The Guardian reports the predictions of Prof Paul Whiteley of the University of Essex. His study forecasts Labour with 291 seats, Conservatives 281, Lib Dems 48 and others 30. The study is based on a model used by the British Election Study which came up with some pretty accurate predictions for 2005 and 2010.
The article explains:
The Essex forecasting model works by combining the number of seats won by parties in the previous election with voting intentions data from polls conducted six months prior to the election.
Whiteley acknowledged that the model forecast a relatively large number of Lib Dem seats, but that appeared to be because there is a stronger incumbent effect for Lib Dem MPs than any other parties, partly due to many of their MPs long holding marginal seats.
It would be good to see some informed debate about this model in the Comments below.
Follow @libdemvoice.org on Bluesky
Like us on Facebook
Subscribe to our feed
Sign-up for our daily email digest
Lord Bonkers: "Something less terrible than the truth"
Two new Makerfield by-election polls gives Burnham 5 and 12 point leads
Reform under water
expats
David Raw
Tom Reeve
Nonconformistradical
Sam Ammar
46 Comments
Rigged poll if you ask me lol you in for a surprize we will see when real poll done
My gut says this is a bit high, but maybe not too high. See for example, this article:
http://www.standard.co.uk/news/politics/libdems-need-to-win-45-seats-for-clegg-to-stay-on-as-leader-10011033.html
which seems to suggest that this is a reasonable number of seats for party members (and MPs!) to expect.
I think we should ban all opinion polling articles in all papers and make every paper, blog and news TV / radio channel print/broadcast/chant every single day the following message in very large letters/very loudly:
GENERAL ELECTION 2015 – WE DON’T KNOW WHAT WILL HAPPEN UNTIL YOU VOTE.
PLEASE COME BACK IN MAY SO WE CAN ANALYSE HOW YOU VOTED.
Possibly the worst outcome of the election there could possibly be for the Lib Dems as a Party.
If the predictions using this model were reasonably accurate in 2005 and 2010, why not this time around? Seems like some good news at last – although too good to be true?
Judy’s cautious response matches my own. If true, fantastic – Ashcroft’s latest polling makes the point about how MPs such as Paul Burstow are defying national trends – but 5 years of political bruises make me cautious. Then again, should be surprised that hardworking and well-liked MPs are doing well?
Basically, this poll is saying that the only losses for the Lib Dems will be in Scotland. LOL. If you believe that, then I have abridge in central London I want to sell.
Given you are ignoring the nature of the work Tsar Nicolas, I take it that you actually were duped into buying a bridge and are now trying to sell it…
“Whiteley’s forecast, based on mathematical modelling, focuses on what happened to seats in previous election rather than the overall share of the vote and is based on a model developed while the British Election Study was based at Essex. This model successfully predicted the outcomes of the 2005 and 2010 general elections.”
It is not a poll….
The model is making assumptions based on previous elections
But none of the previous elections had two incumbent parties. None of them had the Libdems as an incumbent, and none of them had five parties in England and six in Wales and Scotland capable of winning a seat.
The previous predictions had large margins of error of 10-20 seats, so when predicting a party with as few seats as ours it is fairly useless.
I won’t say it’s impossible but if you look at the models from the likes of Elections Etc, their models are predicting far fewer seats for the Lib Dems; and these models are more advanced than Prof. Whiteley’s. Beware of favouring models simply because they give results you’d like to see.
Electoral Calculus, which was pretty good last time, today, has us, the Lib Dems, at 17 seats.
That would be much better for the future of the party, we could reform and recharge in the peace and quiet of opposition.
This isn’t a poll, its a model that combines multiple polls with past election results to predict the election.
Like most models, it is probably wrong in the absolute specifics, but does indicate that a combination of Liberal Democrat resilience in held seats and a lack of cohesion among the other parties may well lead to us holding on more effectively than a universal swing might suggest.
With challenging Labour and Tory candidates’ voteshare being split by insurgent Green and UKIP candidates, in our held seats we might well be the beneficiaries. In other seats, though, this same effect will see us squeezed down to low single-digits even in constituencies that were and might again be very fertile ground for us.
I still don’t get the feeling that we’ll get more than 35 seats. But at the same time, I also don’t feel that we’ll get less than 25.
I’d agree with Simon; it would be lovely if this were true but it’s an election that is unprecedented so any model is going to have to fill in educated guesses about how, for example, people will respond to an incumbency Lib Dem party – will they vote for the local MP they like or against a party they have disliked in government, and so on.
This looks like the right sort of approach, the big problem comes with Fixed Term Parliaments which have effectively extended the Mid Term. Because of the complete absence of speculation about the date of the Election, 6 months before felt more like 2 years before. The changes that used to take place in the last 6 months will be concentrated into a much shorter period. The overall effect will be to boost Labour & the protest Parties in the polls & depress us.
My expectation therfore is that we will do a lot better than this study predicts & Labour a lot worse. The slow decline in Labours polling will accelerate, in fact thats already started.
Some people think naysayers are a pain in the butt. Well, I think yeasayers are a pain in the butt. LDV has made a big fuss about this prediction, because it is a high one. LDV keeps quiet about Electoral Calculus, and any other predictions which come out low. I have no idea whether 17 seats or 48 seats are more likely, but I do recognise the bias of rose-tinted spectacles when I see it. Rose-tinted spectacles, along with Clegg and along with The Prostitute State, will kill this Party.
@theakes
“Electoral Calculus, which was pretty good last time, today, has us, the Lib Dems, at 17 seats.”
Have you ever bothered to look at any of the individual seat predictions? If you did, you would agree that Electoral Calculus is a joke.
All that is clear is that you cannot take ‘national uniform swings’ and assume that the same sort of swing that may occur in Lab v Con battleground seat will apply for example in a Lib Dem v Con seat (especially where we have a good sitting MP or strong candidate). Prior to 1997 you could get 14:1 at William Hills on the Lib Dems getting 32 or more seats (see my tip in the New Statesman at the time). We were at 9.5% in one poll just before Christmas 1996 but eventually polled 17% on polling day (which was 1% less than in the 1992 General Election). But I was personally optimistic as I ran the target seat campaign – and our Campaigns team helped us to more than double our number of MPs to 46 in 1997.
The article says —
“.. The study is based on a model used by the British Election Study …”
Which model? Kate Moss ?
That is unfair to Kate Moss, who would make a more informed guess.
The Prof Whiteley prediction is that “others” will only get 30 MPs in 2015.
Keep in mind that in 2010 there were 29 others and that 18 of those were from Northern Ireland.
So it would be interesting to know what Prof Whiteley thinks is happening in Scotland and Wales in 2015.
His model tells him that the SNP, PC, UKIP, The Greens and Respect will get just12 MPs between them ???
@Chris – so true. As I comment elsewhere I am currently in Australia where a few days ago there was a general election for the Queensland state government. At the previous election 3 years ago the National Liberal Party stormed home with 78 out of the 89 seats and the Labor Party were reduced to 7. But in Saturday’s election this was completely turned around and Labor won the election and will form the new government. The TV “experts” discounted such a result as recently as the day before the election. All of these polls and projections are exactly that: and with non-uniform swing in a small set of seats there may well be a very fine dividing line between a good-ish result and disaster.
I take as a useful starting point that, following detailed analysis of the 2014 local election results, Rallings & Thrasher in the Sunday Times projected 37 Lib Dem seats to be won in 2015.
Theakes. You’d prefer us to win 17 seats rather than 48 and be out of government.
Remind me what the point of politics is again.
Local elections and national elections are completely different.
Many people will still vote for us locally but not nationally (or indeed internationally, just compare council and european election results last year).
Since we increased the number of votes yet lost seats in 2010, it is obvious that a lot of ‘wasted votes’ came our way. Though not good for morale, losing these votes to Labour, Greens, SNP etc will not lose any seats for us.
Tony Dawson: Can you really not think of a worse result for Lib Dems? Nonetheless, I think I can see what you mean, a further coalition with either party under a situation where our overall support would be widely seen to have collapsed would be a lethally poisoned chalice. If a result anything like this did emerge Cameron as incumbent PM would not be obliged to step down, yet according to Lib Dem invented protocol, the largest Party should have first attempt at forming a government. I would hope that the leadership would have the courage and foresight to be quite prepared to walk away if not satisfied with proposed policies. Personally, I do not want the party yet again yoked to the Tories and a Cameron led and inevitably clumsy ‘renegotiation’ with EU leaders followed by a disastrously timed referendum on the EU should in my opinion be out of the question. We should say if that is what you intend you are on your own. I do not think in the circumstances that a formal coalition with Labour would work. If forced into it, I think we should concentrate on dominating one or two select ministries rather than seek representation across the board, but it would be better, at least to begin with, to invite Labour to try to run a minority government.
It seems a fairly safe bet that minority or majority a Labour administration would very rapidly attract intense hostility: I would not expect any ‘honeymoon’ period.
This is an interesting election forecast, but without some idea why it was 14 seats out on the Conservatives in 2010 (they predicted they would get 293 but they got 307) it is hard to decide how much weight to give it. Does it look at the effects of new candidates? Simon Oliver raises an interesting point has being in government for the Liberal Democrats been taken account? (As the British Election Study has been going since 1964 they should have some knowledge of what happened to Liberal MPs in 1979 – where we lost 4.5% of the vote but won two MPs less than in October 1974 with three defeated MPs.)
Simon Shaw and his Rallings & Thrasher projection in the Sunday Times might be closer on 37. I believe we will have more than 28 MPs and hope for more.
We will just have to wait and see. Given the circumstances I would be very happy with a result any where near what is being predicted in the article.
There are two things that really matter at the moment.
1) Is the General Election team targeting its mobile resources at the right seats? That is those seats that we shall win or lose by say 1,000 votes. So, if the team really is basing its resource dispositions of 45 seats, is that a wise frontier at which to be aiming mobile resources?
2) Are we helping 2010 and present day tactical voters ‘lend’ us their vote? A recent poll suggests that 60% of those telling pollsters that at General Election tomorrow they will vote LD say they could still change their mind. The figure for Ts and Ls is circa 40%.
The party will have its own private polling to help it answer 1) above. I was interested to see that Paddy was in Sheffield Hallam at the weekend.
Like us, it also has the Ashcroft constituency polls. These suggest that we shall win circa 30 seats provided nothing goes wrong in the seats that he has not polled, expecting them to be ‘safe’. Many of these will be Scottish seats and he is due to publish polls for those seats tomorrow.
So by tomorrow we shall know whether the Ashcroft polls still point to 30 seats or whether that has gone down.
Ashcroft also conducts Focus groups – this week one in Sutton and Cheam and one in Elmet & Rothwel.
A well conducted Focus group can be very revealing, digging out nuanced positions. Here is how our report card read: ‘Nationally, the party’s own road poster, steering between reckless borrowing to the left and reckless cuts to the right, summed up the Lib Dems’ approach for these participants – for whom staying the course, as urged by the Tories, meant voting as they did last time. But those who did not vote Lib Dem in 2010, or did so in West Yorkshire, were either dissatisfied with this rather non-specific approach or did not accept that the party was still equidistant between Labour and the Tories: “I struggle with the whole Lib Dem thing since the coalition – he might as well be a Tory.”’
Overcoming this to improve numbers voting for us tactically will be very hard. And made harder by our continuing preference for articulating an anti-Labour line nationally. We can safely leave the Tories to do that job for us, without jeopardising the support of Labour voters in our Tory facing marginals.
I still find the ground war and the air war facing in opposite directions.
Bill le Breton 3rd Feb ’15 – 10:27am
“..I still find the ground war and the air war facing in opposite directions.”
Could that be because those in charge of the campaign are all at sea?
When the party’s most famous sub-mariner has taken to the hills in Hallam it raises some questions.
Have the high command made a ‘strategic withdrawal’ to the bunker?
So 281 (Con) + 48 (LD) = 329 ==> working majority
So if that’s true, then the coalition continues for another 5 years.
If we do that well, I’ll be surprised. But perhaps jez and Mr Romanov will also be surprised. We shouldn’t go below 30. It may be that the calculation unduly weights towards performance last time round as the “Others” total also looks a bit low to me. After all, there are 18 seats in Northern Ireland and Plaid Cymru should get three at least, plus at absolute least one for UKIP, so that would leave eight max for the SNP!
Possibly the Ulster seats are excluded from the calculation (the post does at one point say “British”), but in assessing what situation will result from the election, they shouldn’t be. Even if they are, assuming more realistically UKIP gets four seats and the Plaid, independents and Greens get four between them, that would leave 22 SNP, which still seems a bit low.
Is there to be another thread featuring the poll/study reported in today’s “Times” which predicts a LibDem wipe-out in Scotland, or does “The Voice” find that less interesting (or merely less palatable)?.
By the way, that study’s prediction for SNP successes and Whitley’s prediction of only 30 “others” for the whole UK cannot both be right.
@Nick Collins
Personally, I’d have no problem it in being report on the Voice as it will give ample room to discuss how weak the methodology is behind it. Will Orkney really being voting SNP when it was so against Indy (and has voted Liberal for 60 years)? I have my doubts….
Any prediction that looks at only national polls is on a hiding to nothing.
Well, it will be bizarre if LDV does not cover the Ashcroft polling news tomorrow. There is still time for Paul Barker to buy LD seats at 29.5. And the ‘nay sayers’ to sell at 27.
John Tilley I was really getting fed up with your comments but then you made me LOL with all at sea, thank you. After the Europeans it seemed to me that the electorate really hate us and kicked us where it hurts most. If this is indeed the case then it’s obvious that Nick Clegg’s seat is in grave danger. As for other seats it will depend how successful local MPs and activists are at persuading the voters about the good things achieved locally and nationally. It will not depend on where we are in national polls but could everyone in the public eye please stop talking about coalition with anyone in the future as it confirms the impression that we don’t care how many promises we break as long as we stay in power and we’ll be lucky to keep any seats if people believe that.
@Bill le Breton:
“There are two things that really matter at the moment.
1) Is the General Election team targeting its mobile resources at the right seats? ”
I doubt it very much. I remember well Tessa Munt MP and her crew coming into a campaigning fringe meeting at the Liverpool Conference in 2005 and telling us all how Cowley Street spent the last few weeks of the 2010 campaign telling her campaign workers to go and work somewhere else. One has to ask “what political nous have these people?” If they have ‘bough in’ to the Cleggista mantra then the answer is “probably rather little”.
I would say 20 seats, and that’s being generous.
Ashcroft shows 26% swing to SNP in Alexander’s seat , a 16% swing in Gordon (both SNP gains) and Lib Dem.vote collapsing in other seats. So no comfort there. Wonder if Voice will write it up?
Ashcroft’s Scottish poll is now out. The results are grim for us: he projects that we are handily beaten in the two LD seats he polled. Inverness, Nairn etc (21% LD versus 50% SNP) and Gordon (26% LD versus 43% SNP). And we are projected to lose our deposit in every other seat polled – including Glasgow North where we polled 31% in 2010.
Of course the huge surge in SNP support makes it impossible to extrapolate these poll results to the rest of the country but they suggest that the raft of Scottish polls have been broadly right in predicting 1 or 2 LD holds and probably 47 lost deposits.
Hireton
Paul Walter has done a piece this morning.
Paul in Wokingham (or should that be Wokingham in Oz ?)
The arithmetic in Scotland makes a nonsense of the Whiteley’s model. If we lose 9 or 10 seats in Scotland we would have to hold every single seat in England and Wales to even approach the much hyped figure of 48 Lib Dem MPs.
But cheer up because that other Paul confidently says in his comment —
paul barker 2nd Feb ’15 – 7:24pm
” This looks like the right sort of approach, … …
My expectation therfore is that we will do a lot better than this study predicts …”
John Tilley
“The arithmetic in Scotland makes a nonsense of the Whiteley’s model. If we lose 9 or 10 seats in Scotland we would have to hold every single seat in England and Wales to even approach the much hyped figure of 48 Lib Dem MPs.”
A point I tried to make earlier in the tread, but got scorned/mocked/ignored.
Oh com’n – are you kidding?
The Panda joke is about to be shared in Scotland with the Tories.
The best forecast on current data is around 24 and I think that is very generous.
Tsar Nicholas,John Tilley spot on comments.I think we can safely ditch the BES forecast
I Think UNS for Sctoland and England will be still the best way of forecasting the number of Lib Dem seats.However it is worth pointing out that the swing figures fed into the Baxter model are heavily weighted towards You Gov figures because of the high frequency of their polling. I think You consistently under forecast by around a couple of %lib Dem shares .So to get a top end seat forecast I would use ICM.
As far as Ashcrofts constituency question is concerned because this is a new innovation for this Parliament there is no proven track record.
I know the arguments about how our ground war is strong but we need a good air war to get our opinion poll ratings beyond 10% ,and to just help out in the many tight Lib Dem Tory marginals.
@ Tony Dawson @ Bill le Breton ” Is the General Election team targeting its mobile resources at the right seats? ” The answer of course is that the General Election team, controlled by Nick and his paid supporters at the centre, will be targeting its mobile resources at the seats of people loyal to Nick.
in very approximately 35 seats we are 1st, tories 2nd , others way behind. tories have kindly opened up clear blue water between us and them. LD 2010 supporters in these seats have a choice–vote LD or if they vote ed they may get a dose of dave in no 10. I think decent mp’s have an excellent chance of holding on
its not 1 election but 600+ ones– a consequence of FPTP
& we may gain some seats……..
dont be surprised if a balanced parliament leads to a lab-con coalition
they have more in common than they like to admit & 2015 is yet another election thats the one worth losing — just like 2010 as the then governor of the BoE said at the time
it may be worth a bet on us being HM loyal Opposition post may 2015
I read the same study and pleasantly surprised by the optimistic reading for the lib dems. I do however think the model may not be as robust this year.
About 1 in 6 votes cast were tactical votes (http://www.envplan.com/abstract.cgi?id=a43512) which I think helped the Lib Dems. I believe tactical voting will matter a lot in many swing seats but it won’t necessarily be straight for the lib dems. It could just as easily go to the Greens, UKIP or Labour depending on the local context. I think many areas where UKIP may have a chance of getting in will end up with tactical voting.
I estimate around 30% of those who regularly vote locally may sway at the last minute. There may be a lot of mud being thrown, but I think the parties most at risk are those not well funded to make effective rebuttals. Those wavering over Labour may be put off with negative campaigning and attack ads. The study calibrates results off the back of polls where people declare their preference, but suppose I am right that 30% may sway by May, then that would shift results in many marginals where labour/lib dems are running second.