It’s 53-47 after second preferences

This week’s evening Standard YouGov poll on the London Mayor battle and shows a a small increase in Boris’s lead on first preferences. The figures are with changes on last week – BORIS 44% (-1): KEN 37% (-2): PADDICK 12% (+2)

After second preferences the lead is 6%.

This should provide some comfort for the Tory camp following a week a heavy exposure amongst Londoners for the campaign.

UPDATE Meanwhile I have been in contact with the MRUK polling firm which in yesterday’s Sunday Times reported that Ken had a 1% lead on first preferences but were neck and neck after second preferences were taken into account.

MRUK is applying for membership of the British Polling Council and in the meantime will be providing the full data from their political surveys as though they were a member. The information from the S Times poll will be up on Tuesday.

In the meantime they have provided me with some details from their London survey which show that a past vote weighting formula was adopted to ensure a politically representative sample and a “certainty to vote” filter was applied. In this case only those rating themselves as 8-10 were included in their final numbers.

Before the filter Ken had a 3% lead on first preferences – after it had been applied this was down to 1%.

The voting intention question did not prompt by candidate or party name and this is probably why Brian Paddick only registered 9%. There’s a lot of evidence that Lib Dems do better when there are prompts.

The real surprise from the survey was the projected turnout. Of those polled 75% were in the 8-10 category while 60% were 10 out of 10. This contrasts with the 54% “certain” proportion in the last Ipsos-MORI poll and the turnout last time of 36.9%. MRUK say that if the poll had been about predicting turnout then they would have approached it differently.

In the key demographic group of the over 55s Boris was leading 54-35%. This figure, I believe, will prove to be critical. All the evidence is that the older age groups are much more likely to vote than other voting groups.

Overall I think that MRUK is in line with the other the telephone pollsters and, apart from the lack of prompts in their main question, I am reassured about the methodology.

Mike Smithson