Or is he just clutching onto straws?

Reflecting, like just about everybody, on Gordon’s future the Independent’s political writer, Steve Richards, makes an assertion in his column today that “all the polls” suggest that the Tory leads are “soft” and that “..there has been no fundamental sea change” and that “as the last year has shown, fortunes can shift dramatically.”

I believe that this is a completely wrong interpretation from someone who gives that impression that he is is desperately trying to find something that will portray Labour’s plight in a more positive way.

Firstly all the five pollsters that carry out regular surveys of political opinion in the UK are showing broadly the same picture with Tory leads ranging from ICM’s 15% to the 22% from ComRes.

This has remained fairly consistent since the March budget. From April 23rd onwards every single poll from every single pollster has had Labour with double digit deficits. In that period, too, all the polls have had the Tories in the 40s and almost all of them have had Labour in the 20s.

But this is not just a new phenomenon. In the eighteen months between Cameron’s election as Tory leader and Brown’s coronation every poll, bar one, that measured measured voting intention with the two against each other had Labour doing worse with Brown at the helm. What’s happened can just be seen as a continuation with the short interruption of Brown’s media honeymoon.

Then there is also what I have dubbed here on many occasions – the “golden rule of polling”. Based on every single general election, London Mayoral election and by election where there has been polling for nearly two decades “the most accurate poll has been the one showing Labour in the least favourable position”.

Last week in Glasgow the rule was tested again. There were two polls, remember, one with a 14% Labour lead and another with a 17% lead. We all know what happened.

Finally there is the John Major test. Taking ICM, the only directly comparable pollster with those we have today, the Tories only dropped below 30% twice in the monthly Guardian surveys between July 1995 and election day in May 1997.

“Soft” – I know what’s soft about the Richards’s article and it is not the polling.

Mike Smithson