Election Prediction May 2015
So, in case some of you haven't noticed, there's a general election tomorrow. I'm less interested in trying to predict it accurately (that's a job for other people) as I am in trying to capture what seems obvious now, before the votes are in.
I was going to put just "2015" in the title, but I agree with Nick that it is not impossible that we end up having two general elections this year, although I don't think it's massively likely, and I disagree with him about how crucial having the LibDems in Government is to the question.
About a month ago I wrote "the only thing that's very clear about this General Election from the start of the campaign is that it's notably not very clear". That has become even more clear during the campaign, even though the polls have hardly moved.
It looks very clear - almost certain - that no party will win a majority. It also looks very likely that the Tories will get more seats than Labour, but not by much in the usual run of things - and this is not the usual run of things. A Tory+LibDem combo looks likely to be short of a majority too, and Labour+LibDem will probably be well short of one too. Labour+SNP might scrape a majority or might fall short. Labour+SNP+LibDem looks like a clear majority, but is not going to be easy to do. If you consider all the potential results of a general election, there aren't many combinations of votes where there is no workable two-party coalition apart from a grand Labour-Tory one, but that is where most predictions have us ending up.
I think this election has a higher chance of a big surprise result than any I can recall, with the possible exception of 1992. I don't think a surprise result is very likely, but it's more likely than usual. The polls have to make all sorts of assumptions, and they seem more likely to not hold in these circumstances than in previous elections. Within the models, it looks pretty certain that no party will get a majority, and the Tories will probably beat Labour in terms of seats, but there is more than unusual uncertainty about how accurate those models will prove to be. What sort of big surprises could there be? Any party getting a majority seems the most likely. A collapse or triumph for UKIP. A complete wipeout of the LibDems. SNP support being substantially overstated (it seems hard to imagine that it's understated).
What happens politically on that basis is very murky, and the parties have been taking care to make it harder for themselves. It does seem that a minority Government is the most likely outcome, but whether the confidence-and-supply support is tacit or explicit or some combination is not clear. It's not even certain that if the existing coalition retains a majority of seats that it will continue as a coalition Government. It's also not certain that if Labour+SNP have a clear majority they will be willing or able to put Ed Milliband in Downing Street, which seems absurd.
I still don't quite understand what Ed Milliband is thinking in dissing the SNP so heavily on the UK stage. I mean, obviously that's the message they need in Scotland, where it's largely a Labour-SNP fight. But to hammer on that theme to an English audience seems unwise. The Labour-leaning people I know in England seem generally pretty impressed with the SNP, apart from the independence business. Presumably they're guessing that it'll play well with wavering Tory/Labour voters. But I suspect it'll also play badly enough with disaffected potential Labour voters to reduce turnout and/or persuade them to vote Green or even LibDem.
A Labour-Tory Grand Coalition seems almost impossible. But I am pretty sure that we'll get one, or at least some sort of agreement, if the uncertainty about Government and supply becomes so great that there is spectacular turmoil in the markets. It looks likely that things will drag on for much longer than the UK is used to, but having briefly looked in to the constitutional position about Supply Estimates, I think we will start to have genuine problems if we can't get a budget passed before Christmas, and big problems if the details aren't through by the start of the next financial year. If those circumstances start to arise, I'm pretty sure a default compromise budget would be agreed and passed to keep things going. Although I also expect a lot of unedifying brinkmanship.
What'll happen to the party leaders is also massively up in the air, whatever the outcome. With Cameron, the question is merely whether he will go soon or very soon, even if he is PM. I suppose if he wins an outright majority he'll be secure for a year or so, but only if his heart is in it, which it may well not be. Miliband is probably only doomed if he isn't Prime Minister. Clegg will probably go, but is likely to be a little more dignified about it (i.e. take a bit longer) than those two. Farage will stay if he wins his seat, but will probably go if not, and almost certainly if UKIP win one or no seats. Natalie Bennett of the Greens will probably go at the earliest decent opportunity. Leanne Wood of Plaid Cymru is probably secure unless the polls are badly wrong and they lose seats. I've not tracked Northern Ireland much, which might be a mistake - the DUP's 8 and the SDLP's 3 may be important in a way they haven't been in the past. Though I said that in 2010, and they weren't, in the end. The one leader likely to emerge with a massively strengthened position is, of course, Sturgeon.
One thing that often happens in elections is that the news on the day after is how well parties did compared to what was expected, rather than compared to the last election. So if, for example, the SNP fail to win every seat in Scotland, or at least all but a tiny handful, it'll be seen as a bit disappointing. This seems an absurd spin on what will almost certainly be a spectacular win. They only have six MPs at the moment!
The media will be keen to anoint the party with the most seats as the 'winner', if it is the Tories, but they will almost certainly have lost seats, which is hardly winning in my book. The LibDems will lose even more seats - the little party always gets smashed - but will probably regard anything over 20 as not too bad, which seems odd when they will have lost more than half their MPs. Labour may not actually net gain that many seats, but if they gain more than about 15 will claim it as a good result.
My own constituency, MK South, I'm not sure how to call. The polls and predictions and bookies say Tory hold, so if I had to put money down that's where I would. But the Labour campaign has got better, and looks local, whereas the Tory campaign looks like a national-cut-and-paste job. Talking to friends and work colleagues, I've heard examples of the sitting MP being rubbish on bread-and-butter constituency work, which reduces the incumbent effect (because doing bread-and-butter constituency work is part of what it is). It's a classic Tory/Labour marginal, but there is substantial local LibDem support (18% last time). With the national focus on tactical voting, I would expect that to reduce, and it may even collapse ... but of course there's no guarantee that vote would go to Labour. In many ways it's a microcosm of the national scene: the polls clearly say it's a narrow Tory win, but it would not be as surprising as it normally would be to find the predictions wrong.
Everybody is forgetting that there are also local elections taking place in almost all parts of England. So I'm going to largely do the same, despite the fact that these are some of the largest local elections for a long time. Without having paid the campaign (what campaign?) any attention whatsoever, and without really thinking about it, I would expect Labour to gain a bit, largely at LibDem and Tory expense, plus a smattering of UKIP. But perhaps less of a shift along those lines than in the national polls because there's been local elections since 2010.
There's also a local referendum in MK on the neighbourhood plan for Central Milton Keynes, the out-of-town shopping centre in the middle of MK. The publicity for this has been shockingly poor. I only heard about it at all because my village council sent a note round saying that there wouldn't be a Town Council election because nominations matched vacancies, so we'd only have the general election and the local plan referendum to vote on. And almost all friends and colleagues I speak to hear about it from me for the first time. I expect it'll pass. I've had a skim of it (it is a vast and unfriendly document, even if you know something about planning), and I didn't spot anything badly wrong with it. The question is something like "Do you want MK Council to use the neighbourhood plan for Central Milton Keynes to help it decide planning applications?" which strongly invites a reaction like "Umm, I suppose so? Don't they do that already?". So unless there's a clear campaign against it, and there is no mention of it locally, it'll go through on momentum - which is probably by design. My village council was one of the first in the country to produce one of these, creating a minor masterpiece of soft nimbyism, and it was passed with 90% voting in favour. I expect the MK one will pass by a much smaller margin.
Interesting times!
This entry crossposted to http://doug.dreamwidth.org/292626.html, where there are
comment(s) not shown here.
I was going to put just "2015" in the title, but I agree with Nick that it is not impossible that we end up having two general elections this year, although I don't think it's massively likely, and I disagree with him about how crucial having the LibDems in Government is to the question.
About a month ago I wrote "the only thing that's very clear about this General Election from the start of the campaign is that it's notably not very clear". That has become even more clear during the campaign, even though the polls have hardly moved.
It looks very clear - almost certain - that no party will win a majority. It also looks very likely that the Tories will get more seats than Labour, but not by much in the usual run of things - and this is not the usual run of things. A Tory+LibDem combo looks likely to be short of a majority too, and Labour+LibDem will probably be well short of one too. Labour+SNP might scrape a majority or might fall short. Labour+SNP+LibDem looks like a clear majority, but is not going to be easy to do. If you consider all the potential results of a general election, there aren't many combinations of votes where there is no workable two-party coalition apart from a grand Labour-Tory one, but that is where most predictions have us ending up.
I think this election has a higher chance of a big surprise result than any I can recall, with the possible exception of 1992. I don't think a surprise result is very likely, but it's more likely than usual. The polls have to make all sorts of assumptions, and they seem more likely to not hold in these circumstances than in previous elections. Within the models, it looks pretty certain that no party will get a majority, and the Tories will probably beat Labour in terms of seats, but there is more than unusual uncertainty about how accurate those models will prove to be. What sort of big surprises could there be? Any party getting a majority seems the most likely. A collapse or triumph for UKIP. A complete wipeout of the LibDems. SNP support being substantially overstated (it seems hard to imagine that it's understated).
What happens politically on that basis is very murky, and the parties have been taking care to make it harder for themselves. It does seem that a minority Government is the most likely outcome, but whether the confidence-and-supply support is tacit or explicit or some combination is not clear. It's not even certain that if the existing coalition retains a majority of seats that it will continue as a coalition Government. It's also not certain that if Labour+SNP have a clear majority they will be willing or able to put Ed Milliband in Downing Street, which seems absurd.
I still don't quite understand what Ed Milliband is thinking in dissing the SNP so heavily on the UK stage. I mean, obviously that's the message they need in Scotland, where it's largely a Labour-SNP fight. But to hammer on that theme to an English audience seems unwise. The Labour-leaning people I know in England seem generally pretty impressed with the SNP, apart from the independence business. Presumably they're guessing that it'll play well with wavering Tory/Labour voters. But I suspect it'll also play badly enough with disaffected potential Labour voters to reduce turnout and/or persuade them to vote Green or even LibDem.
A Labour-Tory Grand Coalition seems almost impossible. But I am pretty sure that we'll get one, or at least some sort of agreement, if the uncertainty about Government and supply becomes so great that there is spectacular turmoil in the markets. It looks likely that things will drag on for much longer than the UK is used to, but having briefly looked in to the constitutional position about Supply Estimates, I think we will start to have genuine problems if we can't get a budget passed before Christmas, and big problems if the details aren't through by the start of the next financial year. If those circumstances start to arise, I'm pretty sure a default compromise budget would be agreed and passed to keep things going. Although I also expect a lot of unedifying brinkmanship.
What'll happen to the party leaders is also massively up in the air, whatever the outcome. With Cameron, the question is merely whether he will go soon or very soon, even if he is PM. I suppose if he wins an outright majority he'll be secure for a year or so, but only if his heart is in it, which it may well not be. Miliband is probably only doomed if he isn't Prime Minister. Clegg will probably go, but is likely to be a little more dignified about it (i.e. take a bit longer) than those two. Farage will stay if he wins his seat, but will probably go if not, and almost certainly if UKIP win one or no seats. Natalie Bennett of the Greens will probably go at the earliest decent opportunity. Leanne Wood of Plaid Cymru is probably secure unless the polls are badly wrong and they lose seats. I've not tracked Northern Ireland much, which might be a mistake - the DUP's 8 and the SDLP's 3 may be important in a way they haven't been in the past. Though I said that in 2010, and they weren't, in the end. The one leader likely to emerge with a massively strengthened position is, of course, Sturgeon.
One thing that often happens in elections is that the news on the day after is how well parties did compared to what was expected, rather than compared to the last election. So if, for example, the SNP fail to win every seat in Scotland, or at least all but a tiny handful, it'll be seen as a bit disappointing. This seems an absurd spin on what will almost certainly be a spectacular win. They only have six MPs at the moment!
The media will be keen to anoint the party with the most seats as the 'winner', if it is the Tories, but they will almost certainly have lost seats, which is hardly winning in my book. The LibDems will lose even more seats - the little party always gets smashed - but will probably regard anything over 20 as not too bad, which seems odd when they will have lost more than half their MPs. Labour may not actually net gain that many seats, but if they gain more than about 15 will claim it as a good result.
My own constituency, MK South, I'm not sure how to call. The polls and predictions and bookies say Tory hold, so if I had to put money down that's where I would. But the Labour campaign has got better, and looks local, whereas the Tory campaign looks like a national-cut-and-paste job. Talking to friends and work colleagues, I've heard examples of the sitting MP being rubbish on bread-and-butter constituency work, which reduces the incumbent effect (because doing bread-and-butter constituency work is part of what it is). It's a classic Tory/Labour marginal, but there is substantial local LibDem support (18% last time). With the national focus on tactical voting, I would expect that to reduce, and it may even collapse ... but of course there's no guarantee that vote would go to Labour. In many ways it's a microcosm of the national scene: the polls clearly say it's a narrow Tory win, but it would not be as surprising as it normally would be to find the predictions wrong.
Everybody is forgetting that there are also local elections taking place in almost all parts of England. So I'm going to largely do the same, despite the fact that these are some of the largest local elections for a long time. Without having paid the campaign (what campaign?) any attention whatsoever, and without really thinking about it, I would expect Labour to gain a bit, largely at LibDem and Tory expense, plus a smattering of UKIP. But perhaps less of a shift along those lines than in the national polls because there's been local elections since 2010.
There's also a local referendum in MK on the neighbourhood plan for Central Milton Keynes, the out-of-town shopping centre in the middle of MK. The publicity for this has been shockingly poor. I only heard about it at all because my village council sent a note round saying that there wouldn't be a Town Council election because nominations matched vacancies, so we'd only have the general election and the local plan referendum to vote on. And almost all friends and colleagues I speak to hear about it from me for the first time. I expect it'll pass. I've had a skim of it (it is a vast and unfriendly document, even if you know something about planning), and I didn't spot anything badly wrong with it. The question is something like "Do you want MK Council to use the neighbourhood plan for Central Milton Keynes to help it decide planning applications?" which strongly invites a reaction like "Umm, I suppose so? Don't they do that already?". So unless there's a clear campaign against it, and there is no mention of it locally, it'll go through on momentum - which is probably by design. My village council was one of the first in the country to produce one of these, creating a minor masterpiece of soft nimbyism, and it was passed with 90% voting in favour. I expect the MK one will pass by a much smaller margin.
Interesting times!
This entry crossposted to http://doug.dreamwidth.org/292626.html, where there are
comment(s) not shown here.