ICM have released a new constituency poll of Sheffield Hallam, conducted for the Guardian. It shows Nick Clegg holding on to his seat by a margin of seven points over Labour when respondents are asked a voting intention question that includes the names of the candidates standing. Full details are here. As regular readers will recall, previous polling of the seat has shown a much tighter race with Lord Ashcroft’s last poll in Hallam showing Labour ahead by one.

So which poll is correct? Is Nick Clegg likely to hold his seat? The bottom line is while that this piece of evidence does make it look a little more likely that Clegg might hold on, we can’t really be confident what the true position is. The ICM poll had a sample size of 500, the Ashcroft poll had a sample size of 1000. Hence it could well be that there isn’t any difference at all between the polls, that it’s just normal sample variation around a small Lib Dem lead. Its also possible that there has been movement towards Clegg in the days between the two polls as the election looms and people consider a tactical vote.

However a lot has been made of the fact that while both polls had an effort to take account of people’s personal and tactical voting behaviour in their own constituency, they did so in different ways – Ashcroft asks a two stage question, asking people their national preference and then how they will vote thinking about the candidates and parties in their own constituency; ICM asked people the voting intention question including the names of the candidates standing in Sheffield Hallam. Both methods seem to have given a boost to the Lib Dems compared to a generic question, given sample variation and timing we can’t even be certain one did had more impact than the other, let alone which one is more accurate.

One can very easily make a case for one or the other method (Chris Hanretty has a good go here) but really that’s only theorising, we can’t know which way is better unless you test it against some actual elections, and at previous elections constituency polling has been a rare commodity.

In the meantime, Sheffield Hallam remains an interesting race. Normally the idea of party leaders losing seats is regularly drummed up but incredibly unlikely to happen. This time, while my personal expectation is that Clegg will hold on and this poll will probably end up about right, there is a least a non-zero possibility of him being ousted. We shall see.

880 Responses to “ICM poll of Sheffield Hallam”

1 15 16 17 18
  1. May2015 Election [email protected] · 3s3 seconds ago
    New Mirror/Survation poll (via @Jack_Blanchard_) says Labour +1:


    No change on last poll.

    Is this right that it means no change? Thanks. There is a tweet that says Cons are up 2.

  2. MITZ

    Yes, I promise to be gracious in defeat if the bookie’s predictions are wrong, not to mention being pleasantly surprised and relieved, but that is partisan talk not allowed on here :)

  3. ChrisL

    I think it depends on whether you compare to last Surv poll or last for that paper

  4. Nick Clegg’s gonna walk 500 miles
    And he would walk 500 more
    Just to be the man who walked a thousand miles
    To fall down at John Thurso’s door

  5. Survation Mirror Ballot Prompt-voting responses where seat specific ballot papers shown

    CON 31%
    LAB 32%
    LD 10%
    UKIP 15%
    GREEN 5%
    SNP 4%

  6. At this point I am reminded that the polls in the London Mayoral Elections in 2012 under-stated Labour support, by as much as 11 points. Phone polls performed the worst.

    When you also consider that the polls under-stated Labour performance in 2010, a pattern starts to emerge which is rooted in recent history rather than mythology about 1992…

  7. @ Cloudspotter

    Guardian came out for Labour, and I think they are more hard working than 5 years ago.

  8. @CL1945

    No change on the last Survation/Mirror poll. There was a Survation poll at the weekend for another newspaper.

    Best to regard it as Con+2 from an absurdly low 31 in their weekend poll.

    This continues to point to convergence around level pegging at the top end.

  9. I think this will be Bob on for Thursday nights result – maybe shave a point or 2 for UKIP into LD

  10. @cloudspotter

    The Grauniad plumped for Labour

  11. ChrisLane

    The last Survation had exactly the same figures for Lab and Con. So no change.

    Thanks very much.

    I had a thought! Maybe there are shy Labour people out there now due to the attacks, often close to being rude in the most visceral way, on Ed M, by high status people.

  13. Funny how nobody is saying the polls might be under stated for the Cons. lol. Site bias? :-)

  14. Survation have also done “constituency” and “ballot paper” questions. Very similar results.


  15. Survation/Mirror ballot prompt results:

    Con 31
    Lab 32
    LD 10
    Ukip 15
    Green 5
    SNP 4

    So ballot prompting only makes a 1% addition to the LD and Green scores. Tories and Lab fall by 2%.

  16. KeithP
    “try supporting the SDP – its tough advocating a party that no longer exists”
    Ahem. The SDP has two candidates standing in this election. One in Birmingham, one in Kingston-Upon-Hull.

    The leader of the party, Peter Johnson brags about previously being a left-wing UKIP candidate, in his campaign literature.

  17. @Rich

    Everyone is saying that!

    But I think it’s unlikely Tories are being understated in the polls as most polling organisations have gone to extraordinary efforts to filter for factors than may otherwise have understated them – especially Ashcroft and ICM.

  18. RAF

    Surely this is presumably a Survation for the Mirror as was the one last Friday so I would have thought we should be referencing this poll with that one rather than one for another paper at the weekend.
    Do you think there is now too much pollling? Its getting a bit silly now. If political anoraks like us can’t keep up what hope for the rest of the electorate?

  19. @rich
    “Funny how nobody is saying the polls might be under stated for the Cons. lol. Site bias? :-)”

    Actually a few of us are expecting swingback in the actual results, if not in the polls.

    Swingback being a swing towards the government parties, which has been observed in the past but is still a bit controversial. In 2010 we saw swingback towards Labour, who were the party of government at the time.

  20. Am not sure I really believe in floating voters making up that many people. Everybody I speak to seems to have made their mind up long ago.

  21. L Crosby’s McWedge strategy has not moved the McPolls in England and Wales a McInch.

  22. Rich

    Behave. I thought a recent poll had the Tories too low and UKIP too high and said so. Let’s just accept we all see things differently.

  23. @omni,

    Maybe that accounts for the bookies odds looking very generous on Lab. I must admit 4/1 seems very generous, but, as I said earlier, you don’t see poor bookies.


  24. @Mikey

    If the methodology is the same it’s a bit artificial for a pollster to say “only compare it with my last one is X newspaper”.

    It’s obvious the Con 31 in the weekend poll was random sample variation and this is closer to
    where the Tories are.

  25. @profhoward

    Yes it has moved the polls – at the regional level not the national level. What matters is winning the marginals and those are concentrated in certain regions.

  26. Catsowyn

    Not sure if my last attempt get moderated in time so I will try again , to get my prediction in. it may be slightly different but this is it :)

    I’m relying on the Labour ground game, a shortage of Liberal activists and a little bit of LiS resilience. Some shy UKIP voters and Galloway to get squeezed out, Farage to scrape home, Clegg survives with Tory help but resigns within a month, as do Cameron and Bennett

    CON 284
    LAB 274
    SNP 52
    LD 15
    UKIP 2
    PC 3
    Respect 0
    Greens 1

    DUP 8
    Sinn Fein 5
    SDLP 3
    Alliance 1
    Lady Hermon 1

    Speaker 1 (Bet I get that right)

  27. RAF

    Yes agreed. I thought that Con poll was too low. It had UKIP on 17 points which was too high.

  28. Vionn

    Regarding AW’s impartiality, I was rather disappointed in him last year when he announced he was “off to conference” for a certain party – when running such a committedly impartial site it’s probably best not to make your own political leanings so clear. Perhaps it was a slip of the tongue.

    Anthony’s personal politics are no secret and anyone with a minimal amount of google-fu could find them out very quickly. We tend not to mention it on here because of the occasional pleasure to be derived from various idiots attacking him on the grounds of bias and usually getting his politics completely wrong.

    Though I have to point out that anyone professionally involved in politics as a pollster (or an academic or whatever) is quite likely to be going to a Party Conference (or all of them) on work grounds.

  29. rich

    “Funny how nobody is saying the polls might be under stated for the Cons. lol. Site bias? :-)”

    There is nothing stopping you saying it, especially if you have reasons to think it is true.

  30. Omni: interesting. I see what he is doing now: using McWedge to generate a regionally-focussed swing back while keeping the national polls level, so as to induce a false sense of complacency among the opponent. Swing back and swing forward offsetting each other perfectly. Reducing Con vote where it is not needed, increasing it where it is needed.

    I take my hat off to L Crosby.

  31. Catsowyn
    Hopefully I’ll have a prediction for tomorrow morning. It’ll be, as the say on the interwebs, a real game changer. ;)

  32. As before, I think online polls may be too Laboury and too kippy. Just a theory, might be wrong.

  33. @Omni

    It really hasn’t

    E&W Con-Lab swing in Ashcroft’s poll earlier was 4.2%.
    In yesterday YG it was 4.6%.

    Ah but you cry, it’s lower in the marginals! But is it? Ashcroft’s last batch of marginals (excluding Battersea) showed the sw
    Ing at an average of 3.8%.

    Ah but you cry, the swing won’t be consistent across ALL marginals. No. But Ashcroft had lab taking Peterborough. Last week he had Labour taking Bristol West. The point is Labour IS doing well enough in the marginals.

    ComRes also showed Labour doing well enough in the marginals.

    Both Ashcroft and ComRes are phone polls
    whose house effects are hard on Labour.

    I’m not saying Labour will win most seats. But even Ashcroft and ComRes say it’s very possible.

  34. @profhoward

    Exactly. The polldrums aren’t real, Crosby wanted you to think they were real. Exhibit A:


    In fact I wouldn’t be surprised if your real name was Lynton Howard Crosby and you’ve been posting here observing “polldrums” to deceive your enemies. The dead cat was but a decoy. Bravo.

  35. So correct me if I’m wrong but 3 polls today of which:

    – 2 put the Tories ahead, by 1% and 2%
    – the third puts the Tories at jointly the most favourable position as they’ve come up with this year.


  36. Anthony always seems very neutral to me. I always think the site moderation is rather generous to Lab and quite strict on Cons, even worse on UKIP attacks where it is open season most of the time. lol.

    Funny how we all see different stuff. Anyway, I am past partisan now. If EM wins and forms a govt, even with SNP support, so be it, I hope he does a decent job, it’s not like anything we say on here moves the polls nationally…

  37. Rich the first thing you learn when come onto UKPR even as an independent is that any poll not showing a Labour landslide is an outlier, and unless the Conservatives win all 650 seats and 100% of the vote it’s a disaster for them and they are deemed to have failed.

  38. @raf

    But the polls have been moving in the marginals, whether they move enough is another story but the story is *not* polldrums. We’ve only recently started getting group marginal polls like the ComRes, but Ashcroft’s marginals show a change over time. Furthermore you can look at regional crossbreak averages over time and see it has changed

  39. Omni,

    Now look at the corresponding graph for London – goes the other way. Are two different regions really diverging that much or are crossbreaks unreliable?

  40. My latest UK-Elect forecast is here:
    May 5 UK-Elect detailed forecast

    It’s Lab 272 Con 271 SNP 55 LibDems 27 DUP 9 SF 5 PC 3 SDLP 3 UKIP 2 G 1 Ind 1 Speaker 1

    The final pre-election UK-Elect forecast is planned to be tomorrow afternoon.

  41. Vionn, Roger
    Lord Ashcroft also attended the Labour conference, in his role as a pollster. And you could hardly accuse him of being a Labour partisan.

  42. Maninthemiddle

    Someone saying what I think I’ve been seeing as well!!!!!

  43. I’ve just been on the Telegraph site. Their poll chart update has just thrown a wobbly and is showing Con at 29.4% and Lab at 37.3% (4.1% up in 1 day)

  44. @manin the middle,

    Yep, I have lost count of how many outlier Tory leads we have had, but funnily enough never seen a comment on a Lab one. Brilliant!


  45. @MANINTHEMIDDLE Rich the first thing you learn when come onto UKPR even as an independent is that any poll not showing a Labour landslide is an outlier, and unless the Conservatives win all 650 seats and 100% of the vote it’s a disaster for them and they are deemed to have failed.

    And that Tories are sensitive flowers ;)

  46. @hal

    Yes they are diverging because you can see Labour’s support increasing in London over the past 6 months from the pure London polls. The numbers are slightly understated in the crossbreak averages but the trend is the same.

  47. Rich

    Just a few days ago, one of the Survations – loads of reds on here says Cons, yes Cons, that would be the Cons too low

  48. @ ExileinYorks

    It must be a MoE, but a proportion might be true … Although I think it is neck and neck.

  49. I think a lot of people are placing a lot of emphasis on this “regional concentration of marginal seats” idea or as I like to call it the “Its the Midlands Stupid!” approach.

    I don’t know where this presumption that the Midlands held all the marginal seats came from because it simply isn’t true or at least not for this election. Across the Midlands I count 16 Tory held Lab target seats that I think Lab have a halfway plausible chance of winning.

    In the North of England though I count 22 such seats. Then throw in London which contains another 7 and the rest of the South (plus Wales) where there are another 25 and this argument that the Midlands matter more than elsewhere just seems off. As far as I can tell Milliband could bomb in the Midlands but he has the potential to largely make up for it elsewhere. Look at Cameron in Scotland back inn 2010.

    The Katie Hopkins effect perhaps?

1 15 16 17 18