YouGov/Sunday Times and more on that Scottish poll
Polling attention today will unavoidably be concentrated on the Scottish YouGov poll, but for the record the regular YouGov/Sunday Times GB poll is also out - results here. Topline figures are CON 33%, LAB 35%, LDEM 7%, UKIP 15%. That too had a large chunk of Scottish questions aimed at English and Welsh respondents.
English and Welsh people now oppose Scottish independence by 61% to 17%. In the event Scotland does vote YES, only 22% of people in England & Wales think David Cameron should resign, 54% think he should not. On one hand that makes Cameron's position look safe, but my suspicion is that public opinion could be radically different if it actually happened - people are not always very good at predicting their response to hypothetical situations. YouGov also asked people what they think should happen to Scottish MPs during that interim period between voting YES and actually becoming independent - 56% of people in England & Wales think Scottish MPs should not be allowed to attend and vote in the Commons during that period, 62% think Scotland should not elect MPs in the 2015 election if they are becoming independent (the same questions were asked in the Scottish poll with, as you might imagine, somewhat different results - 55% think that Scottish MPs should continue to vote in the interim period, though the 2015 question is quite close - 47% think Scotland should return MPs, 41% think it shouldn't).
Turning back to the headline Scottish poll showing YES ahead, the full tables are now up here. Peter Kellner also has a commentary here, which amongst other things highlights how the biggest shift in opinion over the last month has been amongst under 40s and Labour voters.
Looking at how some of the other tracker questions have changed, the Yes Scotland campaign continues to be seen as more positive than Better Together, but now it is also seen as more honest (back in June people thought YES was more positive than NO, but thought NO was being more honest. Now YES leads on both measures). There's also been a narrowing in the economic questions - back in June 49% thought an independent Scotland would be worse off, 27% better off - now it's finely balanced, 40% think Scotland would be better off independent, 42% worse off.
It remains quite strange that YouGov have shown this sharp narrowing in the race while Panelbase haven't. The difference is not just an outlying poll - while one single poll could be a freak result, YouGov has shown a consistent narrowing in the face over three polls. It's not down to any methodological change - this poll was conducted using exactly the same methodology as YouGov's previous poll. The only recent change in methods was four polls ago, introducing weighting by place of birth, and both YouGov and Panelbase introduced that at about the same time. Things like a differential willingness to respond to polls (people who support a campaign on the up being more likely to click on the email) should affect both Panelbase and YouGov the same - they are both online companies using a panel based system. One possibility is simply that the different trends are down to the same reasons behind the previous differences between Panelbase and YouGov. Both use weighting by Holyrood recalled vote as a core weighting variable, but YouGov also break out a proportion of people who voted SNP in 2011 but Labour in 2010. Perhaps if those people - people's whose loyalty at Holywood and Westminster is divided between the SNP and Labour - had previously been more NO, but have moved towards YES this month, and are more represented in YouGov's sample? It's a possible contributory factor, but such things are rarely so neatly explained. We shall have to wait and see what sort of trend TNS and Survation show in the week.
