Saturday, March 10, 2012

What does the latest YouGov poll mean for the local elections?

I'm a few days late in tracking down the full figures from this week's Scottish YouGov poll -

Holyrood constituency vote

SNP 40% (-4)
Labour 36% (+4)
Conservatives 12% (-1)
Liberal Democrats 8% (+1)
Others 4% (-)

Holyrood regional list vote

SNP 38% (-1)
Labour 32% (+1)
Conservatives 13% (-)
Liberal Democrats 7% (-)
Others 11% (-)

So there's no change of any significance on the regional list ballot, but a drop in the SNP constituency lead from twelve points to four is the sort of thing that would be making us have kittens if we were in the run-up to a Holyrood election. As ever, there are a couple of health warnings here. Firstly, these figures are pretty similar to YouGov's eve-of-poll results from last year that proved to be so misleading. Secondly, we always have to be cautious about headline changes from just one poll, as they could easily be (wholly or in part) an artifact of the margin of error.

But if by any chance the drop in support is real, what could be going on? My guess is that the relentless media focus on the independence referendum since the start of the year is prompting some people to make more of a direct link between their own stance on independence and their party preference, ie. a few people who previously backed the SNP in spite of being anti-independence no longer feel they can do so. If that phenomenon is happening, in one sense it doesn't matter as long as we now view the independence referendum as the only vote that matters. But there's a snag - the local elections are an important way station on the route to that referendum, and they are now just two months away. Scarcely the best time for the SNP to be losing support, and unfortunately the figures for Westminster are somewhat starker...

Westminster voting intention

Labour 42% (+7)
SNP 30% (-7)
Conservatives 17% (-)
Liberal Democrats 7% (-)
Others (-)

Again, it's worth pointing out that these figures are better for the SNP than the ones YouGov were showing on the eve of last year's election, and of course a 30% vote share matches the party's all-time high in a UK general election. Nevertheless, a 7% swing to Labour over the space of a few weeks is clearly disappointing. So the million dollar question - does Holyrood or Westminster voting intention offer the strongest clue to the likely outcome of the local elections? Given the results in recent years, it might seem obvious that the answer is Holyrood, but of course the last three sets of local elections have all taken place on the same day as Scottish Parliament elections. For the last time local elections took place on a stand-alone basis, we have to go all the way back to the mid-1990s. In the 1994 regional council elections, the SNP received 27% of the vote, and in the inaugural unitary council elections the following year, the party received 26%. That was pretty healthy by the standards of the time, and it was certainly better than the SNP's performance in the 1992 or 1997 general elections. But it was also a little way short of the support received in the 1999 Holyrood election, even after that most bruising of campaigns. So the lesson may be that stand-alone local elections tend to produce results somewhere in between Holyrood and Westminster voting intentions. In which case, the YouGov figures might call into question the SNP's ambitions of seizing Glasgow in May, unless there's a much higher localised swing in the city (which of course is perfectly possible).

YouGov also asked how people would vote in the referendum itself. Unfortunately, the results simply aren't comparable with their previous independence poll, for two slightly mystifying reasons - a) don't knows weren't excluded this time, and b) the two possible answers were changed from 'Yes' and 'No' to 'Yes I agree' and 'No I disagree'. However, for what it's worth, here are the raw figures -

The SNP wishes to hold a referendum on Scottish independence in due course. Voters would be asked “Do you agree that Scotland should be an independent country?"

How would you vote if such a referendum were held tomorrow?

Yes I agree 32%
No I disagree 53%

Those results are very much in the 'normal range' for YouGov, which traditionally produces lower figures for the pro-independence side than other polling companies.


  1. If you take this poll to be accurate as one single poll isn't then we are polling higher than we were in 2007 and we will elect more councillors than we did in 2007. Also in Glasgow the SNP fielded far too few candidates in 2007 and let Labour have an over all majority by default. They have learnt how STV works and have adapted the multi-member vote stratagy courtesy of places like Dundee. It will be an interesting election.

  2. Hi James,

    Marcia's beaten me to it insofar as 2007 makes for a clearer baseline if you're looking for signposts for the local elections - the Holyrood poll shows both the SNP and Labour improving on their 2007 performance (though with an overall swing to the SNP). I've taken a stab at a possible projection from this poll and I would think that if the changes in support levels from 2007 are accurate, the SNP could find itself with 133 extra Councillors come May - 496 in total!

    Labour would end up with 43 extra (to 391), the Tories would lose 35 (to 108) and the LibDems would lose 109 (falling to 57). The Greens would lose half their Councillors (sounds bad, but it's only a fall from eight to four) and there'd be 28 fewer others (down to 166).

    One thing I'm worried about is the Glasgow fixation: the result in Glasgow last year was impressive, without a doubt, and it does seem like Glasgow Labour are trying to find new and innovative ways of losing control of the Council but we still don't know what impact Glasgow First will have on proceedings and if they don't hit the Labour vote hard, we risk what could be a major impressive night - nearly five hundred Councillors and possible outright control of Dundee, Angus, Aberdeenshire, Moray and Renfrewshire - could end up being branded a disappointment by the commentariat on the basis of the one that got away.

    Plenty of causes for optimism, but HQ needs to get that Glasgow is one Council out of 32, and remains one of the toughest nuts to crack, even now...

  3. I suspect that what all this bouncing around means is that, as Oldnat has said more than once on UK Polling Report, Yougov's sampling methodology is not very reliable in Scotland. We shall see.

  4. All this means is that YouGov's polling constituency is weighted in favour of unionism. That might sound like a pro-independence person making excuses, but it's the most obvious explanation for the fluctuations.

    As a result, it's difficult to take their results as a true reflection of the populace.

  5. Don't know about the poll, but I have been approached on 4 seperate occassions in the last month by, generally non political, people asking if this "independence thing" was really going to happen and how could it be stopped.

    Straws in the wind but would seem to indicate that, when people start to think about it, "independence" doesn't float their boat.

  6. Welcome back, Councillor Gallagher. I'd be extremely worried for your side of the argument if you hadn't met at least four real people who opposed independence by now.

    A fuller response to your comment can be found here.