But why is this against the trend of other recent surveys?

A new poll by Ipsos-MORI for the Sun today goes completely against the run of other recent surveys and reports that Cameron’s Tories now have a 10% lead.

The following are the shares with changes on the last survey from the pollster before Christmas – CON 42(nc): LAB 32(-3): LD 15(+1)

After a period when Populus, YouGov and ICM have all reported Labour deficits of within 5% today’s survey goes very much against the way things seemed to be going. It also comes after a quiet period for UK politics in general and for the Tories in particular. Generally the main opposition party does better when it is making the news.

Given the effort that Brown has put into the New Year with the “re-launch” that cannot be called that this could be quite worrying. This week he has hardly been off our TV screens with one announcement after another which he has made personally rather than leaving it to the responsible minister like Alan Johnson.

My reading is that Ipsos-MORI’s very strict turnout filter might be a factor. The pollster only includes in its headline figures the voting intentions of those who are 100% certain to vote. ICM includes those who rate their certainty from 70% and above while Populus includes those of 50% certainty or more and weights their response in line with the certainty number. YouGov does not have a turnout filter.

One consequence of the Ipsos-MORI approach is that their non-voting intention questions are based on the responses of a different and broader group of people than the voting intention ones. So we have the apparently contradictory ratings of Brown doing better on “most capable PM” and “best in crisis” in a poll that sees his party 10% behind.

Ipsos-MORI need to get their act together on this one. If the views of the less than the “100% certains to vote” are not deemed important enough to include in the headline figures then the same should apply to other findings. At least we ought to be shown the comparative data.

So good news for the Tories and bad news for Labour. As we always say let’s see what the next poll reports – with a bit of luck there might be something new in the Sunday papers.

UPDATE Some observations from Ipsos-MORI: I’ve had the following email from the firm’s Julia Clark:

Just a little reminder to watch the share rather than the lead. As you know, with a margin of error of +/-3%, our poll today for the Sun is not as far out from the recent Populus poll (and other recent polls) as you imply in your commentary. We have shown an increase in Tory share (Populus was at 37% and our poll shows the Tories on 42%), but the Labour share (33% to our 32%) does not show a notable change for Labour over the last week.

I believe that issues like Peter Hain’s donation scandal could well have given the Tories a boost, as they represent a continuation on from the scandals that rocked Labour before Christmas. Labour continues to flounder a bit, and remain just above their ‘core’ support level of 30%. My point here is that Labour support is unlikely to drop BELOW this figure because it represents the core of Labour voters — however ‘floating’ and otherwise unaffiliated voters who are unhappy with the problems that Labour party is having are more likely to allocate their vote to the Tories or LibDems — and it looks as though they have chosen the Tories.

FURTHER UPDATE: The fieldwork for the poll was carried out on Wednesday and Thursday and MORI have taken on board my point about the non-voting questions and will provide comparative data on their site on Monday.

Mike Smithson