhelpful parts of the electoral map.
Election-based sample distortions of the sort suggested by James Morris, a former Labour pollster, may also account for some of the huge swings we’re seeing in places like Wales, where the Tories have gone from a record-shattering lead to their usual second place in a span of about two weeks.
Thus lies the case for the polls overstating Labour: pollsters may be over-sampling pro-Labour people; a dangerously high percentage of those may not actually vote; and those who vote may not vote where it really counts.
Finally, there’s the fact that reports from the front-line, and the deployment of each sides’ campaign resources, seem to suggest that both parties believe the race is less close than the polls suggest. Of course the plural of anecdote is not data, but it’s another possible sign of a polling miss.
Of course, that would never explain the entire picture. The election has given Corbyn’s usually-hopeless media operation a chance to actually cut through, and the Labour manifesto will have energised their base.
The lacklustre Conservative air war may also have failed to control his first impression on the many, many voters who are so tuned out of politics on a day-to-day basis that they may not even have heard of him until now. As Stephen Bush argues in the New Statesman, the Tories’ assumption that Corbyn was one of their insurance policies is now looking a bit suspect.
(Bush’s piece should also serve as a tonic to this one: whilst the polls may be over-selling Labour, they may not. The Tories may just be tanking.)
Absent a significant turn in the tide over the next week or so, it looks as if next month’s exit poll will be just as eagerly anticipated – and potentially embarrassing for the pollsters – as the last.