Sunday, 21 December 2014

Testing the boundaries (2): Labour vs the SNP

One of the huge unknowns for 2015 is how the post-referendum fall-out will translate into votes in Scotland. The betting markets have been chaotic. Chaos is potentially very lucrative - or very expensive. But the time has come to take a position on this. 

I've looked at the Scottish seats several times over the last few months.  The battleground is as follows. The swings required for substantial SNP success are heroic. A 10% swing to Labour across the whole of the UK would give them a landslide majority. A 10% swing for the SNP would give it five seats.

But the SNP are currently polling well ahead of that level.  The most recent YouGov Scotland poll registers a 21% swing from Labour to the SNP.  So the question is: can they maintain this or will the balloon deflate?

So where are we now?

Ground game

Historically, Scottish Labour has been dominant in its ground game.  No longer.  The SNP has at a conservative estimate seven times as many members as Scottish Labour and possibly more than ten times as many members - Scottish Labour is cagey about its membership, to the point of not releasing absolute numbers of votes in the recent leadership contest:

Many of these new SNP members will be raw recruits with no political experience and some will no doubt be uninterested in the tedium of treading the pavements rather than waving saltires, but even making full allowance for this, the SNP will have far more people power than Scottish Labour.  So far, the enthusiasm seems undimmed, as shown by this recent account from a Labour supporter:

" Yesterday the SNP had 12,000 people pay to attend a rally, and a day later in the same city, Labour could barely scrape triple figures, for free, for a leadership contest. That’s 1/120th of the size of the SNP gathering, or 1/40th of the size of the Radical Independence Coalition event held the same weekend."

This in-built advantage for the SNP is not going to change over the next five months.

Meanwhile, the SNP dominated the Yes campaign and many of its members will have intimate knowledge of who voted for independence.  The No campaign was more pluralistic, which means that Scottish Labour will not have quite such good access to the data from the referendum. 

Party leaders

Both the SNP and Labour have new leaders in Scotland, and both have to endure the shadow of others in their party looming over them.  Alex Salmond stood down as leader in the wake of the referendum result, but even before his resignation took effect, he seemed to repent of his decision and is now going to run for a Westminster seat.  Nicola Sturgeon seems to have a good working relationship with Alex Salmond, but sooner or later there will come a question on which she disagrees with him.  Right now many are still behaving as if Alex Salmond is still leader.  Is Nicola Sturgeon going to look like she has all the authority that President Medvedev had?  It's hard enough following in the footsteps of a hugely popular party leader, especially if you are personally close to him, without having to deal with him remaining ever-present.

This is unlikely to be a problem before the May election, but it is quite possibly going to be important not that long after.  Who will be in charge of tactics and strategy for the SNP in a hung Parliament?  This relationship is one to watch.

Scottish Labour have avoided the worst outcomes from their leadership contest.  In Jim Murphy they have by far the most competent of the three leadership candidates available and he won by a sufficiently wide margin to silence his party opponents.  Neither of these outcomes was preordained.  A complete Scottish Labour meltdown looks markedly less likely as a result.

But it's hardly as though Scottish Labour are marching into sunlit uplands.  The Scots are notably contemptuous of Ed Miliband: he has worse ratings in Scotland for doing his job than David Cameron or even Nigel Farage:

Ominously, only a third of Scottish Labour's current support thought that Ed Miliband was doing the best job.  And of the Holyrood 2011 Labour support, itself the cadre of Labour supporters from an appalling result for Labour, more people thought that Nicola Sturgeon was doing the best job than Ed Miliband. 

Jim Murphy is going to have to inspire the Scottish electorate without any confidence that he will be helped by his grand supremo and with the distinct possibility that he might find himself sabotaged from above at some point.

The shape of Scottish politics

At the last two Holyrood elections, the SNP has come out on top - narrowly in 2007 and by a landslide in 2011.  But in the last two Westminster elections, Labour has hoovered up almost all the Scottish seats, with the SNP trailing far behind in their wake.  Evidently Scottish voters understand the different systems and different purposes of the two different sets of elections.  If 2015 ran true to form, we could expect to see Labour do much better than current polls suggest.

There are, however, strong reasons to believe that 2015 will not follow the same pattern. Scottish politics at the end of 2014 remains dominated by the referendum vote.  Scotland's place in the union is the chief subject for discussion.  While the subject remains that of Scottish identity politics, Labour is in trouble.  In that YouGov poll, 39% of 2010 Labour voters are now recorded as supporting the SNP (though that's not as bad as for the Lib Dems, where 49% of 2010 Lib Dem voters are recorded as supporting the SNP).

This is very much to the SNP's liking, but by electing Jim Murphy as leader of Scottish Labour, Labour has chosen a leader who was one of the main faces of unionism during the referendum campaign. Merely by being who he is, it is going to be hard for Jim Murphy to move the topic of conversation.

There was talk in the wake of the referendum of the independence campaigners standing for Westminster under a unified Yes alliance.  That idea was formally adopted by the SNP, but seems to be being smothered under the guise of co-opting the Yes movement for the SNP, with the aim of getting some prominent independents to stand under this banner also.  This would exclude the Greens and the SSP.  Since these two parties together tally something like a 5% vote share in Scotland, that makes the SNP's task of taking seats that bit harder.  But that's the decision that Nicola Sturgeon seems to have taken.

The dangers of working off uniform national swing

Scottish politics has been upended since 2010 by three large polling movements.  The Lib Dems have dropped from 19% to roughly 5% in the polls.  Labour have dropped from 42% to 25% or so in the polls.  And the SNP have risen from 20% to 45% or more in the polls.  These are enormous movements.

Uniform national swing is designed for small swings.  In the case of the Lib Dems' drop, it simply breaks down mathematically, as I explained here:


The same problem arises to a lesser extent with Labour.

Uniform national swing can be made to work for large rises in the polls (though that would disregard the fact that a fair bit of the SNP's rise has come from the Lib Dems).  But there is an obvious objection to using it here, because we already know that the SNP's rise has been driven by something that we know was not geographically uniform: enthusiasm for Scottish independence.

As I have previously noted, there is something like an inverse relationship between past SNP Westminster success and support for Scottish independence in the referendum:

Uniform national swing is most unlikely to help us judge what seats are going to fall in practice.

I am not the only one to have noticed this disparity.  Stephen Bush of the Telegraph has put together a model for trying to work out which seats will fall where:

While the model is not fully explained and the workings are not fully clear, I like the thrust of what he is trying to do.  I regard its stated assumptions as pretty friendly to Labour, but even so it would see 17 Labour seats fall to the SNP with a further seven in serious jeopardy:

"All in all, the following seats would fall to the SNP: Aberdeen North, Aberdeen South, Dundee West, Dunfermline & West Fife, Edinburgh East, Edinburgh North & Leith, Edinburgh South, Falkirk, Glasgow Central, Glasgow North, Inverclyde, Kilmarnock and Loudoun, Livingston, Linlithgow & East Falkirk, Midlothian, North Ayrshire & Arran and Ochil & South Perthshire.

The following seats would be held by fewer than 1,000 votes: Airdrie & Shotts, Ayr Carrick & Cumnock, Cumbernauld, East Kilbride, Glasgow East, Lanark & Hamilton East, and Stirling."

How will the rest of the run-up to the election pan out?

It's time to review where we are.  We now know how both Labour and the SNP are going to be led into the general election in May.  Political discourse is dominated by the question of Scotland's place in the UK and there is no sign that is going to change.  Jim Murphy has a Herculean job ahead of him to change that, and while he's capable, there's only so much that one man can do.  We can expect him to steady the ship, but he is not well-placed to appeal to the voters who have left Labour for the SNP.

Ed Miliband does not seem to appeal at all to Scots, so I expect the SNP to have a substantial lead over Labour in Scotland at the next election unless the SNP mess things up.  A self-inflicted SNP defeat, however, is not a trivial possibility, given the hubris floating around the nationalist movement: some of their post-referendum defeat rallies have come perilously close to The Triumph Of The Won't.

In the wake of the referendum, I expect turn-out to be higher than at previous general elections, as Scots have rediscovered their taste for democracy.  Unionists are quieter than nationalists, but they trooped out in very high numbers to defeat the referendum.  Equally, the nationalists did well to get many previous non-voters to vote Yes in the referendum and with knowledge of who these people are and with the foot soldiers available for a big ground game, I expect them to get a lot of them to vote again next May.

The betting markets

Let's have a look at the Labour seats that the SNP are competing for as they stand now:

These are all the Scottish Labour seats ranked in order of the current prices on the SNP taking them.

Ladbrokes price the SNP as favourites to get most seats in Scotland at 8/15, and Labour at 11/8.  However, when you look at the individual constituency markets, the SNP are favourites to win only 16 of the 59 Scottish seats (and only three Labour-held seats), and  they will probably need at least 25 to take most seats.  Both Paddy Power and Ladbrokes have set an over/under market on SNP seats at 24.5, priced at 5/6 each side of the line.  There must be value here somewhere, so where is it?

Should I be betting now at all?

Much depends on how confident I am in my judgements as to what is going on.  In truth, not very, but the jelly does seem to be setting.

We are sure to get some Scottish constituency polls at some point from Lord Ashcroft.  At that point, all the value will go out of those constituencies in minutes.  And those polls will be snapshots not predictions, so they won't actually tell us how susceptible the Scottish public will be to the new Scottish Labour leader by the time of the general election.

Thanks to constituency bets placed before the referendum and since, I'm sitting on a nice notional profit already.  I could just sit on that, or if I was cautious, I could close it out.

But I'm greedy.  I think the constituency markets still substantially underrate the SNP's chances.  The world has changed after the referendum and I see no reason to think that it is going to change back any time soon.  The 8/15 on the SNP getting most seats looks about right to me.  So that means that the constituency markets have continuing value.  I'm going to increase my bets on the SNP.

How am I going to choose constituencies?

Given the conclusions I have reached, the first thing to do is not to take the 2010 results too seriously.  Roll up that electoral map of Scotland; it will not be wanted these ten years.  If I am right, we have new electoral coalitions constructed and they will not be quickly broken down.  So those apparently rock-crushing majorities for Labour are looking very vulnerable in reality.

How to locate the SNP's best prospects?  The Stephen Bush article is worth a lot of attention.  You could do a lot worse than putting stakes on a selection of those 17 seats that he has identified.  They're almost all odds against at present.  In my view, the SNP should be odds on in all of them.

One thing to look carefully at is the past share of the vote held by the Lib Dems.  In all post-referendum polls, at least a third of the Lib Dem vote has been recorded as going to the SNP and in the most recent YouGov poll just under half has made that journey.  That has barely been noted, but a third of the 2010 Lib Dem vote is over 6% of the whole electorate and half is just shy of 10% of the whole electorate.  This is a big chunk of the SNP's new coalition.  So seats with big Lib Dem presences in 2010 are well worth additional consideration.

I've chosen the following seats for an additional punt:
  • Glasgow North, which is both in Glasgow (where Yes was victorious) and has a substantial 2010 Lib Dem vote.  
  • Glasgow Central - it doesn't have the Lib Dem vote but it is in Glasgow.
  • Linlithgow & Falkirk East and Kilmarnock & Loudoun, in each of which the SNP already had a quarter of the vote in 2010, giving them a substantial base to build on
  • Aberdeen South.  Aberdeen was an area which broke 60:40 for No, so it's not immediately the SNP's most promising area.  But Aberdeen South had a reasonable 2010 vote for both the Lib Dems and the Conservatives, and the SNP can hope to take a large part of the former and that the latter are unlikely to be willing to vote tactically for Labour.  Something close to 40% of the vote might well be enough for the SNP to take the seat.
I've put most on Glasgow North and least on Aberdeen South.  But there is quite a bit of guesswork in this.

What if I'm wrong about Labour's chances?  Are there any bargains anywhere on this table?  In short, I don't think so.  The bet to make if you think the SNP are going to fizzle is the 11/8 on Labour most Scottish seats.  But that's an awful lot of faith to be putting in the abilities of Jim Murphy.  


roger said...

Very good article. I'm going to put some money on Edinburgh South staying with Labour which did well for me last time. I'd be cautious about Aberdeen going SNP.There could well be something of a backlash

Anonymous said...

Factor in the huge drop in oil price before placing any more money on further SNP gains in Aberdeen.