Thread 2: Do the locals tell us much about a general election?

Thread 2: Do the locals tell us much about a general election?

    Can we extrapolate in the way the Sunday Times does?

sunday times local results.JPG Both the Sunday Times and the Sunday Telegraph carry projections this morning about what Thursday’s results would mean in general election terms.

In the Telegraph Professor John Curtice of Strathclyde suggests that the CON-LAB-LD split was 40%-27%-26% which he suggests would give the Tories an overall majority of 20 seats.

The Sunday Times projection of a 54 overall seat for the Tories is based on work by Professors Colin Rallings and Michael Thrasher, of Plymouth University’ who suggest that the Tories were on 40% national (up one on their calculations last year) and that Labour remained static on 26%.

    But local elections are very different from general elections when Labour finds it easier to mobilise its vote. And of course local elections are often very much determined by local factors which don’t weigh as heavily on voters’ minds when they are voting on what should happen at Westminster.

Turnout makes a massive difference and in the locals this is usually quite low. In such a context it is the best organised parties on the ground that usually end up making most progress. In a general there’s much less variation and the issues are very much determined by how the national media is reporting things.

The other “unreal factor” this year is the Labour leadership. Nobody can really forecast how Labour will go down with the voters when Tony Blair is out of the way. There have been many polls but we’ll have to see how electors take to Labour’s new leadership before coming to firmer conclusions.

Congratulations to ICM.
The pollster for the Scotsman, ICM, appears to be the clear winner in predicting the Scottish parliament election. The firm had an average error of 1.6% on the constituency vote against Yougov’s average error of 2.8% and Populus’s 3.2%.

With the regional list ICM recorded a 1.5% error; Populus 1.3% and Yougov 1.8%.

I have not bothered to calculate the error from non-BPC listed, MRUK, but if the firm or the paper that published the polls had taken me up on my £1,000 wager offer I would have won easily. As I said at the time – ignore.

Mike Smithson
author of “The Political Punter”

Comments are closed.