Despite losing the Scottish independence referendum last September, the SNP have bounced back strongly. They now have support of around 44% in Scottish polls for Westminster voting intention. If the pollsters are right, then they would win around 47 of the 59 seats in Scotland.
This is obviously bad news for Labour, but it is also a headache for the Conservatives.
The illustration shows a new version of the classic two-party swingometer. Although it will be a multi-party election, we can distill it down to a single dimension of Con/Lab swing. Let's assume that support for the Liberal Democrats, UKIP, Greens and SNP remains constant until May. But let the Labour and Conservative vote shares vary, as long as they add up to their current total of 65%.
The swingometer shows that Labour will win an overall majority if they have a lead of at least 4.2% over the Conservatives. But the Conservatives need a lead of 7% to have a majority themselves. They even need a lead of 1.4% just to have as many seats as Labour. This apparent bias is mainly due to two vagaries of the electoral system: low turnout in Labour areas and the failure of the coalition to update constituency boundaries.
The hung parliament possibilities are the five zones between the two overall majority areas. One sector labelled "Con choice" represents the outcome where the Conservatives are just short of a majority and can choose a coalition partner from either the Liberal Democrats or the SNP. They would probably choose to continue the coalition with the Lib Dems.
The adjacent zone, labelled "Con + Nat", is where the only feasible two-party coalition would be the Conservatives and SNP. Nicola Sturgeon appeared to rule this out in November when she said "the SNP will never put the Tories into government". Nevertheless, the scenario could still happen, and it would have to be resolved with hard choices about either doing a deal or having a second election.
The SNP would prefer to think about two other sectors "Nat choice" - where they can choose between the two big parties, and "Lab + Nat" - where there is only one feasible two-party outcome. In either case, we can probably expect a Labour/SNP pact of some sort. Alex Salmond would be in a strong position to drive a hard bargain with Labour in exchange for SNP support.
The swingometer arrow is currently indicating a Labour lead of 1%, which would give a Labour/SNP outcome. Time is running for Ed Miliband to persuade many more people to come over to the Labour side. The Conservatives are hoping for some big pro-government swing, as helped them in 1987. But the needle has to move if Alex Salmond is not going to seize a historic victory from his earlier defeat.
We ran some scenarios by moving Conservative and Labour vote shares, keeping the other parties fixed. The scenarios were parameterised by the Conservative lead over Labour, which can vary from -10% (strong Labour lead) to +10% (strong Conservative lead). The table shows the result of the scenarios at a gridding of 1%.
|Con %||Lab %||CON|
|-4.0||30.5||34.5||240||323||17||1||1||49||19||Lab choice of Lib/Nat|
|-3.0||31.0||34.0||251||313||16||1||1||49||19||Lab choice of Lib/Nat|
|1.0||33.0||32.0||278||286||15||1||1||51||18||Nat choice of Con/Lab|
|2.0||33.5||31.5||285||281||13||1||1||51||18||Nat choice of Con/Lab|
|3.0||34.0||31.0||290||276||13||1||1||51||18||Nat choice of Con/Lab|
|6.0||35.5||29.5||315||254||12||1||1||49||18||Con choice of Lib/Nat|
|7.0||36.0||29.0||324||248||10||1||1||48||18||Con choice of Lib/Nat|