The static campaign

The defining feature of voting intention in this election is how little it has moved. The graph below shows the UK Polling Report polling average for each week so far of 2015.


Things haven’t been completely static – at the beginning of 2015 Labour did still have a consistent tiny lead that faded towards a real tie over the first couple of months. There has been a genuine drop in support for UKIP and the Green party, albeit one that is no more than a point or two, rather than a really tight squeeze on their votes.

The broad picture though, especially over the short campaign, is one of no movement. This is not necessarily unusual – the huge ups and downs of “Cleggmania” in the 2010 election were not typical. Most historical election campaigns don’t show lots of movement (and I suspect some of that we did see is just the legacy of campaigns when there were far fewer polls, so a couple of outliers could more easily create the impression of movement when there was none).

Is there still time for a change, or are we doomed to have election result around about where we are, with the Conservatives and Labour pretty much neck-and-neck? Somewhere between one-in-six and in one-in-five people have postal votes, and many will already have voted, so they cannot change their mind any longer. Between 10% and 20% of people depending on the poll say they don’t know how they will vote, though some of this will be people who won’t actually end up voting but don’t want to say, and some of it will be people who don’t want to give their voting intention to an interviewer (“shy voters”). While it varies greatly depending on how you ask the question a further chunk of people who do give a voting intention say they may yet change their mind before they vote.

There are definitely plenty of people who say they may change their mind between now and May 7th… but I suspect this overestimates the volatility of the electorate and that most respondents who say they still might change their minds won’t do so, they just like to think of themselves as fair minded people who will consider all the evidence before making their mind up to vote for the party they were probably going to vote for to begin with.

Polls are, as ever, just snapshots of opinion now. They can only quantify what respondents themselves know – they can’t tell how respondents might react to, say, the party leaders Question Time Special next week, any as yet unknown and unexpected events in the final eleven days, or people genuinely recoiling one-way or the other at the very last minute. Realistically though, nothing has done anything to substantially change the polls in the last seventeen weeks and the parties are starting to run out of time for anything else to come along.

1,017 Responses to “The static campaign”

1 18 19 20 21

    You cant call “freaking out” negative and opposed to the spirit of this blog. Labour supporters trying to dismiss Ashcroft by saying “well look at Scotland or the North showing Labour behind Conservatives”. But how many people from the North or Scotland do they poll? 90 in Scotland or so, the margin of error must be huge. Other people here argue that a six point lead is likely a tie because the margin of error of 3 percent for each party. But its also possible in mathematical terms that the Conservatives are up 12 percent. I think we should respect all polling institutes, building an average like Realclearpolitics for US elections and then being pretty close to the outcome on election day. Some Labour supporters remind me of republican behaviour before the presidental election in 2012 saying ” the polls are rigged”, ” too many Democrats”, ” the ground game is telling the opposite” etc. We all know how the election ended back then. Obama won by an even bigger margin than most nationwade and state based pre-election polls suggested.


    I felt a bit sorry for Cameron today actually. Obviously he’d taken on board the ‘looking too laconic’ criticism and virtually bounded everywhere. He waved his arms a lot more and had his sleeves rolled up…. Mr Determined and energetic. However I’m not convinced its his natural style so I did smile a little to see it.

  3. Omnishambles,

    I think that some people can naturally be a bit surprised when certain sorts of arguments are used in an unfamiliar context. The SNP’s bargepole policy re: the Tories is so old that it’s become normalised, which is not the case for the most parties’ newfound bargepole policy re: the SNP.

  4. BigD

    You are correct. At least one of them is wrong.

    I hope it’s not the constituency polls or he will stop doing them and no-one else can afford to.

  5. Just checking the automod: Omnishambles.

  6. Bargepole.


    Thanks for your lengthy reply.

  8. I’m stumped. Now that I look at the comment, it’s maybe a bit snarky, but I doubt the software can detect that…

  9. McClane

    Getting a bit partisan there – in a similar spirit, given that the poll of polls remain about tied, one could easily make the same argument the other way.

    My main memory of 2012 is that Nate Silver got it right basically because he discounted any kind of ‘swingback’ or ‘shy GOP voter’ kind of theory about why the polls saying X actually meant Y and insisted that what the polls were saying as a whole was probably about right.


    Don’t worry about him, I’m sure he will be fine on the night.


  11. McClane,

    I think that there’s a tendency on all sides to believe variously anything except what the polls are telling us: that the main two parties are basically neck-and-neck. (Though not in bed with each other yet.) This hasn’t happened in an election since February 1974, which may explain the credubility.

  12. Omnishambles

    “The Lib Dems later clarified that the comments were made in the context of a formal coalition deal”

    That makes it even more bizarre.

    1.Lab/SNP no deal whatsoever – and by some miracle govern for 5 years with no confidence-type vote defeat. LEGITIMATE

    2.Lab/SNP have a formal deal, ditto for 5 years. ILLEGITIMATE

  13. OMNI

    “Re: legitimacy see above. The SNP aren’t using that language but they mean essentially the same thing. Legitimate parties don’t need to be locked out of Downing Street!”

    Not sure I buy that. It’s simply a florid way of saying ‘we’ll vote him out’.

    ““He said he was ruling out a deal with either the SNP or UKIP.

    The Lib Dems later clarified that the comments were made in the context of a formal coalition deal.””

    Well, if he sticks to that I’d agree. I’m partly over-reacting because I am partisan Labour and don’t want Clegg to lock EM out of Downing street, I guess.

  14. Constituency and national polls can both be right even when showing apparent discrepancies because they are snapshots of very different things.

  15. RayFromtheNorth

    True, there is very little to choose between the two except who scores a few points in the weeks after May 7.

    But given Labour are aiming for 1. at least Clegg isn’t totally gearing this against EM.

  16. MAGPIE

    Interestingly, the polls as they turned out in 2012 did actually underestimate the incumbent support by a few points – it turned out that of all the tens of polling companies, the one which was the most Democratic-leaning actually turned out to be the one that was closest to the true figure.

    So a CON lead of 4 points or so might not be completely out of the bounds of reality.

  17. I’m pretty sure that the door to 10 Downing Street famously doesn’t have a lock. Of course you can probably lock the security gates at the end of the street.

  18. Geoff,

    That’s true, which slightly undermines my point. Though I think it was partly the fact that they were all applying slightly spurious weightings to correct to what they thought would happen that partly caused that bias.

    I’m not sure one can easily translate that into UK politics though.

  19. The implicit argument that there are two classes of MP — one that can legitimately participate in the process of government formation, one that can not — sets a fairly dangerous constitutional precedent. Nigel Dodds identified that over the weekend.

    On the other hand, to bring this back to olling-related comment, there’s consistent support in E&W that essentially declares any SNP representation in Westminster second-class. That’s the context for a full-blown constitutional crisis.

  20. @ Magpie,

    I’m partly over-reacting because I am partisan Labour and don’t want Clegg to lock EM out of Downing street, I guess.

    On any set of seat totals in which the Lib Dems can possibly go in with the Tories, Clegg will try to go in with the Tories because he’s Nick Clegg.

    On any group of seat totals in which they can’t put together a majority, his opinions are irrelevant.

    We knew all this by the time the Coalition budget came out in 2010. None of it is news, nothing has changed except the probability of Clegg losing his seat. Don’t waste any energy worrying about it.

  21. MAGPIE
    I’m not sure one can easily translate that into UK politics though.

    Here’s a short interview with Nate Silver on the difference between the USA and UK. Though it was quite interesting.

  22. @ Catoswyn

    “LOL. I think you can get a bit excited….. why not indeed. Break open a bottle.”

    Would be nice but I rather over did it in my 20’s and i am down to about 4 possibly 5 brain cells, im saving one of those for election night and my prized bottle of Port.

  23. BigD – of course they can! Some seats drastically out perform or underperform the national swing. In 2010 the swing from Lab to Con was 5%, yet there was a swing of 14% in Cannock Chase. So yes, it is perfectly possible that Labour could be outperforming the national swing by a large amount in a single seat.

    Of course, the more pertinent point is not that Ashcroft is showing a decent swing to Labour in one seat, but that on average across ALL his constituency polling he’s showing a decent swing to Labour.

    LRR – I think it does have a lock on the inside, you just can’t open it from outside.

  24. I am grateful to those who replied to my earlier post about Ashcroft’s polls assuring me that they were totally above board and in no way slanted to the Tories. What a surprise it was that they were followed by a poll with a six point lead to the Conservatives.

    Is this really just a coincidence? Seriously what are the odds of it happening by chance?

    Perhaps I should admire the trusting nature of those who think that Ashcroft would publish polls that boosted the Labour party. I wish that I were so innocent.

  25. @ Holgate,

    There’s consistent support in Scotland for the dissolution of the union. There’s consistent support in Cornwall for Cornish devolution (albeit not very much support).

    If one group of malcontents with a constitutional hobbyhorse don’t constitute a “crisis”, I don’t see why any other group should. That the Prime Minister is a member of one of the groups is embarrassing and a good argument for replacing him with someone more sensible, but it’s not a national emergency.

  26. Thanks Anthony. I should have rephrased my point. As you say all Ashcroft constituency polls are showing a decent swing on average to Labour but not in the national poll.

  27. I am a Labour supporter and I prefer Yougov (it may have changed tonight) because of consistency, this does not dismiss Ashcroft poll.

    I find all this talk about the democrats polling nonsense and irrelevant.Based on what i see on the ground and the polls,thing are very tight both parties have everything to play for ,I have been to marginals (London) and they are behaving as you expect the Tories will lose some to Labour depends on how many?

  28. Too much guff on here, time to take a more realistic look.

    First of all, as pointed out by most, Ashcroft’s national and his marginals cannot both be right. Simply not possible that Labour is e.g. in contention in South Ribble, High Peak, ahead in Cannock, while the national picture is 36-30.

    Now look at the sample sizes. In each case around 1000. But that’s both for the national and each of the constituencies. I think that should give a pretty good idea of which is more likely to be accurate in general. Though not all the constituency polls will be accurate, again we need to look at them in aggregate.

    And taking off any kind of partisan hat, I find myself only really trusting Opinium, Populus and YouGov. Who happen to have the largest sample sizes in each of their polls, and poll regularly. I suspect like many YouGov is a bit too Labour, and Opinium possibly a bit tory. But they swing around much less than Lord Ashcroft polls. And frankly I just don’t trust most of the others.

    A final thought. I wonder if some of the forecasts such as Election Forecast are double counting shy tories. Like many here I expect the tories to lead by around 2 come polling day, and I think they are already a nose in front. But if ICM and various others are already building their polls to detect that, then the seat analysis could be a bit skewed.

    So still polldrums unless anyone says otherwise.

  29. KiIllary

    Ashcroft has posted polls that showed Labour leads. He posted two today. His constituency polls are consistent with a 4 or 5% swing from Con to Labour.

    However, if you are talking about national polls he showed a four point Labour lead on 22nd February at which time the consensus was neck and neck.

  30. Johnmo

    US election – yes I agree the analogy is flawed. The poll aggregators were all pointing to a clear Obama victory, some Repubs were saying all the polls were wrong because they were all over sampling democrats. They had I think Gallup to rely on as that was out on it’s own predicting Romney

  31. JOHN MO
    ‘things are very tight both parties have everything to play for…’

    Spot on.

  32. @bill patrick


    @little red rock
    “I’m pretty sure that the door to 10 Downing Street famously doesn’t have a lock. ”

    We spend ~£38bn on defence and 10 Downing Street doesn’t even have a lock?

    This is an outrage

  33. LRR

    ‘How big a difference is there if one compares the SVI in the Constituency polls with the National polls?’

    The CVI and SVI are just about the same in this round. So no clues there.

  34. Never mind I just saw Anthony’s post.

  35. @DAVE84
    You’re talking sense son.

  36. There seems to be more heat than light here today.

    Idea! Could we use some of that energy to agree how to decide when there is sufficient polling evidence to call a statistically significant lead for either the blue or the red side?

    I like to keep it simple (mainly because I’m not very good with stats). My hypothesis is that if Tories and Blues are equal pegging then the chances of one side showing a lead are:-

    One Poll- 50%
    Two Consecutive Polls – 25%
    Three Consecutive Polls – 12.5%
    Four Consecutive Polls – 6.25%
    Five Consecutive Polls – 3.125%


    1. If a particular polling company shows one side consistently leading (and other polls show switching leads) it could mean that a) that company has got it right or b) that company has a bias in its polling method

    2. If four consecutive polls from 4 separate polling companies show one side winning- then that’s quite likely to be significant. Less than that probably best to wait for confirmation from the next poll.

    3. If five consecutive polls from 5 separate polling companies show one side winning- it’s almost certainly significant.

    4. We also need to decide what would lead us to decide that any particular polling companies polls are showing a bias and therefore should be excluded from 2 & 3 above.

    I know my attempt is very amateurish but I would be interested to see how other people would go about answering these questions.

  37. @ Dave 84,

    I wouldn’t advise trusting Populus.

    This is unfortunate, as they are a tracker and I find their Labour leads aesthetically pleasing. But there are problems with their methodology.

  38. @Dave84 Now look at the sample sizes. In each case around 1000. But that’s both for the national and each of the constituencies. I think that should give a pretty good idea of which is more likely to be accurate in general.

    I asked about this recently and someone kindly explained that there isn’t much difference with two 1000 samples.

    There’s a difference between the 1000 and 2000 sample polls. And if you consider that with don’t know/ not 10/10 likely to vote taken out, some of these ‘1000’ polls are down to 500 before the final result is calculated.

    However apparently the size of the population doesn’t have as much impact as the size of the sample. So 1000 out of 60,000 is not statistically too different to 1000 out of 50,000,000.

  39. @Cloudspotter

    Thank You.

    I suppose we should expect SVI and CVI to converge as the election draws near.


    I was wrong. It has a lock.
    The key is under the flower pot next to the wheelie bin at the back.

    But you must NOT tell anyone. I promised Dave that only he, I and my North Korean Secret Service handler would know.

  40. Looks to me like the Big Two are still tied. Which means that leads of 6% in each direction are exactly within the margin of error. And we’ve had them in both directions. Which doesn’t suggest to me that “the polls are wrong”, but the opposite. Level pegging on 33% (as per the UKPR rolling average) should be showing up as polls ranging from Tories 30% Labour 36% to Tories 36% Labour 30%, with the majority of polls bunching around the 34/32, 33/33 and 32/34 area.

    That’s exactly what we’re seeing. Ergo, system working normally. Parties are tied. Polldrums.

  41. Has Smithy been on with his YG prediction yet?

  42. @rayfromthenorth

    You know how in the US they give out those “I voted!” stickers in polling stations after people vote?

    I think voters of Clegg-approved parties should get an “I’m legitimate” sticker. Or maybe “It’s my birthright”.

  43. Omnishambles

    I think you’re on to something there, quick get it registered before Paddy Power nicks the idea.

  44. Not yet – just pondering at the mo

  45. Am I the 1000th comment?

  46. @Smithy

    Is “pondering” code for “waiting to see if Tom Newton Dunn is gloating on twitter tonight”?


  47. I think one thing we can hopefully agree on is that some polling companies reputation will be damaged as a result of this election. They can’t all be right. This might seem a bit unfair given how complex and different this election is to any other.

    I think Populus reputation is probably most at stake. Most other pollsters have flirted between small Conservative leads and small Labour leads with the odd tie. Populus is the only poll that has consistently shown Labour leads this year (save for 4 ties).

    So if Labour win the popular vote they can claim to be the only pollster to reflect this throughout the campaign. If on the other hand, as expected, the Tories win the popular vote then they have got it spectacularly wrong.

  48. @Matt M – thanks. However still incline to believe Ashcroft constituency polls more than many other polls as they are frequent enough to draw conclusions from, and those are in line with the average of many others.

    So just YouGov vs Opinium as the likely winners we think?

  49. With ten days to voting day I am reminded that the last, and only time, I voted in an UK election was 1970. In that election Edward Heath won 330 seats with 52.4% support, Labour 288 with 43.1% and the Liberals 6 seats with 7.5%.

    This was also the first election that the SNP retained a seat won in a 1967 by-election, Bernadette Devlin’s Unity Party won 2 seats for the Nationalists in Northern Ireland (after a spectacular by-election win in 1969), Gerry Fitt won a seat for Republican Labour and 1 seat was won by the Protestant Unionist Party, forerunner of the DUP. There was also an Independent Labour win, but I am not sure from where?

    In this 2015 election SNP could be poised to wipe out both Scottish Conservative and LD representation, reducing Labour to a rump and denying them majority party status and possibly even minority government status.

    In England UKIP will be the third party by vote, but may not get any seats or as high as 8, but the impact of the UKIP vote will be to turn certain Tory safe seats into Tory-Labour marginals which may in and of itself deny Prime Minister Cameron the ability to return to number 10, unlike Edward Heath forty five years ago.

    As for the Green party who did not even exist in 1970, they are going to obtain around 5 times more votes than 2010, retain at least one seat and continue to increase their local government representation across England.

    As of this very moment Conservative and Labour are on 66.3% combined and have shrunk in support as the election has progressed, just as I predicted might happen, which compares with these two parties obtaining 95.5% of the vote on June 18th 1970.

    And if I listed some the seats that Labour and Liberal lost in 1970 the names would sound very familiar in terms of which marginals are up for grabs in this 2015 election.

    Finally the Liberal Democrats will go from third ranked party in the House of Commons to fourth, possibly fifth.

    And once again when one views the Lord Ashcroft constituency polls released earlier today Green lead the LD in two of four seats – so where is this 8%, or 9% or 10% that LD are supposed to get?

    Once again I believe weighting to 2010 results will make many pollsters come a cropper on election night when the votes start to be counted.

  50. Maclane

    Both sides have said the polls are rigged

1 18 19 20 21