Less of a snapshot, more of an identifit
Frustrating, isn’t it? We’re now just over four months away from the Holyrood elections, and yet the understanding we have of public opinion in the lead-up to polling day is at our most limited in sometime. Even Scottish Westminster polls appear to have evaporated. Is next May to be the Election That Pollsters Forgot?
Well, there’s time yet, but for now, all we have are sub-samples of UK Westminster polls to work from. These aren’t great, but let’s see what we can pick up from YouGov, ComRes and Angus Reid.
The three polls put Labour between 37% (ComRes) and 47% (YouGov), with Angus Reid putting Labour on 42%.
The SNP are on 30% according to ComRes, 34% according to Angus Reid, yet 19% according to YouGov.
The Tories are on 11% according to Angus Reid, 19% according to ComRes and 24% with YouGov.
But there’s hardly any deviation in the LibDem polling: just 6% with Angus Reid, 7% with YouGov, and 8% with ComRes.
Let’s look at a slightly bigger sub-sample: the outcome of the Westminster Election. Labour took 42% of the vote in Scotland, the SNP just under 20%, the Tories took 16.7% and the LibDems just under 19%.
Now, bear in mind that the UK polls have Labour and the Tories neck-and-neck. At the election, the Tories took 36.1% of the UK vote to Labour’s 29%, with the LibDems on 23%. Seeing as the Liberal Democrats are now lucky to get half of that in the opinion polls, it looks like Labour’s current good fortune has more to do with the LibDem collapse than anything else.
So there are a couple of ideas we can pick out from the sub-samples.
1. While Labour seem to be benefiting from the LibDem collapse in England, and a similar collapse is underway in Scotland, those benefits are not coming to Scottish Labour. Even the YouGov poll sees only a 5-point gain to the LibDems’ 12-point loss, and the ComRes poll actually has Labour going backwards – something that would be unthinkable at the UK level right now!
2. With the exception of the YouGov poll sub-sample, the SNP are in a decent position to capitalise on the LibDems’ misfortune. It’s no surprise: as the LibDems progressed in the 2001 and 2005 Elections, so the SNP went into reverse. When the SNP made advances in 2007, the LibDems unexpectedly went backwards and as the SNP increased its Westminster vote share in 2010, Scotland was the only part of the UK where the LibDem vote share declined. The two parties appear to occupy a broadly similar space so when Charles Kennedy was LibDem Leader and successfully articulated policies that resonated with that large crossover group, that spelled trouble for the SNP. When the SNP had a more attractive message (or messenger), that hit the LibDems. Now that the LibDems are positively unpopular, it stands to reason that it’s good news for the SNP.
3. While 11% looks far too low, and 24% looks ridiculously high, the Tories look to be on a slow, yet upward trajectory. Clearly, memories of Thatcher have loomed large yet David Cameron’s accession to the Premiership has not seen the sky fall in, so maybe, just maybe, people – at least, those who would have switched to the Tories at some point over the last fie years had they been from or living in England – are starting, little by little, to defrost towards the Conservatives. The question is, when the decisions being taken actually start to take effect, can that continue? I’m not sure.
4. The LibDems are buggered. Fine, these are only sub-samples, but they’re speaking with one voice and it’s not good for them. Besides, even the full poll paints a picture of the party having lost a half to two thirds of its support in May. Now, it could be that people are no longer willing to admit to voting LibDem (remember the ‘Shy Tories’?), but if that’s the case, and if Liberal Democrats are the voters who dare not speak their name, that’s a massive credibility problem.
But that’s all well and good – for Westminster. It doesn’t tell us anything about Holyrood. Does it?
Actually, it might. There was, once upon a time, a Golden Age of Scottish Polling, when the Herald paid SystemThree to carry out a monthly opinion poll which produced figures for Westminster and both Holyrood votes. Alas, the last of these was carried out in December 2003, but just as they gave us a snapshot of public opinion then, they also gave us something wider: they told us how more or less likely people were to vote for a particular party, depending on what ballot paper was in front of them.
And this is what we found: for every one percentage point that the SNP picked up in the Westminster section of the poll, they picked up 1.27 points in the Holyrood Constituency vote, and 1.25 points in the Regional. The LibDems have a Westminster:Constituency:Regional ratio of 1:0.93:1.16 (interesting that their polls had them doing better in the Regional vote than the Constituency vote when the ballot box suggested otherwise, but still); the Tories W:C:R ratio was 1:0.82:0.76 and Labour’s was 1:0.85:0.71.
So if there were a Holyrood poll out right now, what I think it’d suggest is that it’s neck-and-neck between Labour and the SNP on the Constituency vote, with Labour just shading it. The SNP, however, would have a five-point lead on the Regional vote.
And therein lies the rub: that slight advantage on the Constituency Vote would see a slew of marginals head Labour’s way, and I estimate that if I had an actual poll to run a projection on (this is what the lack of polls has reduced me to), Labour would have 50 seats.
At this point, I’m going to lob in a hand grenade. There’s one party that’s conspicuous in its absence from all of these calculations up to now – one party I don’t even have enough data to make an algorithm for: the Greens. I’ve had to count them as part of the ‘Others’ column for the polling, but for the extrapolation, they do need to be handled quite distinctly. So, I ran three versions: one which gives the Greens half of the ‘Others’ total – generous, but as the only ‘Other’ party to remain in Parliament after 2007, and having avoided tearing itself apart or being involved in either defamation or perjury proceedings, that might be fair. The second assigns them one third of the total – broadly in line with the ‘Others’ column in the polling data we do have. In the third, the most miserly outcome for the party, I assigned them the same percentage of the ‘Others’ vote that they picked up in 2007 – around 28% of all Regional votes not cast for the Big 4 were cast for the Greens.
Now this matters. The unknown strength (or otherwise) of the Greens determines the D’Hondt calculations, which won’t work if we just stick an x in. So with three different scenarios, we get three different outcomes.
Given Labour’s constituency-based strength, they get 50 in all three scenarios.
In the 28% scenario, the SNP get 49 seats (2007 in reverse). In the 33% scenario, they get 48. In the 50% scenario, they get 45. So the better the Greens do, the more problems they cause for the SNP.
The Tories get 18 seats – except in the 50% scenario, where they only get 17. The LibDems, meanwhile crash to ten seats… only nine in the 50% scenario.
And as for the Greens, the as yet unknown variable? They could get only one seat, they could stay where they are on two, they could end up with seven.
So you’d think that Labour would be laughing… but there’s another twist in the tale.
Obviously, having made it this far, the SNP have shown that minority government is do-able and there’s nothing stopping Labour from at least trying to follow suit if an outcome like this actually takes place. But legislation still needs to be passed and the minority government has had its hairy moments: the Edinburgh Tram scheme, the Creative Scotland Bill, the 2009 Budget.
Unless the Greens do really well, it’s the Tories who hold the balance of power. A Labour-LibDem Coalition would yield only 60 seats, five seats short of the magic number – 65. Labour plus the Tories gets to a minimum of 67.
The problem is that the Tories do not want to coalesce with Labour. Annabel Goldie has said that she might be willing to entertain Coalition after 2011, but her preference would be the SNP. After the 2009 Budget fiasco, Goldie proclaimed the Chamber that “Scotland is already badly served by one Labour Government, we don’t need two”. Labour may no longer be in Government at Westminster, but while the two parties will happily vote together to stick it to the SNP on occasion, it’s doubtful she’s changed her tune that much.
But the real question is the other parties’ attitude to the Tories. Labour are probably on the same page as them on most of the key issues, but on the ones where they differ, the gulf is too wide. Besides, Annabel Goldie would have to explain why the Tories want Labour out at Westminster but in at Holyrood, while Iain Gray would have to explain why the Tories are evil at Westminster but tolerable enough to do business with in the Scottish Parliament. Similarly, the SNP and Tories are not bedfellows on many issues, and the SNP still has a ban on a formal Coalition with the Tories. However, the two don’t want Labour in office all that much and of all the parties, the SNP has probably found the Tories the most amenable to deals on matters such as the Budget. A Coalition involving the Tories might not be on the horizon, but an SNP/Tory Confidence & Supply agreement? Maybe.
That said, if the Greens do well, it’s all moot. They hold the balance of power, with perhaps seven seats that could move at will between a Labour/LD alignment and an SNP/Tory arrangement. The prospects of the Greens getting on board with the latter may seem slight, but it’s a voting combination that isn’t unheard of in the Scottish Parliament, and even Green heroine Caroline Lucas is open to working with Tory MPs when she thinks it might be advantageous. I’m assuming, of course, that the LibDems still won’t want anything to do with the SNP and judging by the noises made by Mike Rumbles and his ilk (referring to the SNP being out next May – the voters will be the judge of that, Mikey Boy), I think that’s still a reasonable assumption to make.
Indeed, the Greens seemed happily aligned with the SNP at first, but relations aren’t what they were and on the issues that matter to that relationship, the Greens are particularly exasperated. My guess is that the grass isn’t any greener on the other side of the fence though, and even if the Greens did come to at least a loose short-term arrangement involving Labour and the LibDems – not dissimilar to the one reached between the Greens and the SNP in 2007, which I suspect is as much as the Greens will go for – it wouldn’t last long.
We have limited information: I’ve taken averages of three different sub-samples of three polls for a different institution, subjected the figures to an algorithm that may or may not be right, plucked figures out of thin air for the Greens, then run them all in a projection using a uniform national swing (which doesn’t happen) against notional figures for the new boundaries. In fact, there are probably more variables in play than there are cold, hard facts. But right now, it’s all we have.
And what we have suggests that the SNP will be ahead on votes, but Labour will be ahead on seats – which makes talk of legitimacy far more complicated to begin with, and it becomes a real race to see who can secure the necessary deals to get them to 65.
What we have suggests that a party which no one wants to work with will hold the balance of power.
Or that the balance of power will be held by a Green Party that will work with anyone who’ll deliver progress on their objectives – but turn their back quickly if expectations aren’t met.
Hmmm. This is going to make my prediction post at the weekend a bastard to write…