Survation have a new poll out in Sheffield Hallam which gives a ten point lead to Labour. Naturally this has produced a lot of crowing from people who don’t much like Nick Clegg and some possibly unwise comments from Nick Clegg about the poll being “bilge”, commissioned by the Labour affiliated Unite (which is was, but it shouldn’t make any difference to the voting intention figures). Tabs are here.

The poll has been compared to Lord Ashcroft’s one last year which showed Nick Clegg ahead in his seat, albeit, only narrowly. The reason for the difference is nothing at all to do with who commissioned the polls though, and everything to do with differences between the methodology Ashcroft uses and the methodology Survation use for all their clients (Unite, and anyone else).

One difference that people commented on yesteday is that Lord Ashcroft uses political weighting in his constituency polls, but Survation do not. This has the potential to make a sizeable difference in the results, but I don’t think it is the case here – looking at the recalled vote in Survation’s poll it looks fairly close to what actually happened, weighting by past vote would probably have bumped up the Lib Dems a little, but the reason the Lib Dems are so far behind is not because of the weighting, it’s because more than half of the people who voted Lib Dem in 2010 aren’t currently planning on doing so again.

However, there are other methodology differences that probably do explain the gap between the Ashcroft poll and the Survation one. If we start off with the basic figures each company found we get this:

In Survation’s poll the basic figures, weighted by likelihood to vote, were CON 22, LAB 33, LD 23, UKIP 9
In Ashcroft’s poll the basic figures, weighted for likelihood to vote, were CON 23, LAB 33, LD 17, UKIP 14

Both had a chunky Labour lead, in fact, Ashcroft’s was slightly bigger than Survation’s. Ashcroft however did two things that Survation did not do. He asked a two stage question, asking people their general voting intention and then their constituency question, and he reallocated don’t knows.

When Lord Ashcroft does constituency polls he asks a standard voting intention question, then asks people to think about their own constituency. This makes a minimal difference in most seats, where people’s “real” support is normally the same as how they actually vote. In seats with Lib Dem MPs it often makes a massive difference, presumably because tactical voting and incumbency are so much more important for Lib Dem MPs than those from any other party.

This is a large part of the difference between Survation and Ashcroft. In Ashcroft’s second question, asking people to think about their own constituency, he found figures of CON 18%, LAB 32%, LD 26%, UKIP 14% – so the two-stage-constituency-question added 9 percentage points to the Lib Dems. Survation actually asked people to think about their constiuency in their question, probably explaining why they had the Lib Dems 6 points higher than Ashcroft in their first question, but I think the constituency prompt has more effect when it is asked as a second question, and respondents are given a chance to register their “national choice” first.

The other significant methodological difference is how Survation and Ashcroft treat people who say don’t know. In their local constituency polls Survation just ignore don’t knows, while Ashcroft reallocates them based on how they voted at the previous election, reallocating a proportion of them back to the party they previously voted for. Currently this helps the Liberal Democrats (something we also see in ICM’s national polls), as there a lot of former Lib Dems out there telling pollsters they don’t know how they will vote.

In this particular case the reallocation of don’t knows changed Ashcroft’s final figures to CON 19, LAB 28, LD 31, UKIP 11, pushing the Lib Dems up into a narrow first place. Technically I think there was an error in Ashcroft’s table – they seem to have reallocated all don’t knows, rather than the proportion they normally do. Done correctly the Lib Dems and Labour would probably have been closer together, or Labour a smidgin ahead, but the fact remains that Ashcroft’s method produces a tight race, Survation’s a healthy looking Labour lead.

So which one is right?

The short answer is we don’t know for sure.

Personally I have confidence in the two-stage constituency question. It’s something I originally used in marginal polling for PoliticsHome back in 2008 and 2009, to address the problem that any polling of Lib Dem seats always seems to show a big jump for Labour and a collapse for the Lib Dems. This would look completely normal these days of course, but you used to find the same thing in polls when Labour were doing badly nationally and the Lib Dems well. My theory was that when people were asked about their voting intention they did not factor in any tactical decisions they might actually make – that is, if you were a Labour supporter in a LD-v-Con seat you might tell a pollster you’d vote Labour because they were the party you really supported, but actually vote Lib Dem as a tactical anti-Tory vote. The way that it only has a significant effect in Lib Dem seats has always given me some confidence it is working, and people aren’t just feeling obliged to give as different answer – the overwhelming majority of people answer the same to both questions.

However the fact is the two-stage-constituency question is only theoretical – it hasn’t been well tested. Going back to it’s original use for the PoliticsHome marginal poll back in 2009, polling in Lib Dem seats using the normal question found vote shares of CON 41, LAB 17, LDEM 28. Using the locally prompted second question the figures became CON 37, LAB 12, LDEM 38. In really those seats ended up voting CON 39, LAB 9, LDEM 45. Clearly in that sense the prompted question gave a better steer to how well the Lib Dems were doing in their marginals… but the caveats are very heavy (it was 9 months before the election, so people could just have change their minds, and it’s only one data point anyway.) I trust the constituency prompted figures more, but that’s a personal opinion, the evidence isn’t there for us to be sure.

As to the reallocation of don’t knows, I’ve always said it is more a philosophical decision that a right or wrong one. Should pollsters only report how respondents say they would vote in an election tomorrow, or should they try and measure how they think people actually would vote in an election tomorrow? Is it better to only include those people who give an opinion, even if you know that those undecideds you’re ignoring appear more likely to favour one party than other, or is it better to make some educated guesses about how those don’t knows might split based on past behaviour?

Bottom line, if you ask people in Sheffield Hallam how they would vote in a general election tomorrow, Labour have a lead, varying in size depending on how you ask. However, there are lots of people who voted for Nick Clegg in 2010 who currently tell pollsters they don’t know how they would vote, and if a decent proportion of those people in fact end up backing Nick Clegg (as Ashcroft’s polling assumes they will) the race would be much closer.


287 Responses to “Polling in Sheffield Hallam”

1 2 3 4 5 6
  1. @Hawthorn

    I think there isn’t much chance (on the current figures) of any joint policy plans that would be acceptable to parties in a coalition.

    If I’m reading the mood music right then the current coalition has left large parts of both parties deeply unhappy, and so it goes across the spectrum.

    That is why I suspect we will see C&S and another election, probably with several new party leaders, later in the year.

    Clegg – possibly not even an MP
    Miliband – maybe hanging on, but only if he leads a C&S government
    Cameron – two bites of the cherry without a clear win? Again, only survives if he’s leading the government…

  2. ROSIEANDDAISIE

    Yes, a cherry picked result, and I’m not reading any trends into it, but still an interesting observation

  3. Cant see natalie surviving the debates either.

  4. RAF “Obviously bring on 33 means 2/3 of people still prefer someone else”
    Not ‘people’. Voters (or ‘weighted voters’, at least)
    Not ‘someone’ but ‘some other party’.
    Not ‘bring’ but ‘being’.
    Do have a care.

  5. Not surprisingly, most of the discussion about this week’s Ashcroft constituency polls has been about their implications for the election and for coalition formation.

    However there is another little matter for those interested in using polling data to project potential seat tallies in May. This is what the Ashcroft data tell us about the quality and performance of the various seat-projection models. About a fortnight ago in comments posted at 4.37 pm on Jan 16 at 4.37pm and at 3.21pm on January 19 I suggested an approach for quantifying this. The Electionforecast (EF) eam had tackled this by plotting their projected VI against those published in Ashcroft’s November and December batches of constituency polls. I pointed out that corresponding graphs for most other models would look broadly similar and suggested instead using a quantitative measure of accuracy based on calculating the Euclidean Distance [1] between the projection and the poll.

    With kind acknowledgements that the original suggestion came from UKPR, the EF team have now adopted a similar approach to assessing their performance in the most recent batch of Ashcroft polls:

    http://blogs.lse.ac.uk/generalelection/predicting-the-polls-scotland/

    I have also repeated the exercise (with outcomes a little different from theirs). In addition I have extended the analysis to other models making it possible to compare the accuracy of each in turn.

    Summary (for those who don’t want to get bogged down in technical detail): My analyses suggest that the VI-reallocation algorithms being used by May2015 seem to be performing rather worse than those for Electionforecast and Electoral Calculus (EC). Moreover EC is almost certainly hampered by failing to take full account of constituency polling data. So, Electionforecast probably has the best-informed database for use as a starting point for making projections

    Background: In my January posts, I calculated that – using my accuracy measure (average batch Euclidean Distance (ED) between poll and projection) – Electionforecast and UNS performed a little better than Electoral Calculus. In the December batch. it turned out that EF and UNS had mean accuracy scores of 9-10, while EC was somewhat less accurate (> 13). EF managed to get the ED accuracy measure down to 5-6 after incorporating the poll results into their database. The lower the ED score the closer the model is to target, so this post-test just confirms that the model improved once the database has been updated.

    So, what does this kind of analysis reveal about the accuracy of the different models in predicting the January batch of Ashcroft polls?

    Electionforecast (Predicting SVI: Mean Euclidean Distance score – 8.3; Median 7.1; Predicting CVI: Mean ED score – 8.8; Median 8.0).

    Electoral Calculus (SVI: Mean ED score – 8.4; Median – 8.1; CVI:Mean – 9.3; Median – 9.8).

    So for this batch the two models perform equally accurately, perhaps with modest advantages for EF As a comparison I used the same measure to assess the accuracy of a UK-wide UNS model [2] and a Scotland-only UNS model [3]. Surprisingly, the latter was only a little less accurate (mean scores only, here) than the more sophisticated models (SVI: 11.7; CVI: 12.0). More predictably, the performance of the former was dreadful (SVI: 36.7; CVI: 37.3). But then no one seriously believes that Labour’s support has increased in Scotland in line with UK-wide figures.

    Turning next to May2015, this model doesn’t offer VI projections for each of the main parties and so cannot be evaluated using exactly the same approach. They only mention the predicted winner and runner-up and the margin. However, in the spirit of the ED accuracy approach, these margins can be compared with the Ashcroft margins across the different constituencies generating an average discrepancy score [4]. Using this measure EF, EC and UNS-Scotland all performed equally accurately (with all revised ED accuracy scores for both SVI & CVI falling between 10 and 12). UK-UNS again performed very badly (SVI: 49.0; CVI: 47.8). But more tellingly May2015 turned out to be less accurate than the other three models (SVI: 17.6; CVI: 18.4).

    What this means is that the May2015 VI-reallocation algorithms seem to offer less accurate projections of actual polling results than the other models do (with the obvious exception of UK-UNS). This is shown most directly by the fact that they wrongly predicted Labour wins in 8 of the constituencies and often by margins of 10-20%.

    So, what is the overall assessment of the different models? Electionforecast comes out best [5]. In the absence of specific constituency polling data, its VI reallocation methods are amongst the best on offer. Once constituency data becomes available, it is quickly incorporated into the database.

    May2015’s algorithms are less accurate at handling unpolled constituencies and are likely to be a little further off target in these cases. However, like EF, they immediately use constituency polling data and so the accuracy of the model can be expected to increase as more constituencies are polled.

    Electoral Calculus performed a little worse than Electionforecast in the December batch. However, this pattern was not repeated in January so the jury is still out on its algorithms. That said, a case can be made that Electoral Calculus is less accurate as a model than Electionforecast. This is because it doesn’t seem to make systematic use of constituency polling data once it becomes available. (As an illustration, look at its current projections for Thurrock). For this reason, I am inclined to be a little more sceptical about the projections offered by this model.

    Disclaimer: I have no connection at all with any of the modelling teams mentioned here.

    [1] For example on Jan 30 Airdrie and Shotts projected VIs of (6, 35, 2, 2, 53, 1, 1) in the 7-dimensional VI space: (Tories, Labour. LD, Ukip, SNP, Green, Other). The corresponding Ashcroft SVI point in the same space was (8, 38, 1, 4, 46, 2, 1). Subtracting Ashcroft from EF, the ‘discrepancy vector’ is (-2, -3, +1, -2, +7, -1, 0). The Euclidean Distance between the two is the square root of ( {-2 squared} + {-3 squared} + {+1 squared} + {-2 squared} + {+7 squared} + {-1squared} + {0 squared}) which is Sqrt (4+9+1+1+4+49+1+0) = SQRT (69) = 8.31. So, the EF projection was 8.31 units off target within this 7-dimensional VI space.

    [2] This is based on the assumption that in each constituency the VI changes are exactly in line with recent averages in national polls (e.g., Labour up from about 29%’in the election to about 32% now; SNP showing an increase of just 2-3% in line with UK-wide polling data, etc. etc.)

    [3] This uses recent recent YouGov Scottish crossbreak averages as posted by @OldNat, @Statgeek, @NC and others. The assumption here is that Labour constituency VIs have dropped in line with Scotland-wide averages, SNP support has risen dramatically and so on.

    [4] On Jan 30, May2015 predicted that in Airdrie and Shotts Labour would beat SNP by a margin of 8.06%. As indicated above the actual polling result was SNP 8 points ahead of Labour: a discrepancy of 14.06 units. By calculating a new ED measure over the 15 polls it is possible to created a revised margin-based accuracy measure for use with this model. (This is not directly comparable with the measure used above.)

    [5] The EF team’s ED figures are slightly different from mine. This may be due to details like my inclusion of ‘Other’ as a dimension, and my own exclusion of Plaid Cymru VIs from all of the January calculations. The models are all accurate in predicting 0% PC VIs in Scotland, and it is my view that including these figures introduced spurious accuracy. It would not be appropriate for an accuracy measure to be enhanced by succeeding in predicting that PC will not do well in Scotland.

  6. @Alan Kennedy

    “Labour only 5% ahead of the Tories in Scotland – who would have predicted that a year ago?”

    Most of the national Scottish polls, as opposed to crossbreaks in UK polls, have the Tories anything between 11-13% behind Labour, and that’s with Labour performing at almost historically low levels (i.e 27%). It would appear, post referendum, that the Tories have declined too in the face of the SNP surge but obviously not to anything like the extent that Labour have. Clearly, they have much less to lose in Scotland too and can, to some degree, sit back and enjoy watching the spectacle of Labour’s decline, but that shouldn’t disguise how poorly they are doing themselves. They are the sole centre right party in the whole of Scotland yet attract the support of barely one sixth of the population.

    Moving on, I wonder if this Kings Fund think tank report on the NHS will create any political waves? It is an even handed report by a wholly independent and expert group, and while it contains some criticism of Labour’s claims about creeping privatisation, and praise for some of Jeremy Hunt’s recent initiatives, it delivers a withering and devastating critique of Lansley’s reforms which, they conclude had “contributed to the increasing waiting times and declining performance that patients are experiencing.”

    Is this a gift to Andy Burnham?

  7. Addendum

    I have just looked at the latest Electionsetc projection and that reminded me that I meant to say something about my complete failure to mention this model in the accuracy assessments just posted.

    The reason I haven’t included this model in the exercise is that Electionsetc makes no public projections at all about events prior to the General Election. It is therefore impossible for others to benchmark the model’s interim performance.

    As part of the projection process, I understand that Stephen Fisher does work out projected VIs for individual consituencies. However, these are then converted into ‘probabilistic seat values’ for the final projection. So, for example, Labour might get the equivalent of 0.4 of a seat from one constituency, 0.3 from another, eventually building up to the published totals.

    Since he doesn’t share the interim steps in his calculations, there is no way for outsiders to evaluate how the model is doing right now. Of course, like the EF team, he could easily carry out the calculations himself and post information about the outcomes.

  8. It’s Grim for NC. )-:

  9. Tom Chadwick

    5 years of power and a possible job as a Minister ?’

    I would not expect MPs such as Kennedy, Sanders, Pugh, George even Farron to be interested in being so contaminated.

  10. Crossbat,

    “It would appear, post referendum, that the Tories have declined too”

    Unless a proportion of Tories have bought the LiS line;

    “Vote SNP get Cameron”.. It would be ironic if Labours appeal to their own traditional supporters fell on deaf ears but galvanised tactical voting against them.

    Probably the most bizarre result we could see, though definitely not one I am predicting, would by 59 Scottish seats split;

    51 SNP, 1 LibDem, 3 Labour, 4Tory……

    Peter.

  11. This is on the Politics Home website – the full article seems to be in the FT:

    “Ed Miliband is over twice as likely to emerge as Prime Minister after the general election than David Cameron, according to new research. An ‘outcome predictor’ developed by polling firm Populus and PR company Hanover found that while the Labour leader has only a 2% chance of leading a majority government after May’s poll, he is likely to have a greater choice of smaller parties with which to go into coalition than Mr Cameron.”

  12. The betting odds have moved very sharply this week towards Con’s most votes and most seats and DC as PM after the election.

    I Don’t gamble (other than Stock Exchange) myself but an interesting “straw in the wind”

  13. GRAHAM

    @” so contaminated.”

    I’m interested in your concept of Government without Cabinet Ministers.

    How would that work?-some sort of collective approach-or just leave it to the Civil Servants ?

  14. The irony of FPTP for Labour is highlighted by a thread on pb.

    FPTP, which gives such disproportionate advantage to parties with densely packed supporters in numerous small urban constituencies , punishes them disproportionately when those voters defect:-

    “It’s looking entirely feasible that the SNP could have the third largest Westminster delegation despite finishing sixth in votes. By contrast, UKIP (whose own figures are down for the third consecutive month) could easily end up with millions of votes but only a handful of MPs at most. ”

    David Herdson
    pb

  15. @TOH

    The betting odds have moved very sharply this week towards Con’s most votes and most seats and DC as PM after the election.

    This might be because the publicaton of the Ashcroft constituency polls have brought them up to speed about the implications of what has been happening in Scotland.

    However – as spelt out above – the Electionforecast model seems to have anticipated these polling results pretty accurately and their seat projections have barely changed over the last couple of months. So, I don’t think there are any new date to underpin this change in the posted odds.

    On the contrary, if anything I would expect this evening’s UKPR Polling Averages to show that Labour’s margin is a little more comfortable this week (though the improvement may not show up after integer rounding).

  16. @John160

    “Ed Miliband is over twice as likely to emerge as Prime Minister after the general election than David Cameron, according to new research….”

    Of course, that forecast is based on the reading of the current polling tea leaves, and these may change over the next three months, but a Miliband Premiership is indeed looking the likelier of all the potential myriad of outcomes.

    The polls look stuck at the moment, as Portillo observed on Neil’s programme last night, and I was amused by the the plaintive look on Neil’s face when he posed this rhetorical question. “If Miliband is so hopeless, and the Tories so much more trusted to manage the recovering economy, why aren’t they at least 5% ahead in the polls?”. (or words to that effect)

    An answer there was none and all we were left with was Neil’s rather despairing face peering, sadly and forlornly, into the camera.

    TV gold.

    :-)

  17. COLIN
    ‘I’m interested in your concept of Government without Cabinet Ministers.

    How would that work?-some sort of collective approach-or just leave it to the Civil Servants ?’

    I have suggested no such thing – simply that the individuals referred to would be unlikely to wish to participate in a Tory – led Government and are likely to oppose the continuation of such an administration. Some of them probably feel much as Franz Von Papen did in 1937 circa.

  18. UNICORN

    Not disagreeing with you, just thought it was interesting.

    At the moment I agree there is little movement in the polls. I have always expected the movement to come late ie just before the election, and strongly in the Con’s favour. We shall see in if I’m correct in due course.

  19. GRAHAM

    Ah- I see -only Tory Ministers are “contaminated”

    My mistake-I thought you had some interesting thoughts on Parliamentary Government.

  20. TOH

    Did you see the change in PB’s polling average for January?

    Maybe the reason?-being a site for political betting .

  21. CROSSBAT11

    ““If Miliband is so hopeless, and the Tories so much more trusted to manage the recovering economy, why aren’t they at least 5% ahead in the polls?”. (or words to that effect)”

    Can I suggest it is because the voters are not actually in the polling booth yet and they are not actually making a realchoice yet .

  22. COLIN
    The irony of FPTP for Labour is highlighted by a thread on pb.

    And, if we accept that his “personal” support for AV was real, the irony for Ed is that he could not persuade his MPs to support that “Yes” campaign.

  23. Colin

    Good morning to you, bright and sunny here.

    Yes of course that’s right, still interesting though.

  24. GRAHAM
    Some of them probably feel much as Franz Von Papen did in 1937 circa.

    Quite so, although I suspect he reached that stage by Summer 1934.

  25. Great post! Thanks!

  26. @UNICORN

    Wonderful that all your hard work in evaluating the different predictive models has been acknowledged and utilised :)

    I really enjoy reading your posts although I can only guess at some of the techniques you are employing (biologists tend to feel safest with regression and t-tests! ). Your love of making sense of the numbers shines through and IMO is very engaging.

  27. Unicorn

    I’m going to have to read your posts several times to be clear on some of the details – but that’s a compliment, not a criticism!

  28. Following up the discussion on SDLP vs Sinn Féin in Foyle I thought I’d dig out some numbers from the new Council elections in May 2014.

    Caveat that these are aggregate first preference STV figures from a best fit of Electoral Areas to the Westminster Constituency.

    I have used the electoral areas Ballyarnett, Faughan, Foyleside, The Moor, Waterside as my best fit.

    This gives totals of:
    Sinn Féin 12035
    SDLP 11745
    Unionists 7083
    Independents / Alliance / Others 5094

    This shows that it’s no walkover for SDLP any more but Sinn Féin is less likely to attract crossover support from the other parties than the SDLP.

    Note one of the successful independents is a convicted member of the Real IRA and also a member of the 32 County Sovereignty movement (Real IRA political wing) so some of those votes may transfer to Sinn Féin although given the internecine hatred there, possibly not.

    I’d still expect an SDLP hold based on those numbers.

  29. This from Parliament’s quietest coalition – the Labour/Co-Op coalition: http://labourlist.org/2015/02/miliband-unveils-plans-to-allow-workers-to-buy-out-companies-and-turn-them-into-co-ops/

    It’s certainly an interesting concept.

  30. Bantams (from last night)

    Very little IMO to do with the coalition, as a LD I know we became toxic when we promised to stop student fees, always thought it was stupid considering the chances of a hung parliament and the limited prospect of being able to implement it whilst power sharing. Whether you agree or not with the principle is irrelevant the fact we backtracked has put the party back by at least a decade.

    This is very widely believed and it may reflect what activists felt on the ground, but it’s not actually true. Because of the You Gov ‘daily’ polls we began to be able to monitor public opinion in a much more detailed fashion at this time and they tell a different story:

    https://d25d2506sfb94s.cloudfront.net/cumulus_uploads/document/rthm9kcfzj/YG-Archives-Pol-Trackers-Voting-Trends-with-UKIP-040215.pdf#page=3

    Lib Dem support began to fall from the election onwards and by the start of August it was already down to 13%. The Browne Report[1] wasn’t published till October and up to that point there was very much the impression that the Lib Dems had an opt-out from voting for an increase. It was explicitly stated as such in the coalition agreement, so people had no reason to expect Clegg’s switch[2].

    Though Lib Dem support did then fall a bit more from October, then reaching its 2011 level of 9-10%, most of the damage was done before well before that and can only have been caused by the act of coalition and, perhaps more importantly, the enthusiasm with which Clegg and the Lib Dems participated. Defectors may have felt retrospectively justified by the u-turn on tuition fees, but they had already left.

    [1] Which let’s remember was set up by Labour and there’s no indication they wouldn’t have accepted its findings – they introduced tuition fees after all. Like most government policies in recent times, tuition fees were basically intended to benefit the financial services industry at the expense of the populace and the country, but that hardly makes it less likely that Labour would have ‘reluctantly’ adopted Browne’s proposals.

    [2] It also should be pointed out that this change came very much from Clegg and his coterie, no doubt wanting to show they were ‘responsible’ – ie blindly following whatever the conventional wisdom was at the time in the Westminster Bubble. In actual fact slightly more Lib Dem MPs voted against tuition fees or abstained than supported them.

  31. Crossbat11

    I was amused by the the plaintive look on Neil’s face when he posed this rhetorical question. “If Miliband is so hopeless, and the Tories so much more trusted to manage the recovering economy, why aren’t they at least 5% ahead in the polls?”. (or words to that effect)

    An answer there was none and all we were left with was Neil’s rather despairing face peering, sadly and forlornly, into the camera.

    Maybe he’s on a success bonus?[1]

    [1] This is just a sophisticated way of saying LOL.

  32. Populus

    Lab 34 (-), Con 31 (-), LD 8 (-), UKIP 16 (+2), Others 11 (-2)

    Polldrums!

  33. Populus – Lab 34 – Con 31 – LD 8 – UKIP 16 – Grn 5

  34. Populus Scotland Crossbreak – SNP 40 – Lab 31 – Con 16 -LD 9 – UKIP 1 – Grn 2

  35. Number Cruncher’s graph shows a very slight uptick in Labour voting intention for Scotland recently: http://i57.tinypic.com/xfte9v.jpg

    If it continues at that rate I’d guess the SNP and Labour will be on roughly equal vote share by May – of course, there’s no guarantee that it will.

    Lib Dems, UKIP and Greens having a major battle for fourth place down there around the 5% mark, while the Tories seem to have suffered since the referendum too – probably losing tactical voters in all directions – Lab, Lib and SNP.

  36. As ever, this isn’t a place for policy debate

  37. @Mrnameless – the link to the cooperatives story is interesting. This is an area that is quietly being rediscovered.

    Invented in Victorian times, to meet the needs of ordinary workers and communities in an age where private business tended to wield untrammeled power, coops and mutuals were one answer to how to combine profit based business activities with social value. To be absolutely clear, they don’t guarantee anything in terms of good management or community benefit – just look at the mess the Coop group got into – but at least they provide a mechanism for broadening the benefits when they work well.

    It’s a little discussed success area of new Labour that they opened up company registration to a new breed of social enterprise structures, and to be fair to this government, they have continued the moves.

    I’m involved in a couple of Community Interest Companies and I’m anticipating helping to establish two cooperatives shortly, all with the intention of developing commercial business to generate profits for communities.

    We’ll never get away from the conventional shareholder model, which is still a good way to generate commercial investment, but there is a growing social enterprise and cooperative sector. In these days of austerity, communities will have to move away from grant funding towards other models of income generation, and anything that can be done to encourage this is welcome, in my view.

  38. Harry SP

    If you look at the long term picture, rather than just 92-97, it’s undeniable that there has been a secular decline in the trend line of Tory vote in Hallam since the 50s, with an acceleration from the 80s onwards.

    http://en.m.wikipedia.org/wiki/Sheffield_Hallam_%28UK_Parliament_constituency%29#/image/File:HallamGraph.png

    Peaks and troughs as national popularity waxes and wanes of course, but 97 was just a large drop superimposed on a long term trend.

  39. PS

    Of course, Labour didn’t benefit from that secular decline in the Tory Hallam vote. Labour’s own vote declined. The beneficiaries were the LDs.

    Look at it this way and Hallam is a perfect example of the Lib/Alliance/LD long term game of offering an all-things-to-all-folk safe repository of protest vote for those disillusioned by the main parties. And if Clegg loses in May, that will be the natural end game for a party that finally had to come down on one side or the other.

  40. Populus isle of Wight Crossbreak:

    Lab 4 Con 6 LD 2 Ukip 2 Grn 1

  41. @CB11 / Peter Cairns

    The Conservatives (in Scotland) have regained their losses from the referendum period. See:

    http://www.statgeek.co.uk/polling/periodic-averages/twelve-month-trendlines/

    Perhaps we’re seeing a tad of UKIP to Con and even the rarest of movers; Scottish Lab to Con movement. Or that’s what the chart seems to suggest. It might be churn.

    @Mrnameless

    “It’s certainly an interesting concept.”

    My instincts suggest it’s a doomed idea. In most cases the existing owners would not sell without the price being right (i.e. above market prices, so the new owners are already paying over the odds). In addition, most companies are run by those who know the company best. Some because they created it from scratch and experienced every teething problem and niche-specific problem that comes with creating and running a business.

    To hand a company over to people who do not have that knowledge and experience dooms the company to failure at worst and mediocrity at best (short of it being a company that is easy to run for plenty profit…not too many of those about). With mediocrity, the competition (perhaps in another city, but perhaps even in another country) wins and the company fails in the longer term.

    In the case of the company being put up for sale, while the new owners might immediately get rid of the management, they tend to keep key skilled or experienced staff. In short, they keep the best bits of the company and get rid of the less good bits, which sadly includes some staff (the pruning of staff isn’t always fair).

    Politicians really ought to stay away from business (and businesses should stay away from politics). It’ll never happen.

  42. Mr N

    NC’s graph also shows a slight uptick for Cons in Scotland recently.

    Of more interest, I think, is any patterns that are developing in 2015 (the “slight upticks” bit), rather than in a graph dominated by the 2014 SNP surge.

    Granted the much more limited value of my YG Scottish crossbreaks since 5 Jan, they seem to show Con and Lab generally rising and falling together, but Con gradually creeping ever closer to the Lab position, as Lab falls are greater than rises, while the opposite is true for Con.

    Of course, even if that perception were to be confirmed by statisticians, it would be of little use in terms of seat predictions, where Unicorn has shown that the best guide is currently Election Forecast.

  43. I think it is fair to say that it is likely that the Populus polling is not taking account of the full drop in Labour support in Scotland.

    On the same sophisticated cigarette packet model I employed yesterday, if Populus are showing Labour support in Scotland 6-8 points higher than YouGov, then that would show Labour about 0.5-1 points too high.

    Knocking that off the Labour total still leaves Labour fractionally further ahead with Populus than YouGov, but not different in any meaningful way.

  44. OldNat, yes, my post yesterday WAS about Sheffield Hallam and the consitutency specific polls on this seat.

    The most important point I made is that when a poll with a good sample size shows a ten per cent lead this lead cannot credibly be explained away by methodological issues.

    Labour is in the lead in Sheffied Hallam.

    I also made the point that the Conservatives are not out of it in this seat if there is a swing to them as the election approaches, which I think is not at all unlikely.

  45. Populus Crossbreak Barnard Castle –

    SNP 0 – Lab 0 – Con 0 – LD 0 – UKIP 0 – Grn 0 – PP 0 **

    To be honest there doesn’t seem much interest here at the mo.

    Could be the excitement of the farmers’ market I s’pose……

    [** PP = Pups’ Party – they’ve bin wuffin’ at a few people but no serious campaigning as yet]

  46. Frederic Stansfield

    It was your asking Shadsy for the current odds on UKPR that seemed odd. :-)

  47. @Unicorn

    Your insistence for using this site for non-partisan discussion of polls is most irregular.
    Why can’t you be like the rest of us and engage in partisan blather and wishful thinking?
    This abject behaviour has now come to the attention of outsiders and if you don’t desist there is an imminent danger that this site will get a good name.
    Please stop showing so many of us up with your hard work, intellect, technical grasp and impartial approach, it’s sooo unfair.

  48. @Guymonde

    +1

  49. Guymonde

    Hear, hear. Though I fear it is even worse than that.

    Unicorn shows all the signs of being qualified to make his subversive posts.

1 2 3 4 5 6