Shouldn’t there be separate polls and a separate swing calculation?

Whenever a poll comes out we all rush to make the same point that the Tories Cameron’s need to have a 5% margin to win most seats and 10% to secure a majority. This is finally getting through to the markets and now a Labour majority has now taken over from a Tory one as second place in the betting on the outcome of the next election.

Yet are we missing something? All these projections are based on the concept of a uniform national swing (the UNS) which includes Scotland where politics is developing very differently from the rest of Great Britain.

With the Scottish Parliament elections due to take place on May 3rd there has been a big surge in support for the Scottish National Party with the main three parties following a very different pattern from elsewhere. Yet pollsters and forecasters still seek to calculate national swings as though this is an homogeneous nation.

Surely the time has come for polls to separate off Scotland and for the main seat calculators from Anthony Wells and Martin Baxter to work on an England and Wales only basis.

Looking at the detail from recent polls that show separate Scottish data and generally you find that the Tory increases that we have seen nationally have not really happened North of the Border. The party seems to be at the same level or below the 15.8% of the vote that was chalked up in Scotland on 05/05/05.

Thus December’s Populus survey put the Tories on 12% in Scotland while YouGov, for the Telegraph had them at 14%.

So Conservative gains in Scotland that simply bear no relation to the political reality there are included in the calculations yet the overall Tory standings in England and Wales are being diluted by being lumped in with Scotland where they only have one seat.

From a statistical point of view there is a danger of distortion in stripping out the Scottish numbers because other weightings that apply on an overall basis might be affected as well. But it could be that Labour’s 10% “comfort zone” that the Tories need for a majority is actually 8-9%.

For this means that the Tories are doing badly in Scotland where it does not really matter to them – but are doing better in England where a 1-2% uplift could be crucial in many target seats.

As we get closer to the General Election the big betting focus will be on seats and issues like different parts of the Kingdom performing differently could be crucial.

The PBC POLITICAL FORECASTER OF THE YEAR COMPETITION. Entries have to be posted by 2359 GMT tomorrow night.

Mike Smithson