Or rather, different polling. Each night it’s the same: we await the latest polls to see if any air has leaked from the Clegg bubble or Cameron has regained the Big Mo or Gordon come alive and then we plug the numbers into Electoral Calculus or UK Polling Report or the BBC and see What It All Means in terms of how the numbers become seats at Westminster.
I wonder if most of it means very little. This goes beyond, but includes, the caveats we always apply about uniform national swings being a crude measurement and all the rest of it. Consider Scotland – a useful study and not just because it’s the part of the country I know best – the other day I posted about a poll from which it was possible to draw the conclusion that no seats or perhaps just two or three would change hands at all. The topline figures were Labour 40, SNP 20, Lib Dems 19, Tories 16 but this, on reflection, only scratches the surface of what may or may not be going on.
Consider these seats: Ochil & South Perthshire (SNP need a 0.6% swing from Labour), Aberdeen South (LD need a 1.6% swing from Labour), Angus (Tories need a 2.2% swing from SNP), Perth & North Perthshire (Tories need a1.7% swing from the SNP), Dumfries & Galloway (Tories need 2.9% swing from Labour), Dundee East (Labour need a 0.5% swing from SNP), Edinburgh North & Leith (LD need a 2.3% swing from Labour), Edinburgh South (LD need a 0.5% swing from Labour), East Dunbartonshire (Labour need a 4.4% swing from LD), Dundee West (SNP need 7.3% swing from Labour) and Stirling (Tories need a 5.5% swing from Labour).
Each of these seats, and a couple more, is notionally “in play” next month but none of the national (ie, Scottish) polls offers much in the way of a clue to what will happen in any of them.
And the same is true for many seats and regions in England too. If the election remains a true three-horse race the utility of the national polling seems likely to diminish still further.
So, pollsters, here’s what I’d like to see: a poll of 20 Liberal-Labour marginals and another poll of 20 Tory-Liberal marginals and, heck, another poll of 20 Tory-Labour marginals.
Then we might have a better idea of what’s going on. Right? This would also seem more useful than just polling “The Top 100 Marginals”.
Comments