Democracy in the UK #4: Calling it

Who's going to be Prime Minister?

It's tricky, because polling averages are showing Labour and the Conservatives both stuck at around the 33-34% mark. That hasn't changed for a long time, and if it's reproduced at the election there'll definitely be a hung parliament - so the question turns on predictions about how politicians will behave, which are always dicey, rather than just interpreting polls.

So, first things first, what are the possibilities? I've taken six election models and looked at their predictions about who'll win seats in the new parliament. Three (The Guardian, May2015, and Electoral Calculus) average national, regional and constituency polls to show what would happen if an election was held today; three more (Election Forecast, Elections Etc, and Polling Observatory) are genuine forecasts of what will happen on May 7th, which predict organic shifts over the remainder of the campaign based on history.

323 seats are needed to win a confidence vote, and probably about ten more to produce a sustainable government. Tallying the possible governing alliances in each of those models produces this (click to enlarge):

The first takeaway is this: David Cameron is exceedingly unlikely to be PM after this election. Elections Etc is easily the most optimistic model about the Tories - it predicts they'll win 292 seats, ten more than the next highest prediction. Even on that prediction, they'll only just be able to scrape through a confidence vote if they pull together all possible partners -  the Liberal Democrats, UKIP, and the Democratic Unionist Party from Northern Ireland. That won't happen: UKIP and the Lib Dems do not get along, especially over Europe. So the only way Cameron could survive a confidence vote would be if left wing parties let him, by abstaining. The SNP has sworn repeatedly that it won't do that. It would be staggering if the Labour Party did, though it's conceivable that if there was a leadership contest in the party they wouldn't want an election immediately. (I owe that point to Dan Berman.) But in most of these scenarios there's an at-least-mathematically viable Labour government, so it's hard to see why Miliband would be seen to have done so badly that he should be challenged. All in all, I don't think there's any way the Tories can form a government with things as they currently are.

So the negotiations will centre on the prospect of a Labour government and who might prop it up or be brought into it. One possibility getting some airtime recently is a Labour-Lib Dem coalition: Labour would supposedly prefer, for both principled and cynically electoral reasons, the Lib Dems to the SNP. But on none of the predictions would an alliance of those two parties even come close to a majority. Labour and the SNP don't actually fare much better. A formal coalition has already been ruled out, though an informal arrangement could just about get a Labour minority government through a confidence vote. But Labour don't want to be - or be seen as being - entirely at the mercy of the SNP to survive knife-edge no-confidence motions, and more importantly, without any formal arrangement they'd need to find 50-60 votes on each and every bill to get it through. That's not impossible - the ALP managed something similar for three years, and though there'd be a lot of horse-trading required, it's likely that most Labour policy could get support from either the SNP or the Tories - but it's unlikely to be workable.

For much the same reason, an alliance of Labour, the SNP and the minor left-wing parties - the Welsh Plaid Cymru, Northern Ireland's Social Democratic and Labour Party, and the Greens - would be very scrappy. Miliband would struggle to pass the parts of Labour's programme which those parties characterise as "Tory-lite", and though it could probably get through with Tory and Lib Dem help, running your government by relying in parts on two different party groupings with totally opposed views about appropriate policy would be extremely unstable.

The Labour/Liberal/SNP column is solid green: those three parties together would control a solid majority. One possible arrangement is that Labour and the Lib Dems enter a formal coalition which would control about 300 seats. From that number, workable minority government is possible, especially with a bloc of about 60 seats controlled by left-leaning parties. But the 300 would have to be highly disciplined - if Lib Dem rebellions happen too frequently, Labour will be back to difficult negotiations to get the 50-60 votes it needs. And in that scenario - and in fact, less strongly, in any coalition scenario of this sort - it's an open question whether Miliband and the Labour leadership would be willing to give out ministerial positions to another party for the sake of being a not-quite-as-small-minority government. One other factor which we have almost no information on is the Lib Dem-SNP relationship: people aren't talking about it, since obviously those two parties can't form a governing alliance. But in a three-party arrangement of this sort, there'd have to be at least an informal deal with the SNP probably involving some concessions to their program. Whether the Liberal Democrats are willing to be in a coalition making those concessions is an interesting question - I suspect they would, but we don't really know that much about it.

On balance, I think this last possibility is most likely: Miliband as PM at the head of three-party alliance, which the Lib Dems are formally involved in and the SNP supports from the outside on confidence and supply. Even that, though, wouldn't be a super stable relationship. How long it lasts will depend on the SNP's calculation of its self-interest: it could at any time withdraw its support and trigger new elections by supporting a no-confidence motion. Doing that could trigger a further surge in their support, if it's seen as a result of Labour ignoring Scotland's interests and voice, or it could cause a wave back to Labour if voters start to take seriously the message that a left-wing government can only be guaranteed if Labour wins Scottish seats. I'm not sure which of this is more plausible - it will obviously vary depending on the time and circumstances - let alone which Nicola Sturgeon and Alex Salmond think is more plausible. So it could be an exciting year.

Here is a version of the spreadsheet, with a manual input to add to the forecasts extra seats switching to the Conservatives from the Lib Dems and Labour, and from Labour to the SNP. (You can enter negative numbers if you want to swing the other way; at the moment those are the only inputs as I think Labour and SNP gains from the Lib Dems are fairly settled.) You can access it in Google Docs here. May2015 and Electoral Calculus have tools where you can see how changing vote shares would translate into seat swings; obviously you can see what their models say would result on those pages, but if you want to see the impact in all the models you can use this one.