Death to the swingometer.
Reading the political coverage in the UK press, it's the same old same old Labour versus Tory cage-fight, with a bit of Tory-lite or Labour-lite in third place depending on who's in power.
The election's always decided by a handful of floating voters in a handful of swing seats, they say, and right now the story is they're switching to the Tories. Except they're not.
Polls rarely tell you anything about the churn. If Labour are down 2 and the Tories up 2, what's actually happened? Perhaps that 2% moved straight over, but it's unlikely. Maybe 2% moved from Labour to the Liberals, while another 2% moved from the Liberals to the Tories. Some will certainly have joined the Labour column, even in times like this. 2% in that scenario is just a net loss, after all.
The closest you normally get to this is polls that ask about switches since the last election. Earlier this week Comres did just that, and the Independent reports the results here. More than a third of Labour's 2005 voters are now going to vote for another party, and of those, more are now voting Green than have gone Tory.
It makes sense: the disillusioned are likely to be those who wanted Labour to renationalise the railways or to stay out of foreign wars, and why would they switch to the Tories? It matters, too. If Labour are losing more votes to us that they are to the Tories, shouldn't they change their right-wing approach on these issues, rather than trying to stick to Blairism at all costs?
But this is a bit too complicated for most commentators, and for election night coverage too. Stay tuned for discussions that still feature that tired old swingometer. It's easier than reporting a true multi-party system.