An update on the Lord Ashcroft constituency polling from last week. Looking at the tables for Doncaster North – Ed Miliband’s seat – the weighting appears to be seriously askew, weighted as if the seat had a Conservative majority at the previous election. I’ve spoken to Ashcroft’s office and it appears to be a human error – typing the wrong weighting target in the wrong box (it’s something I’ve done myself in the past, though thankfully I’ve not done it and published the results!).

Lord Ashcroft’s team are reweighting and republishing the results, but obviously they are likely to make a significant difference to the findings, probably knocking down the reported level of Conservative and UKIP support and bumping up Labour. I’ll update when the new figures are out.

UPDATE: The corrected version of Lord Ashcroft’s polling is now up on his site here – topline figures in Doncaster North are now CON 13%, LAB 54%, LDEM 4%, UKIP 25%. UKIP are still in second place, but they are now a far larger twenty-nine points behind Labour (and given some of the reaction to the earlier poll it’s worth noting that Labour are even well ahead of the Tory and UKIP vote combined). There is no risk to Ed Miliband here.


123 Responses to “Update on the Lord Ashcroft Doncaster poll”

1 2 3
  1. I don’t think YouGov would make the same mistake as Ashcroft because PK seems to double check anything unusual or counter-intuitive before the poll is released.

  2. Presumably Ashcroft doesn’t collect and collate these figures personally, so he employs one of the usual polling companies? If so, shouldn’t any disapprobation be directed at that company rather than Lord Ashcroft himself?

  3. @spearmint

    “Does anyone understand why Peter Kellner thinks the incumbency effect…”

    I have no understanding of his thinking on the issue but one obvious reason for a greater but very skewed effect is balkanization as it doesn’t just cause ethnic enclaves it also creates a “school catchment colony” effect in constituencies as the most affluent concentrate in a few seats instead of being spread all over like they used to be.

    This will help Lib and Con in a *few* particularly leafy seats in or on the edge of the cities (although at the same time there will be a reverse effect in seats that have become or are on the road to becoming ex-Con).

    Hence the Lib wipeout predicted in the polls won’t happen everywhere and you could probably guess easy enough which seats it won’t happen in and which seats it will.

  4. COLIN

    The thing I find staggering in the Summary is that those few loyal souls who still declare a VI for LDs have amongst them, no less than 72% who might”end up voting differently” !

    I think there’s actually a number of things going on here. Firstly, and probably most important, they’re being stereotypically liberal and saying you might change your mind is part of that (even if you won’t).

    Secondly they are probably more likely to vote tactically than other groups and so may be waiting to see what the situation is nearer the time in their particular constituency.

    Thirdly many may still be unhappy with their party and its actions in coalition and so alternating between saying Lib Dem and DK (2010 Lib Dems still have a higher percentage of DKs). So even when they are saying they will support them again, there will be some doubt about it.

  5. ‘There is no risk to Ed Miliband here.’

    What a very great pity!

  6. Interesting latest odds at Ladbrokes (maybe they’re as good a place to look as any for VI, after all they are putting their money where their mouth is when it comes to predicting the outcome of the next general election).
    Cons most votes & Cons Most seats>evens
    Lab most votes & Lab Most seats>2/1
    Cons most votes & Lab Most seats>3/1
    Lab most votes & Cons Most seats>16/1

  7. Pete B –

    We don’t know who that was (for the avoidance of doubt, while Ashcroft won’t formally confirm it, it’s widely known that Populus do most of his regular polling.

    However, in this case I’m given to understand that the Doncaster North poll was NOT carried out by Populus (I don’t think the Sheffield Hallam or Thanet ones were either). No idea who did do it, but it wasn’t Populus.

  8. I have not had chance to look closely at Peter Kellner’s analysis, but I have not seen any mention of the polling that suggests that Labour is doing better in marginal seats. This would skew UNS in favour of Labour (up to a certain number of seats), and would be a counter-factual to incumbency bonus (since as marginals, they would often have first time MPs).

    I also remember from my PPE days (cue boos and hisses) that political scientists tend to be a bit sceptical about the idea of large personal votes for MPs. Unless an MP has a very high profile (which can cut both ways) I would expect the average voter to only have a vague awareness of the quality of a particular MP and vote for the party. I am pretty sceptical about the incumbency bonus idea, which seems to be one that requires the average voter to be as big an anorak as the people on politics forums.

  9. @ BERNARD SIMPSON
    Be careful Bernard, we cannot take the lead WITH Ed Miliband running Labour.

  10. AW
    “No idea who did do it, but it wasn’t Populus.”

    I see that it was telephone interviewing – does that narrow the field at all?

  11. Has anyone bothered to check the Ashcroft poll for South Thanet that showed UKIP in second place to the Conservatives, and Labour in 3rd?

    – A Concerned Resident :-)

  12. @ CROSSBAT11
    “I mean, just how many riders and provisos does it take before the whole argument becomes, to put it diplomatically, academic self-indulgence”?

    Crossbat, I agree with exactly that comment from PK that prompted the above. I must have written posts 500 times saying something similar. (With admittedly rather less knowledge). Anything, literally anything can happen between now and May next year. What the polls tell us is, at the present time, very many people don’t know what the hell to believe and are between some vain hope in Nigel Farage,(God help us), they dither between your party and mine.

  13. Pete B

    If so, shouldn’t any disapprobation be directed at that company rather than Lord Ashcroft himself?

    Well legally you’re responsible for your agents, but I don’t think anyone’s actually heaping much blame on him, quite the opposite for being so ready to admit he was wrong and so quickly. The original result was just at the bounds of plausibility and might have been due to local factors (Doncaster Council are a permanent feature of Private Eye’s Rotten Boroughs column) and the attacks in the media.

    Ashcroft was also quick to point out it was not just the VI that was affected:

    In a nutshell, Labour lead UKIP by 29 points in Doncaster North, not twelve; Miliband leads Cameron as best PM by 14 points, not one point; Miliband’s constituents would rather see him as PM than Cameron; they give him the highest ratings of the four main party leaders, not the third highest; and they trust Miliband and Balls more on the economy than Cameron and Osborne, not the reverse.

    It remains whether those in the Press who were so keen to publicise the original ‘facts’ will be as quick as his Lordship to make the same corrections.

  14. @ Mr. Jones,

    Is there any reason to think the 2010 Lab/Con marginals are particularly enclave-y, though? And in addition to the enclave effect you’ve got confounding variables like the BME voters moving out into the suburbs and whatever the heck is going on in Sheffield Hallam, which should be the quintessential example of the phenomenon you’re describing but is somehow at risk from Labour.

    It’s an interesting theory but I’m not convinced there’s been so much movement of that sort in the last four years- in contrast to the last two decades, say- to fundamentally alter the validity of the UNS calculations, especially since the boundaries haven’t changed.

  15. lurkinggherkin

    Has anyone bothered to check the Ashcroft poll for South Thanet that showed UKIP in second place to the Conservatives, and Labour in 3rd?

    – A Concerned Resident :-)

    The weighting was OK for Thanet and Hallam. That’s what convinced me that Doncaster was wrong, because their weighted figures were so close to 2010 and Doncaster so far out.

    I did have some concern about the reallocation of all DK and Refuseds to their original 2010 vote though. If you only reallocate 50% as is more usual you get:

    Sheffield Hallam

    Con 19%

    Lab 30%

    Lib Dem 29%

    UKIP 13%

    Green 10%

    Thanet South

    Con 33%

    Lab 26%

    Lib Dem 6%

    UKIP 32%

    Green 3%

  16. Roger Nexico.

    Interesting thoughts-but who knows the truth.?

    That is the lesson which Martyn so rightly says none of us is willing to accept.

    I get the impression that both Ashcroft & Kellner are beginning to accept it too. And if those two don’t know what is going on , how are we supposed to.

  17. @ Lurker,

    I am pretty sceptical about the incumbency bonus idea

    Fortunately the incumbency bonus is measurable, so we don’t have to treat it as a matter of belief: it is definitely a real thing that has existed in past elections.

    The questions we have to answer for the next election are:

    1. Have the 2010 intake behaved differently from previous cohorts of MPs in a way that means they won’t benefit from incumbency?
    2. Has the electorate changed so much since 2010 that incumbency no longer benefits politicians?

    On the general principle that the future tends to look a lot like the past and interesting things don’t happen often, my own view is that neither is likely.

    Frankly the 2010 Tories do seem a little flaky as a group- an awful lot of them are standing down after a single term, and one of them was Louise Mensch- and Ashcroft’s marginal polls do suggest that most of them are not enjoying an incumbency bonus. But Ashcroft’s marginal polls don’t name the candidates, and it also seems likely that an intake elected on the back of the expenses scandal and particularly resistant to control from CCHQ would be especially good at constituency services. So barring further evidence, I’m going to say it’s a wash.

    And if 2010 wasn’t a rabidly anti-incumbent election I can’t see any reason why 2015 would be, even allowing for the general anti-politics mood.

  18. One place where there might not be an incumbency bonus is where a new MP is being challenged by the previous incumbent. I’m not sure how many seats that applies to, but it does happen.

  19. @Spearmint

    I seem to remember that the Ashcroft polls were showing an incumbency disadvantage for the new Tory intake, also discussed here

    http://lordashcroftpolls.com/2014/09/liberal-democrat-battleground/#more-6424

    “So what is going on? Much of the discussion about this subject seems to assume that the incumbency effect is something that just happens to a politician. But it is not – it is something he or she creates (or doesn’t). Incumbency is not so much an advantage in itself as an opportunity to build a profile, make a reputation, and achieve the things that will ultimately be rewarded on election day. I suspect new MPs, knowing they will have a battle to hold on to their often marginal seats, have in the past worked harder than most, thereby creating what has become known as the “first-time” effect. I further suspect that some have registered the existence of the phenomenon without having understood the reasons for it, and now think it will be bestowed upon them gratis by a benevolent electorate. The MPs who enjoy the biggest boost from incumbency will be the ones who earn it.”

    So it depends if they heeded that warning and started to earn it since that warning.

  20. And it looks like someone looked at the latest marginal polls and found the same thing – no incumbency bonus

    http://may2015.com/featured/there-is-no-first-time-incumbency-effect-for-tory-mps/

    I don’t think the Tories trained their new intake on how to be a good MP. Certainly the case in my constituency where we have one of the new intake, and the previous one who had to resign over his expenses was much better.

  21. My own view is that 1st time incumbency has been massively overhyped …. it certainly was the case in 2001 when half the labour mps didn’t need to be around Westminster because of labour’s huge majorities….there is clear evidence that incumbency for the first time did help labour retain seats… I have always had my doubts about it, certainly as a key determinant in enabling the tories to cling on.

    more interestingly, i know it’s fashionable to pooh pooh electoral calculus and say that “of course uns is crude and unapplicable and useless”, but I think the electoral calculus website does tell us something, even if it isn’t very sophisticated. In terms of direction, the baxter website is telling us that there has been a noticeable strengthening of the labour position in november after a torrid october/early november for mili…. baxter has labour 10 short of a majority which is more generous to the reds than the betting
    markets…. time is running out for the blues. obviously fisher and electionforecast with their swingback models are much less generous to labour.

    http://www.electoralcalculus.co.uk/homepage.html

  22. A good example of ‘incumbency bonus’ (always assuming that it exists!) is the LD prospects in Scotland. I cannot imagine a situation, for example, where Charlie Kennedy loses his seat, even though the Scottish cross-breaks regularly show his seat to be a prospective SNP gain. Michael Moore may not be in quite such a strong position in the Borders, but even there I’m pretty sure that the LDs will hold on in what is now a three way possibility – LD/Con/SNP depending on the %ages.

    The crossbreaks from Ashcroft give Labour only four seats in Scotland – and one of them is Kirkcaldy and Cowdenbeath, which, as we now know, is to be vacated this coming May by its present incumbent. Again, I’m not convinced that Labour will fall to anything like those levels – though I may be wrong.

    Personally I believe that there is an incumbency effect – where an MP makes him/herself seen or heard on a regular basis. This used to be easier for the old Liberals, as virtually each one had to take some sort of ‘shadow’ post and was therefore seen and heard regularly (albeit only after the ‘official opposition’ had had their say).

    If an MP is doing his/her job well, then after a few years most people in the constituency will know someone who, or some organisation which, has been helped by said MP. At least, that’s been my experience over the years in several different constituencies.

  23. Roger Mexico

    Thanks for that. The Ashcroft figures for TS didn’t look wildly unexpected but the Doncaster glitch did make me wonder….

  24. @spearmint

    “Is there any reason to think the 2010 Lab/Con marginals are particularly enclave-y, though?”

    Lab-Con no, transitional – either becoming more Lab or more Ukip

    “And in addition to the enclave effect you’ve got confounding variables like the BME voters moving out into the suburbs”

    Yes, lots of confounding variables.

    “and whatever the heck is going on in Sheffield Hallam, which should be the quintessential example of the phenomenon you’re describing but is somehow at risk from Labour.”

    Yes, generally red liberals escaping to more affluent areas are outnumbered by green liberals already there so they tactically vote Lib. and despite them being annoyed over the coalition I’ve always thought they will do so again in 2015 despite what they tell the polls now.

    So either Hallam is proving my theory wrong or the proportion of red to green oop north is getting to the point where they might risk voting Lab instead. I think there’d need to be a very clear lead for them to risk it though.

    “I’m not convinced there’s been so much movement of that sort in the last four years- in contrast to the last two decades”

    Agreed. It’s been a gradual effect over the last 20 years or so. I think what’s happening now is the effect on UNS has become too obvious to ignore so people who don’t want to admit to themselves that the country is balkanizing will be coming up with all sorts of other explanations for why the voting is going the why it is.

  25. sorry i should have stressed “1st term incumbency” that’s what people are talking about since 147 of 306 tory mps are new intake tories from 2010. the argument is that their 1st term incumbency will mean that the tories will cling on even if they seem, polling wise, to be stuffed.

    of course, Charlie Kennedy, who’s been an MP since 1983 has a massive incumbency bonus.

  26. I said at the time of the Reinhardt & Rogoff spreadsheet disaster that, as an engineer, I’m shocked at data leaving anyone’s office without being properly checked for obvious mistakes first.

    If we make mistakes as bad as the one that Ashcroft has made, people die or people lose a hell of a lot of money. In Ashcroft’s case, his mistake has driven no small part of the news agenda this week. I assume that there will be equally big headlines to the effect that the numbers had not had so much as an elementary check before they went out, and that Miliband is in fact on course for a thumping win with an increased majority.

  27. @Spearmint: “So barring further evidence, I’m going to say it’s a wash.”

    I haven’t come across this expression before. Can you (or anyone) enlighten me?

  28. Anthony Wells

    However, in this case I’m given to understand that the Doncaster North poll was NOT carried out by Populus (I don’t think the Sheffield Hallam or Thanet ones were either). No idea who did do it, but it wasn’t Populus.

    Well that explains the slightly different layout that those three polls had and why there was a separate report for them. It does mean we don’t get to checkup on future reallocations though. You can’t quite work it out from the figures but I wonder if Ashcroft normally uses the old Populus formula of 50% for Lab and Con, 30% for LD. But not the 100% that whoever did these ones used. I suspect they may be a more consumer orientated pollster and misunderstood what was wanted.

  29. @Spearmint er al.

    What puzzles me about the incumbency effect is not whether but why. The statistics seem to be plain enough to say that first time incumbents are associated with lower swing, but association isn’t the same as causation. Everyone has a theory (new MPs try harder, local parties who’ve won gain supporters, some voters like to give a new MP a second chance etc) but I’ve not heard of any actual research on the causes of the effect.

    I’m of the mind to agree that it will probably continue, but there’s always the niggling doubt that the true cause of it might be something we’re not watching, and could be surprised by.

  30. @ John Kay,

    It means two opposing factors have canceled out, leaving you back where you started. I think the etymology must come from “washed out”- ie. everything gets swept away and the slate is wiped clean. Probably an Americanism.

    @ Postageincluded,

    Good points.

    If you had data on how frequently MPs held surgeries you might be able to use that as a proxy for “trying harder” with regard to constituency services and test that hypothesis. It’d be hard to collect the data though; I bet none of them want to find themselves at the bottom of that league table!

  31. Spearmint

    In addition to the points that Richard made about the latest batch of Tory MPs[1], there another couple of things to point out. The first is that the big incumbency boost after a first term is usually a double one. Not only does a new hard-working MP get credit for it, but there is the unwinding of the previous incumbent’s personal bonus to come off the opposition vote.

    But the expenses scandal meant that many ex-MPs will have lost theirs before the election and so the Tories will already have had the benefit. And some of the better ex-MPs may actually be standing again – and still have a positive bonus[2].

    Also MPs in this Parliament haven’t had the chance to build up the personal vote. On a previous thread someone pointed out that the 1997 intake part of a massive majority, were encouraged to spend as much time in marginal constituencies as they could, building up the vote. With no majority the government has to keep its backbenchers in Westminster for the votes.

    There will be some who buck the trend of course, the hard workers and local heroes like Halfon. But there seems quite a lot to suggest that these will not be enough to make a difference.

    [1] Even with Mensch gone, they do seem to be a rather entitled-feeling bunch. All that talk of A-lists can’t have helped, or endeared them to many activists.

    [2] The classic case is Stroud where David Drew is standing again. Ashcroft’s constituency question showed an 11 point lead, up from 1 point for the general one.

  32. I wonder whether the absence of ‘negative incumbency expenses effect’ might provide some counterbalance.
    In my constituency there were certainly people who refused to vote for ‘Mrs Expenses’ even though that would be their preference in normal circumstances. Some at least of these will revert.
    I’m sure my constituency is far from unique.

  33. MrJones, Spearmint,

    Sheffield Hallam is not my seat, but I live a few hundred yards from the boundaries, campaign there often, and can offer some insight.

    Really, it’s a seat unlike any other. A very large constituency, it covers the relatively student-filled Crookes ward, the upper-working/lower-middle class suburb of Stannington, the leafy suburbs of Ecclesall and Fulwood and the Tory-inclined and retiree-inhabited villages of Dore and Totley, plus a whole load of the Peak District.

    In any other situation but this election, a seat as visually rich and middle class as Hallam would be Tory without a second thought. But the grandeur belies the reality. A huge public sector workforce lives in Hallam – they are doctors, consultants and managers at the city’s several hospitals, teachers at the city’s schools or lecturers and administrative staff at the two universities.

    They largely consider themselves to be left of center, albeit fairly moderately, and are often unhappy at what they consider the coalition to have done with their places of work. They would enthusiastically have both rejected New Labour and endorsed the left-of-Labour promise of the Lib Dems in 2005 and 2010. They made up a huge chunk of Nick Clegg’s voters, and have largely left him behind.

    There are also 12,000 students, largely concentrated in Crookes’ private rented housing and the University-owned student villages in Fulwood. A smaller contingent of a few hundred lives off Ecclesall Road South, near the Greystones pub. They are not by inclination Labour voters – those who were students here in 2010, albeit now replaced – broke strongly for Clegg. But they are generally left wing and strongly dislike Clegg for reasons that are less to do with tuition fees per se and more to do with general feelings of dishonesty and incompetence.

    This leaves Nick Clegg in the curious position of having a large chunk of his 2010 vote disappear, while a number of 2010 Tory supporters (who didn’t want that left-of-Labour bloke) and non-voting right wingers who saw his election as inevitable have come tactically to his aid for this election. It’s resulted in a fascinating flip of messaging from the Lib Dems – their leaflets, which once spoke of keeping the Tories out, now proclaim “It’s Lib Dem or Labour!” with bar charts aflutter.

    Also to Labour’s credit and advantage is the personal nature of the campaign and their choice of candidate. Nick Clegg is of course very well-known as the Deputy Prime Minister, but has few roots in Sheffield and lives in Putney. Labour’s candidate Oliver Coppard was brought up in the seat and lives and works in Sheffield, and it’s as much his campaign as it is Labour’s.

    The low base in Hallam and lack of central oversight given its non-target status has meant that Labour have had to be inventive with their resources to grab attention. Coppard’s literature and posters resemble classic Labour material only in the small logo in the bottom right. The slogans are not One Nation or Together, but Better for Sheffield. The Neo Sans font is nowhere to be seen, which gives everything a more personal feel than standard Labour boilerplate. The leaflets are rigid and matte, again different to the easily-discarded flimsy glossies that come through doors from all sides in more traditional marginals.

    Canvassers call on behalf of Oliver Coppard, the Labour candidate, not the other way round, and wear t-shirts bearing his name more prominently than the party. Canvassing is regular and spread all over the seat. It feels, when compared to other campaigns in which I’ve participated, a lot like a by-election which is very removed from normal politics.

    The Tories in Hallam, once dominant, have been forced back into the hinterland of third place by boundary changes and demographic shift over the years. They lost their last councillor in the seat in 2008, and since then have regained no ground. Somewhat ironically given my earlier statement, students are actually a relatively healthy Hallam demographic for them – they mostly come from further afield than Sheffield, and, being largely middle class, contain a fair few Tories.

    Not, of course, that they are anything close to winning in the student areas. There is only one ward in Sheffield Hallam where the Tories still come second – Dore and Totley – while in Crookes and Stannington they no longer even stand. The bulk of their voters have either died, moved away or switched to the Lib Dems over the years, with little prospect of a return.

    Now, the Others. Sheffield Hallam is one of the least friendly seats for UKIP in South Yorkshire. Between its middle-class Labour supporters, the Lib Dems, largely liberal or tribal Tories, Greens and the student population, there’s nowhere they naturally appeal to any degree except in Stannington, where they placed a decent third in May.

    The Greens enjoy some appeal among students, but their vote is generally quite soft. Usually they can be relied upon to come round to Labour when reminded of the two-horse race nature of the seat and of the opportunity to kick Clegg. Their candidate is a former SDP man and teacher – largely lacking in green credentials when compared to Coppard.

    As far as I’m aware, no other candidates have been announced as standing. Apologies for the long post, but the seat is fascinating to watch and will no doubt be a very exciting place to be during the next five months. Keep an eye on here for updates as we progress.

    TL;DR: Sheffield Hallam is hard to compare to other seats, but it’s interesting in its own weird way.

  34. Looking at election forecast.co.uk it seems their prediction of a hung parliament has been over 90% for 3 consecutive days now, I’d say that was as close to “calling it” as you can get this far out. If they are right, it’d be quite a feather in their cap from a modelling point of view.

    At the moment it’s hard to tell if they are out but the way the lib dems VI prediction is pretty much dropping linearly leads me to suspect they won’t recover in line with historical trends and this drop will happen all the way to the election as the overestimate decreases with remaining time left to enact a recovery.

    It’s one of the large areas of uncertainty and may well not follow historic trends (or at least become a new set of data which will have a significant effect on the model post 2015). Other areas will be SNP, greens and UKIP, how well do minor parties (not sure I want to call the SNP a minor party any more, if the polls are right, they will have a completely new role in politics in future, and they might have very different characteristics as a localised party) who see a massive increases in support hold onto this support 6 months out from an election? Historically we have little precedent so a historic data based model may fall flat on it’s face (until we actually see relevant data after which I’d expect the model to perform better).

    The trouble with historic based models is there is a vast amount of data with only a handful of elections to make predictions based on the data. I’d expect these models to perform a lot better in 30 years time after we get a better idea how minor parties perform and the relative likelihoods of swing back vs swing away. Right now I think the chances of stuff happening that is historically unprecedented is too high to hang a hat on these types of models.

    Where they might be of use is in making better than UNS predictions. The other half of the model tries to look at Ashcroft type localised data and make seat predictions from that. How SNP target seats in particular perform might be far away from UNS predictions and I hope someone at least can capture than information beforehand.

    I’m sure all the parties will have their own models of where their battleground seats lie, who can commit resources most effectively come the election will be important, although I doubt more than a handful of seats can be swung from a loss to a win by a huge localised campaign in a tight election winning a dozen of these as a bonus could make a huge difference in the makeup of the next government.

  35. Postageincluded – the evidence of the incumbency effect comes from two factors. One that parties that had an incumbent MP in a seat when they did not have one at the previous election get a bigger swing, the other than parties that do not have incumbency at an election when they did at the previous election (because the MP is retiring, or was defeated) have a bigger swing against them*

    That gives us a steer as to causes. It isn’t something special about new MPs or parties who have won the seat, its not unique to new MPs at all – it’s the transition from the state of having incumbency to not have incumbency, or vice-versa. So it’s not that second or third term MPs don’t have as much incumbency effect as a first term MP… it’s just they already had it at the last election as well, so it’s already factored into their vote.

    All this suggests its something to do with the fact of being a local MP. We can’t tell, but I expect it’s from being a recognised name, being in the local paper talking about local stuff, turning up at local community events, dealing with people’s constituency problems and all that stuff.

    (*The thing that is special about first term MPs is when an MP gains a seat from sitting MP of another party… so one party gains an incumbency effect and another loses it, so it looks particularly large. Obviously it doesn’t apply to all first term MPs)

  36. *shudder*

    There but for the grace of god go we all.

  37. Bum – just wrote something and then my internet connection went down. Can’t be bothered to write it again, suffice to say there is a nice paper in Electoral Studies that runs the numbers for anyone with access to academic journals

    Are you sitting comfortably? Estimating incumbency
    advantage in the UK: 1983–2010 – A research note
    Timothy Hallam Smith

    Bottom line is average Conservative incumbency effect is about 1%, but has increased to about 1.6% at recent elections. Labour average is about 1.8%, Lib Dems about 8.5%(!).

    Obviously these are small for Con & Lab parties, so don’t overestimate the impact of incumbency… it’s a small factor that will tilt the scales a little, it wont transform anything. Nevertheless, at any election there will be a couple of handfuls of seats tight enough for it to make a difference, and at an election with lots of 1st term incumbents in battleground seats that may have a modest impact.

  38. mrnameless

    Sounds like the balance of green liberal to red liberal is possibly quite close then.

  39. Mr Wells says — ‘ However, in this case I’m given to understand that the Doncaster North poll was NOT carried out by Populus (I don’t think the Sheffield Hallam or Thanet ones were either). No idea who did do it, but it wasn’t Populus. ‘

    I find this amazing.
    First that Ashcroft does not clearly say who does his polling (despite protestations of transparency)
    and
    Second that he varies his pollsters yet we are supposed to compare his polls.

    Surely the pollster should do everything including the opportunity to ‘transpose’ and not leave it to the commissioner?
    Surely the commissioner should just commission and not be responsible for doing any adjusting for past this that or the other or if someone reads the Sun?
    Why on these clearly sensitive polls does the pollster not only change but change to someone who is totally unknown?

  40. It was a mistake by Ashcroft – that’s all. It would be a real shame if this incident causes him to reconsider spending his own cash on some quite invaluable data.

  41. HookesLaw – it’s a strange state of affairs, as he sort of straddles the boundary between being a client, and being a white label pollster.

    Ashcroft doesn’t actually conduct the polling in his name, he just pays the money to commission other people to do it for him… like a client.

    However he also acts like a white label pollster, putting it out under his own brand and – I assume – making decisions about weighting and methodology used, but getting someone else to actually do the heavy lifting. A pollster outsourcing fieldwork to another company in this way is nothing unusual at all – pollsters will often pay an external companies fieldforce to actually make the telephone calls, or buy access to a panel provider’s respondents to get the sample. So ComRes and Populus used to often use ICM’s call centre (I don’t think Populus do these days, I’ve no idea about ComRes). Survation certainly used to use Toluna’s panel for their surveys, again, no idea if they do now. This is all normal… however, it isn’t very transparent.

    As a more general comment, transparency rules these days mean the weighting and reallocations used by pollsters are extremely clear and open. Sampling – how, who and so on, tends to be more opaque…

  42. Mr Nameless

    Did you ever read the comments policy?

  43. @ Mr. Nameless,

    Thanks for the overview. I’m staggered by the scale of the Tory collapse in a seat which they held solidly in every election until 1997 and in which they won more than twice the Labour vote share in every election since- after the coalition was formed I thought Clegg might be vulnerable to a Tory challenge, but until Oakeshott’s constituency polling/trolling came out I would never have thought in a million years he might be vulnerable to a Labour one.

    Socialist Republic of South Yorkshire indeed!

    @ Anthony,

    Thanks as ever for cutting through our psophology with data. Has anyone looked at Roger’s pseudo-incumbency-bonus-to-former-MPs theory?

    @ HookesLaw,

    Because of capitalism? Lord Ashcroft is paying for them; he can release or withhold any information he likes. If his new pollster isn’t a member of the British Polling Council he doesn’t even have to release the tables.

  44. @ Spearmint

    Isn’t it the time of month for your graphs what with it being the 1st? These ones should be quite interesting.

  45. @ Shev II,

    It is, it is, but I’m procrastinating copying in all the data by talking to you lot. ;)

    They should appear mid-week sometime.

  46. Spearmint

    It’s hard for outsiders to understand how reviled the Tories are in South Yorkshire. Hallam was traditional Tory ground. 15 years ago, it was reported as being in the top ten constituencies in the country by average income. It’s a beautiful area and anyone with money in this region gravitates there.

    But even some dyed-in-the-wool Tories there were shocked by what happened to South Yorkshire in the 1980s. 50,000 steel jobs went in Sheffield in 10 years. Unemployment in mining villages in the Sheffield hinterland hit 25%. You can’t live cheek-by-jowl with that and not have your world-view affected by it.

    Plus, the last Hallam Tory MP, Irvine Patnick gained notoriety. He was one if the old “hang em and flog em” brigade and that never say easily with the majority of Hallam Tories. After Hillsborough Patnick acted as a go-between for SY Police to get stories into the tabloids accusing Liverpool fans of bestial behaviour. Put the lot together and it’s little wonder that the seat has swung violently away from the Tories over the past couple of generations. The Tories won 71% of the vote in 1951 and 55% as recently as 1979. Since then it’s been a constant slide, to under 25% in 2010. Wouldn’t surprise me to see them slip under 20% in May.

    And folk say Labour has a problem in not having a voice in certain areas of the country…

  47. @Lefty

    I second that.

    My clan are from South Yorkshire, and the toxicity of the Conservatives is as you describe .

    Labour may be complacent. but I suspect no other party has the on-the-ground troops and machinery to take them on seriously.

  48. There’s a Comres phone poll at 10pm.

  49. Chevin Guy,

    Obviously my experiences colour my impression of the seat, but they were wondering “What’s going on in Hallam” and I think I gave a relatively objective account. The Lib Dems and Tories are doing badly and Labour are doing well.

    If anyone has a view of the seat which contradicts mine, I’d be happy to get a more rounded perspective but by necessity I have a limited view.

  50. Bear in mind people that phone polls have tended to show lower numbers for Con and Lab and higher numbers for Ukip in particular. If there has been a dip in Ukip support post R&S this may be the poll to show it.

1 2 3