bbc

A crumb of comfort for Michael Howard?

Most of the assumptions on how many seats each party is likely to get for a given vote share are based on appyling the predicted changes on a national basis and doing a computation in each seat to see what happens there. It’s rough and ready but, in the past, has been a reasonably reliable guide. The best known calculator is the one produced by Martin Baxter.

At the last two elections Labour and the Liberal Democrats have picked up seats over and obove the uniform national swing because of the willingness of supporters of both parties to switch allegiance in order to defeat the Tories.

The big questions for spread-betting gamblers and others trying to forecast the coming election are whether the same will happen again; whether the tactical switches of 1997 and 2001 will remain; and whether we might just see some tactical voting in the other direction with Lib Dems voting Tory to get a Labour incumbent out.

More data just available from last week’s Populus Poll shows that where Labour cannot win, 35% of current Labour voters would vote tactically; where the Lib Dems could not win, 37% of its current supporters vote tactically; and where the Tories 36% of its voters would vote do the same.

The interesting split is the Lib Dem 37% which goes 20% saying that would vote Labour to keep the Tories out and 17% would vote Tory to keep Labour out. That last figure is probably the most surprising and, if this actually happens on polling day, could see a few extra Tory gains over and above the national swing. It would also support the tactical vote unwind concept.

The willingness of more than a third of Tory and Labour voters to switch to the Lib Dems to keep the other main party out is good news for Charles Kennedy.

The spread -betting markets, meanwhile, have seen a shift to Labour. IG Index Commons Seats Spreads LAB 355-362 (+2): CON 190-197 (-2) : LD 68-72:

© Mike Smithson 2005