Scottish referendum
A referendum defeat has made the SNP more of a threat on polling day, not less
I’ve previously described Scotland as Labour’s red wedge, and a place where only one party is likely to make gains at the general election.
But the referendum has changed all that. With ostensibly rock-solid Labour areas voting in favour of independence, and the SNP’s membership soaring in the past few days, it’s clear the election result in Scotland might not be so one-sided. The excellent website electionforecast.co.uk currently suggests the SNP has an 85% chance of gaining one extra seat on polling day, and a 64% chance of gaining two. It’s timely to consider how and from whom the SNP could increase its current tally of six constituencies.
Here are five likely targets:
The first two in the list aren’t actually seats where the SNP is the main challenger. But the figures I’ve used are the number of votes the party came behind the winning candidate in 2010 – and you can see that numbers are smaller than in the other seats in the graph, where the SNP is second place.
Argyll & Bute is a fascinating seat: a four-way marginal, where in 2010 the Lib Dems came first ahead of the Tories, Labour and the SNP. It’s a brave person who’d predict the outcome of that particular contest in 2015, but the constituency is represented by the SNP in the Scottish parliament and voted 41.4% in favour of Yes at the referendum.
The SNP came third in Caithness, Sutherland & Easter Ross in 2010, behind the Lib Dems and Labour. If the Lib Dem vote collapses here, as polls suggest it might do across the UK, where would it go? Or more precisely, who would a collapse benefit more?
In Gordon, the SNP will be in a straight fight with the Lib Dems. This is the seat electionforecast.co.uk thinks the SNP has an 85% chance of winning. Within the constituency boundaries are some of the SNP’s traditionally strongest territories, including a central portion of Aberdeenshire and a northern chunk of Aberdeen. Both of these areas voted No in the referendum, however, while Aberdeen council is run by a frankly unlikely coalition of Labour, the Tories and some independents.
I suspect the result here will depend on the turnout of traditional Lib Dem voters, and the size of any collapse in the party’s support.
Falkirk and Glasgow Central will be head-to-head fights with Labour, and both may well be bad-tempered and bloody. The majorities seem on paper too large for the SNP to overturn, at least in one go. But Glasgow voted Yes in the referendum, and Falkirk wasn’t far behind (46.5%). Falkirk has been a running sore for Labour since the sitting MP Eric Joyce resigned from the party in 2012. In Glasgow Labour supporters deserted the party line in droves to support independence. It’s the uncertainty of the outcome in these contests that will arguably do more to help the SNP than any statistical or historical precedent.
In the arithmetic of the next parliament, a bigger block of SNP MPs seems almost perverse amid the present talk of accelerated devolution and a new constitutional settlement. It would be ironic indeed were the 2015 election to deliver to Westminster even more nationalist MPs from a part of the country to whom their rivals are willing to cede more power. But that’s one outcome that feels more likely now than this time last week – and another example of how losing the referendum doesn’t seem to be hurting the SNP that much at all.
I think http://electionforecast.co.uk/ are actually predicting that SNP are likely to win four extra seats.
The table at http://electionforecast.co.uk/tables/SNP_seat_gains.html is suggesting that the SNP has an 89% chance of winning Gordon – i.e. an almost certainty of winning that particular seat. But they are also odds-on (60%+ chance) to win Argyll & Bute, and Falkirk. That is, they are more likely than not to win these three extra seats.
Then, adding the 40% possibles (Caithness, Sutherland & Easter Ross, and Ross, Skye & Lochaber) and a few others that are less likely, that leaves the SNP with a central estimate of four extra seats (i.e. ten in total) and 90% confidence that they will get something between six and fifteen. Of course, statistically, the SNP might not even keep all of their six existing seats, but the worst is 66% in Dundee East.
This is all based on pure probabilities – see the histograms here: http://electionforecast.co.uk/graphics/2015_seat_histogram.svg – so may not work very well for such small numbers. For example, the statistics suggest UKIP are most likely to win two seats (with a 90% confidence interval of between zero to 5 seats) but http://electionforecast.co.uk/tables/UKIP_seat_gains.html shows only Clacton as nailed-on at 85%, with Oldham East and Saddleworth next at just 30% (where Labour are rated as 60% likley to keep it). The next one in the longer list is Thurrock at just 10%. The map and cartogram only show Clacton in UKIP purple.
Gordon was an 85% probability when I wrote this post, but the fact it’s already gone up to 89% I think confirms my conclusion about the SNP’s prospects!