October poll round-up: time to think local
All four of the main UK parties bounced around the polls this month with an unpredictability that prompted many a newspaper headline.
Labour rose as high as 38% and plunged as low as 30%. That kind of inconsistency – beyond the margin of error – suggested there was a lot of churn going on among the samples of voters used by the various polling agencies. The Tories also ranged over a gap of eight points, peaking at 36% and touching 28%. The Lib Dems shuttled between a high of 11% and a low of 6%, while Ukip varied between 13% and a remarkable 25%.
The averages for October show how such extremes (particularly in Ukip’s case) were mostly one-offs (or outliers, to use the polling jargon).
Labour’s average for the month was 33.9%. The Tories’ was 31.8%, Ukip’s 16.0% and the Lib Dems 7.9%. The gentle trend in the narrowing of the gap between Labour and the Tories continued:
Labour’s monthly average hasn’t been this low since 2010. Ukip’s has never been this high. The last time the Tories ranked lower than 31.8% was in July 2013. The Lib Dems are actually up on September’s figure, albeit by a tiny 0.2 points.
Type these averages into an online swingometer and you’ll get a result that implies Labour remains on course to be the largest party in a hung parliament. But you’ll also get a result that suggests Ukip won’t win a single seat and that the SNP won’t gain any more seats. The events of the past few weeks – Ukip’s success in Clacton, Labour’s crisis in Scotland – have meant both of these assumptions now look like being false. And I suspect the national opinion polls will continue to mask the success Ukip and the SNP may have in individual seats.
This obviously makes for a misleading depiction of what could happen at the general election.
October was the first month where the national polls took on a distinct touch of unreality. It has reminded me of the state of play in the US just before the 2012 presidential election. Here, nationwide polls were implying it was too close to call, and these kept generating the most headlines. But polls in the swing states turned out to be a much better reflection of what was really going on, which was that Obama was winning where it mattered, and he duly got himself re-elected.
Maybe this is the tactic to adopt in the UK from now on. We should think local, and not get too bothered by the national picture. For one thing, it’s increasingly obvious there is no national picture. Instead, the election is going to be a patchwork of mutually exclusive contests at a regional and constituency level.
Whether these will lead to one party emerging as an overall winner feels immensely unlikely. Whether even two parties together would command enough seats for a majority in the House of Commons is starting to seem doubtful.