The polls are superficially comforting for Labour at the moment. Several may put support for the party at its lowest level since the days of Michael Foot but surely the vagaries of the electoral system will prevent a 1983-style meltdown?
Our own Poll of Polls currently puts the Tories on 33.8 per cent, the Lib Dems on 28.6 per cent and Labour on 27.6 per cent. If repeated at the election on a uniform swing, these figures would leave Labour as the largest single party in a hung parliament, 59 seats short of a majority. Labour would be left with 267 seats, the Tories with 259 and the Lib Dems with 93.
But here's the catch. The concept of uniform swing, which assumes that the swing to or away from a party will be indentical in every constituency, is a notoriously poor guide to predicting election results. Differences in constituency size and variable turnout mean that we rarely see anything like a uniform swing in practice.
With this in mind, the American psephologist Nate Silver, who predicted the correct result between Barack Obama and John McCain in 49 of America's 50 states, has devised an alternative method -- and it's not good news for Labour.
Silver's approach works by works by assigning shares of one party's 2005 vote to another. As he explains at his blog FiveThirtyEight.com:
[W]hat happens if 10 percent of people who voted for Labour in 2005 defect to the Conservatives, 15 percent of Labour's voters defect to LibDems, and 10 percent of the Conservatives' voters defect to LibDems?
Applying this to a series of possible outcomes gives us a radically different picture to that offered by uniform swing calculations. For instance on a uniform swing, if the Tories win 34 per cent of the vote, the Lib Dems 29.1 per cent and Labour 26.9 per cent this is the result:
Conservatives 271 seats
Labour 253 seats
Lib Dems 93 seats
But if we use Silver's method we get a very different result:
Conservatives 304 seats
Labour 214 seats
Lib Dems 101 seats
Such a result would leave Labour with just five seats more than Foot in 1983. Silver adds some disclaimers to his method, it does not account for voters exiting or entering the electorate and cannot directly account for tactical voting.
But all the same, it's a salutary reminder that Labour's election prospects are far worse than they may appear.