Our latest voting poll puts Labour two points ahead of the Conservatives. Labour have 40 per cent of the Great Britain vote, up two from our previous poll. The Tories are on 38, down five.
The interviews were conducted from Thursday to Monday. As the fieldwork was frontloaded, most of it took place before the Corbyn wreath controversy became a major story (and all of it pre-dated the Westminster incident).
Elsewhere, the Liberal Democrats are unchanged on eight per cent, although the rounding obscures an increase of nearly a percentage point. This is not statistically significant on its own, but taken together with evidence from elsewhere, it appears that the party is deriving some benefit from the infighting within the two larger parties.
UKIP is up two points to five per cent, the combined Green Parties are down a point to two per cent and the SNP’s share of the Great Britain vote is unchanged at four per cent.
We also asked a number of questions on both topical and more fundamental issues, which will be published in the coming weeks and months, as will our new leader ratings numbers.
The methodology is largely the same from the one described back in April, but with a change, as it evolves from “experimental” to “production”.
Starting with this poll, we are introducing a crossbreak (and corresponding weighting) by ethnicity. Pollsters have traditionally shied away from publishing this breakdown because of concerns about how representative it is, and because of persistently undersampling ethnic minorities. This is understandable but, with non-white eligible voters now representing more than 10 per cent of the eligible population, far from ideal.
We found older BME respondents a bit harder to sample as expected, but having reviewed this and the previous sample, there is little sign of the usual problem of reaching the excessively politically engaged among them.
As such we are happy to publish the ethnicity crossbreak, with the usual caveats about crossbreaks. Since our raw sample was close to target for ethnic minorities, the effect of the weighting change is extremely small.
Number Cruncher interviewed 1,036 UK eligible voters online between 9th and 13th August. Data are weighted by age, gender, education, ethnicity and region to match the profile of the eligible voting population. Headline voting intention figures are additionally weighted by likelihood to vote, and are for Great Britain only. Image credit: Garry Knight.