YouGov/Sunday Times - CON 39, LAB 42, LD 9

Tonight's YouGov poll for the Sunday Times has topline figures of CON 39%, LAB 42%, LDEM 9%. It's the first time a poll has shown Labour back in the lead since the Conservatives moved ahead in the polls after the veto. Usual caveats apply - it could be that the brief Conservative boost in the polls has now started to recede... or it could be an outlier. Certainly today's ICM poll for the Sunday Telegraph (see post below) showing a six point Tory lead doesn't suggest that the veto bounce is subsiding. We'll have a better idea next week when we see both the YouGov daily polls for the Sun, plus the monthly Populus poll for the Times (and possibly the ICM/Guardian poll, assuming they aren't going to do it over Christmas weekend!).

While I am here, I expect less frequent visitors will be asking why there is such a difference between ICM and YouGov's figures, or more specifically, why there is such a difference between their Labour figures, as their figures for Conservatives support are almost identical. While some of this is probably normal random variation, and either an outlier against Labour from ICM or an outlier in favour of Labour from YouGov (or both!), part of this is also due to methodological differences. The most important of these are that ICM weights people by their likelihood to vote while YouGov does not, and YouGov ignores people who say they don't know how they will vote while ICM reallocates a proportion of them based on how their voted at the 2010 election. There may also be less quantifiable differences in their weighting and sampling, but the upshot is that ICM tend to show some of the worst scores for Labour of all the polling companies, while YouGov tend to show some of their better scores (and the opposite with the Lib Dems - ICM invariably give the Lib Dems their best scores, YouGov tend to give them their worst).

I'll do a full report on rest of the questions in the YouGov/Sunday Times poll tomorrow.