ICM has CON 5% ahead while Populus gives LAB a 1% lead

It’s been a huge day of polling on both sides of the Atlantic and I’m only now catching up with the first UK telephone polls of 2012.

ICM had with changes on December – CON 40%(+3): LAB 35%(-1): LD 16%(+1) . So a CON-LAB gap exactly the same as YouGov yesterday but with very different shares for the LDs.

Populus for the Times has with changes on last month – CON 37%(+2): LAB 38%(-1): LD 13%(+1)

Although the top line lead numbers are very different the direction of travel for each of the three main parties is the same. CON & the LDs up with LAB down.

Given that Populus uses the same broad methodology as ICM and, indeed, usually uses ICM calling centres for its field-work, you would expect the two to come out with broadly similar numbers. My guess is that the differences are largely methodological with the ICM approach to turnout being the key divider.

Unlike Populus ICM discounts by 50% the views of those in its surveys who did not vote at the 2010 general election. We’ve seen some evidence from other polling that LAB has more 2010 non-voters amongst its current supporters and their views are adjusted accordingly. Whether this is the right approach or not I do not know but ICM did get the AV referendum outcome exactly right.

Coming up later is the first phone poll using live interviewers on the Florida primary where all the betting action is at the moment. The two polls we’ve seen so far use the automated rob-call system where respondees keying their answers on their phone key-pads.

Mike Smithson @MikeSmithsonOGH

Recent Threads