This week’s YouGov Sunday Times poll results are here, and have topline Voting intentions of CON 33%, LAB 38%, LDEM 9%, UKIP 12%. As you’d expect, the poll has questions on the Liberal Democrats for their conference, as well as some about Ed Miliband and the Unions.

By 45% to 34% people think it was the wrong decision for the Lib Dems to go into coalition with the Conservatives, though this is mainly driven by Labour voters. Amongst the remaining Liberal Democrat supporters three-quarters still think it was the right thing to do, but of course that’s largely because those opposed to the coalition are no longer voting Liberal Democrat! On the principle of coalition, 22% of people think it is better to have coalition governments that force parties to compromise, 53% think that a single party government is better.

22% think that the Liberal Democrats have been a positive influence on government, 25% a negative influence and 43% don’t think they’ve had much influence either way. Asked more specifically about Clegg’s claim that the Lib Dems have prevented the Conservatives from being more right wing, 46% think this is true (30% think it’s a good thing, 16% a bad thing). Asked which best reflects their view, 36% think that by entering coalition Clegg was doing what he thought was best for the country, 44% that he was betraying his principles for power regardless of the interests of the country.

Turning to Labour and the Unions, on balance people still think that Ed Miliband is too close to the Unions (32% think he is too close, 17% too distant, 20% about right, 31% don’t know). Miliband’s proposals to change how trade unions affiliate members to Labour are widely supported and by 43% to 14% people think he is right to try and reduce Labour’s links with the Unions (although a further 20% think that he isn’t actually trying to do this). Despite this overall he is not seen as handling his party’s relationship with the Unions well – only 25% think he’s done it well, 46% badly.

Only 16% of people think Miliband is ever likely to be Prime Minister, 70% think it is unlikely. Even amongst Labour supporters only 42% think he is likely to be Prime Minister, 45% unlikely. This is a strange finding given Labour’s consistent lead in the polls and when polls ask people which party they expect to win the election, far more tend to say Labour. I suspect this is speaking more of people’s difficulty in visualising Ed Miliband as Prime Minister, rather than their considered prediction.

Also buried away in the poll was a repeat of the “bedroom tax” question from back in March. Back then YouGov found more people supported the policy than opposed it, since then opinion has switched round, and there are more people opposed (48%) than in support (40%).

As I mentioned yesterday, there was also a second Scottish poll in the Sunday papers, Panelbase in the Sunday Times. They had topline figures of YES 37%, NO 47%, practically unchanged from their last Sunday Times poll which had topline figures of YES 37%, NO 46% (it also demonstrates pretty conclusively that the answers to referendum voting intention in the Panelbase poll for the SNP were influenced by the two preceeding questions). Panelbase have done the most regular Scottish polling over the last year and a bit, and leaving aside that SNP poll have shown very consistent figures, with YES support between 34%-37%, NO support between 44%-47% and no obvious trend in either direction. I’ve updated the page on Scottish referendum polls so far here.

Today’s YouGov voting intention figures for the Sun are CON 33%, LAB 37%, LDEM 9%, UKIP 13%. The four point lead is towards the lower end of YouGov’s recent averages, and suggests the government’s defeat over Syria had no lasting effect upon party support. Perhaps the 10 point Labour lead at the weekend was a freak co-incidence, a normal outlier that just happened to appear straight after a government defrat, or perhaps it was a genuine but short lived effect in a poll conducted when the media was plastered with news of the government’s troubles. We’ll never know. Tabs are here.

There were also some extra Syria questions for the Sun which found opposition to military action hardening up in the aftermath of the vote. 69% of people are now opposed to the idea of a British missile attack on Syria (up from 50% just before the vote), 73% of people now think Parliament was right to block action (up from 68% straight after the vote). Tabs here.

Also out this morning was another Scottish referendum poll, this time from TNS BMRB (the successors via various mergers to System Three, a far better known brand in Scottish polling!). TNS show referendum voting intentions of YES 25%(-8), NO 47%(-5), Don’t knows 28%(+13). Changes are since TNS’s last referendum poll in February, though this one is apparently the start of a more regular series of polls. John Curtice already has an analysis up here – note the caveat about the sample. TNS BMRB don’t use any political weighting (I assume for similar reasons to MORI’s worries about levels of false recall), but it can result in samples like this one, where more people claimed to have voted Labour in the 2011 Holyrood election than claimed to have voted SNP.


Looking again at Populus’s AV polling earlier this month I spotted something interesting. As you’ll recall, Populus asked two questions – one a bare question which found YES on 33% and NO on 37%. The second briefly describing AV in the question which gives NO a higher lead, as we’ve seen in other polling. Here Populus found 46% saying they’d vote NO, 29% YES.

However, until I looked at the tables I hadn’t spotted something more interesting about Populus’s second question: they gave people a four point answer scale, asking people both how they’d vote, and how certain they were, and the NO vote appears to be much firmer.

Amongst the 46% who said they’d vote NO, 31% said they would definitely vote NO, 15% would likely vote No but haven’t completely made up their mind. Amongst the 29% who said they’d vote YES, only 11% said they’d definitely vote YES, 18% that they hadn’t completely made up their mind.

This could well also be a pointer towards which side will be more motivated to turn out, though my own expectation is that outside London and the other areas with no other elections, turnout will be more driven by the people who are turning out to vote in the other elections that day.

In my round up of AV polling earlier this week I said that while different companies were showing different topline figures, all the companies that had polled in the last month or so were united in showing the tide moving in the direction of the NO camp.

ComRes’s monthly poll for the Independent on Sunday is out now, and follows the same trend. The YES campaign are on 34% (down 6 since February), the NO campaign are on 37% (up 7 since February). Don’t know is at 28%. Notably while YouGov’s polls that preface their question with an explanation of FPTP and AV have been showing a NO lead for a long time, this is the first poll asking the bare referendum question to have shown NO in the lead.

Topline voting intention meanwhile is CON 37%(+1), LAB 40%(-2), LDEM 11%(nc), Others 12%. Changes are from the previous online ComRes poll a month ago, since on average their online polls seem to show higher support for the Conservatives than their phone polls.

UPDATE: There is also a YouGov poll on AV for Sky News out tonight – voting intention (using the bare referendum question, without the explanation of the systems used in the AV tracker for the Sun) is YES 37%, NO 32%, Don’t know 24%, Would not vote 7%.

Different companies and different questions are showing a wide variety of different results on the AV referendum, but they all show the tide moving in the NO campaign’s direction.

At one end YouGov’s latest poll, prompted with explanations of what the AV and FPTP systems are, has a 17 point lead for the NO campaign. At the other extreme Angus Reid have a 6 point lead for the YES campaign, and ComRes’s last poll – now almost a month ago, had a 10 point lead for YES.

Most of this difference is down to question wording – in the last month YouGov and Populus have both done parallel surveys, one explaining the systems to respondents and one not. In two parallel surveys for the No2AV campaign YouGov found the YES campaign ahead by 3 points when people were asked just the raw question, but when they were given the explanations of the systems NO led by 11 points.

In a similar exercise, Populus found a 12 point lead for YES when they asked just the raw referendum question, but found a 14 point lead for NO when they told people what the two systems were.

It may be that these differences gradually vanish as we head towards the referendum itself and the public become more aware of what AV and FPTP actually are… or it may be that the referendum campaign produces more smoke than light, and the public don’t actually end up better informed at all!

While polls are showing different overall figures though, they are showing the same broad trend.

Angus Reid have been polling roughtly monthly since January. In January they had YES ahead of NO by 37% to 20%, with 37% don’t knows. Their most figures are YES 32% (down 5 since January), NO 26% (up 6), and 35% don’t know.

ICM asked AV voting intention for the Guardian in both December and February. In December they found YES on 44%, NO on 38%, Don’t know on 18%. By February their figures had moved to YES 37% (down 7), NO 37% (down 1), don’t knows 27%.

YouGov ask AV most frequently, with data every fortnight, so we have more granularity there on the ups and downs of the campaigns. Since last summer YouGov had been picking up a gradual trend towards NO, at the start of 2011 YouGov had a NO lead of around 9 points. However, at the start of February YouGov picked up a shift towards YES, with the AV campaign briefly narrowing the gap to just one point (it looked like a rogue result so we ran it two days in a row to check – it wasn’t!).

It was to be purely temporary though, since then the campaign proper has started and YouGov have been recording a strong trend towards NO. A fortnight ago they recorded a 7 point lead for NO, then a week ago an 11 point lead for NO, today a 17 point lead for NO. The last one may turn out to be an outlier, but the trend is undeniable.

Finally ComRes have been asking about AV monthly for the Independent on Sunday. In January they had YES six points ahead (36% YES, 30% NO, 34% d/k), in February they had YES increasing their lead to ten points, at the same time as YouGov were picking up that sharp but temporary move towards the YES campaign. The next ComRes poll is due this weekend in the Indy on Sunday, so if they follow the same trend as Angus Reid, ICM and YouGov we’d expect it to show a drop in the YES campaign’s lead.