Populus show Labour gaining from conference season
Populus's monthly poll for the Times is out. The topline figures, with changes from their last poll which was conducted prior to the conference season in mid-September, is CON 40%(-1), LAB 30%(+3), LDEM 18%(nc). The poll was conducted between the 9th and 11th.
This poll suggests that Labour have come out on top from the conference season. This makes it a sharp contrast from YouGov's polls prior to the weekend, and most notably to ICM, who use very similar methodology to Populus and who showed the Conservatives up to a lead of 19 points. This poll was conducted a few days after the flurry of polls at the end of last week, so it could be a result of the Conservative conference boost receding... or it could be an outlier. Having had some polls showing the Conservatives coming out the winners from conference season and now this showing Labour advancing, we really need to wait for some more polling to know for sure.
Other questions offer less good news for Labour - David Cameron now leads Gordon Brown as the best person to deal with the recession "right now" by 45% to 30%, and as the best person to lead the country after the next election by 48% to 28%.
Populus also split their sample by public and private sector. Unsurprisingly the Conservatives do better amongst those in the private sector - their support stands at 36% amongst public sector workers, 41% amongst private sector workers and 49% amongst the retired.
The private and public sector employment split raises another question too though, which some regular readers have emailed me to ask if I'd address. Regular readers will remember Ipsos MORI's methodology review after their wrong call in the London elections. MORI's conclusion was that their phone samples were including too many public sector workers, and since last Summer they have been weighting by public and private sector employment to control it. In MORI's weighted samples 12% work in the public sector and 44% work in the private or voluntary sectors.
Populus's sample had 57% of people in work (so much the same as MORI's samples), but the Times report says a third of them are public sector - implying 18% are public sector and 38% private or voluntary. Populus's sample would seem to have a lot more public sector workers than MORI's samples.
I raised this last time Populus gave a public/private sector split, which produced the same sort of contrast with MORI. The contrast could be a result of different ways of classifying people - I know Populus do it by providing a list of public sector employers (including Northern Rock) and asking if people work for any of them (an excellent approach to it!), but don't know how MORI approach it. Alternatively their sampling or weighting methods could genuinely be producing samples with different proportions of public and private sector workers. As Andrew Cooper of Populus said last time round, this will not necessarily make any difference to the voting intention figures since Populus use past vote weighting to make sure their figures are properly representative, but it's something worth keeping an eye upon.
