Sunday round-up
Results of this week's YouGov poll for the Sunday Times are here, with questions on the usual grab bag of subjects - most notably on tax avoidance and energy. The topline voting intention figures are CON 32%, LAB 43%, LDEM 9%, UKIP 9%, so despite several YouGov polls in a row showing single-figure Labour leads things still appear to be averaging up around the 9-10 point lead that has been the norm for six months or so now. Leader approval ratings are Cameron minus 19 (from minus 20 a week ago), Miliband minus 18 (from minus 14), Clegg minus 56 (from minus 58) - Ed Miliband's conference boost is still evident, but continues to gently unwind.
On tax avoidance, 64% of people think it is unacceptable to avoid tax compared to 26% who think it acceptable... however, 42% of people say that they personally would avoid taxation if they had an accountant to show them how. 71% of people think that high tax rates on business and the rich encourage tax avoidance, however only a minority (19%) think that this means tax rates should be reduced. The majority (52%) think that government should crack down on avoidance rather than cut taxes.
Turning to energy, the energy companies themselves are by far the most widely blamed for increasing energy prices - 58% think they are most to blame, compared to only 17% saying rising gas and oil prices and 11% the cost of carbon emission targets. Asked about shale gas and fracking people were evenly split - 32% think it should go ahead, 30% that it shouldn't. 38% said don't know, probably indicating it is an issue that many people have very little awareness of.
In the Mail on Sunday there was also a Survation poll, which had topline figures of CON 30%(+1), LAB 43%(+2), LD 8%(-2), UKIP 12% (nc). Changes are from the previously published Survation poll on the 23rd September. The Mail on Sunday's write-up appears to be a prime example of how not to report opinion polls. It begins with a subheading of "Conservatives 13 points behind Labour, one point ahead of UKIP Party" which is clearly untrue, though probably an innocent error. The rest of the article though is worse - Conservative support hasn't "fallen to 30", it has increased to 30. They haven't "dropped five points in ten days", they have increased one point.
In the absence of any other polls from about ten days ago showing the Conservatives at 35 I can only assume that the 5 point drop comes from comparing the poll to the YouGov poll conducted on the 11th of October. This is doubly wrong - first it is deliberately cherry picking an unusually high score as a point of comparison to exaggerate the movement. Secondly (and assuming there was not some unpublished Survation poll they are comparing it too), they are comparing polls from different companies using different methodologies that produce consistently different results. Since April YouGov's polls have shown an average Conservative support of 33% (the 35% was either a blip or a party conference publicity boost), since April Survation's polls have shown an average Conservative support of 30%. In other words, the poll does not show Conservative support "plummetting", it shows Conservative support at exactly the same level that the pollster in question has been showing them at for months and months.
