Political Betting has the latest results from a new AngusReid poll. Their topline figures are CON 38%(-2), LAB 26%(nc), LDEM 19%(+1). As ever, Angus Reid show a lower level of Labour support than other companies (for my thoughts on why see here), but the narrowing gap between the Conservatives and Labour echoes that elsewhere.

Angus Reid have also produced some figures for groups of marginal seats – note that these are properly weighted seperately from the main poll, and not just cross-breaks. It was also a larger poll to ensure they were decent sized samples.

Amongst 150 Labour/Conservative marginals these figures showed a swing of 12% from Labour to the Conservatives, a lot higher than the swing recorded in the country as a whole. As far as I’m aware every poll of marginal seats in the last year or so has shown the Conservatives enjoying a larger swing in the marginals to varying degrees. I’m always slightly wary of the exact difference (and the extent we can draw meaning) but the pattern is at least strongly consistent across studies by all different pollsters.

Amongst Liberal Democrat held seats the shares of the vote, with changes from 2005, are CON 33%(+4), LAB 16%(-3), LDEM 39%(-7). Here I would urge more caution – I have severe reservations over marginal polls carried out in Lib Dem seats.

Essentially, I have rarely if ever seen such a poll that doesn’t report a big drop in Lib Dem support, even when subsequent elections don’t reflect that. Often they also show increases in Labour support even when the party is down. For example, in November 2004 Populus did some private polling for Michael Ashcroft in Lib Dem held seats being targetted by the Tories – they found the Lib Dem support down by 8 points. The following year the Lib Dem share of the vote in those seats remained static.

My guess is that this is because people do not factor in local candidates or tactical decisions when answering voting intention questions (or perhaps do not consider these things at all until an election is called). Naturally enough, these factors weigh the heaviest in Lib Dem constituencies.

In the PoliticsHome marginal seat polls in 2008 and 2009 I used very heavy prompting, and asked people two versions of the voting intention question to push people towards really considering how they would vote locally… and it massively increased Lib Dem support (and showed them holding their own against the Tories in some parts of the country). I’m not necessarily saying those were more accurate – it’s never been tested at an election and for all I know I could have overegged it – but I always take voting intentions in CON -vs- LDEM marginals with a great deal of salt.


126 Responses to “Angus Reid marginals data”

1 2 3
  1. Ron Hist and Old Nat

    Thanks for the information that the CON-LD data are based on all 60+ currently (or presumably notionally) LD-held seats.

    Taking into account the ragbag factor (from Cornwall to Orkney and Shetland, via eg Bristol, Hornsey, Sheffield and Fife), and accepting all of Anthony’s strong reservations, the LD->Con swing looks pretty much in line with the average current polling figures, rather than wildly different like LAB->CON. This is apparently in keeping with LDs’ own confidence about holding most and even gaining some, once the real GE campaign begins.

    However, I find it difficult to believe that in the seats they hold, the LD average 2005 vote was 46%. That would mean a lot of seats held with absolute majorities. Seems unlikely to me. But, unfortunately, I am a long way from my 2005 Times HofC Guide at the moment, so not able to do even a quick rough check. I would be interested (and grateful) if anyone is able to check AR’s 46% figure for 2005.

  2. George Gardner

    You probably misinterpret Brown’s “popularity” in Scotland.

    Given that the choice at UK level in that Presidential election for a Parliament which has nothing to do with our domestic affairs, who would you expect us to prefer?

    A Brit from Scotland whose core base is in Scotland and the North of England, or a Brit from England whose core vote is in the English shires?

    Neither looks much good, but in a forced vote, I’d marginally prefer Brown to Cameron.

    We’ve now had our own Parliament for 10 years. The implications of that, are much more deeply embedded here than they were 5 years ago.

    Scottish politics are not just different because we have the SNP as a player.

  3. @GEORGE GARDINER

    “There are supposedly odds on GB becoming the next First Minister of Scotland, giving up his Kircaldy and Cowdanbath seat in the commons following a UK Election defeat and re-standing for Labour in Scotland at a later date – at least so I am told”

    Ha Ha Ha- ‘OLDNAT’ are you there……….

  4. @OLDNAT

    “Scottish politics are not just different because we have the SNP as a player.”

    Well *Scotland* is different from England and that has NOTHING to do with the SNP.

    But scottish politics is pretty much the same game (aside from the labour introduced reforms on parliament and electoral system) up there as my Brother-in-law often moans to me.

    For that he blames squarely the awful Salmond and Swine – ney: who he says are hated much more than the UK PM.

    He has remarked on Brown going back home and running Scotland.

    As i still predict a minority Cameron government with Brown falling on his sword/ stepping aside within 72 hours I reckon he would….and he would walk it.

  5. From memory the last big non AR marginals poll showed a swing of 8.5% against a national swing of 6.5% or so I recall. That was before the current tightening. I remember posting at the time that the differential wasn’t that great, and if the polls tightened but the marginal differential stayed the same the Tories would not make a majority. This AR poll shows a much wider gap, both nationally and in the marginals, and I don’t think it resolves any of the issues. Essentially, you either believe AR are accurate and we are on for a huge landslide, or the other pollsters are more in line, and things are very tight. It’s not at all scientific, but the ‘feel’ of this campaign just isn’t the same as 97 – I just don’t get the sense that cameron is going to be swept in with a huge majority with rejoicing crowds etc etc. I don’t think the Tories sense this either if they are honest. I refuse however, to write off AR just because of how I feel. I’ll just keep looking forward to election night and a resolution of this debate.

  6. @Wolf, seems about right to me. Ya can find results on wikipedia in each constituencies article.

  7. @GARETH

    “Anthony, what we need is a poll of marginal polls so we can follow what is happening in the seats that make a difference and a projection based on that. I’m the guy who suggested you put poll of poll figures to one decimal place so no surprise there.”

    Having a running average of ALL marginal polls would seem to make the site more consistent and give a fairer snapshot of what might be happening.

    It is a little strange that we only report the marginal data (whether a sub set or a valid sample) collected by one polling organisation- and the one that is clearly the outlier from the average of everyone else.

    So come on Anthony:

    go back as far as you can and dig up the marginal polls there have been and calculate a running average as of today that can then be modified with each new one.

    There are going to be quite a few over the next 10 weeks…..

  8. The actual figures in the marginal polls would be almost irrelevant, as they would be spread over such long intervals.

    What would be useful would be the relative swings in the marginal polls, compared with GB wide polls at the same time.

  9. @George Gardner ‘I believe the polls are now showing the Tories at their bare “core” vote’

    What’s the belief based on George? The recently published Social Trends data would suggest core Conservative support (not the same thing, I grant you)to be around 31/32%, rather than 39%.

    Along similar lines, yesterday’s Guardian ICM poll had some fascinating data on flexibility of voting intentions, which pointed to the maximum potential support of the three main UK parties as C 43, Lab 34, LD 29. That does more than any other poll I have seen to illustrate Labour’s problem in getting anywhere near a parliamentary majority, even with their huge electoral system advanatage.

  10. I would like to propose PR on this site in order to give the English more of a chance ;)

  11. @NEIL A

    “What would be useful would be the relative swings in the marginal polls, compared with GB wide polls at the same time.”

    On reflection I agree with this- presumably every MP will have a GB poll linked to it/ run along side it.

  12. Rob Sheffield

    I’m basing my comments on the polls and media commentary in Scotland.

    You base yours on your brother’s comments.

    Enough said.

  13. Rob – unfortunately every company polls different groups of marginal seats (and often one company will poll different groups of marginal seats because different clients have requested different things, or because the situation has changed over time.

    For example, when ICM did a poll of marginals last month they didn’t bother doing some seats with small Labour majorities because they were assumed to be Conservative gains anyway. The polls have tightened since then, so if ICM did the same exercise next month they might include some tighter marginals.

  14. @OLNAT

    Do you realise how pompous you just made yourself sound ?!

    I think he bases *his* comments on the polls and the media commentary in Scotland and then relays that to me whenever we speak.

    I know you continuously try to make yourself out to be the ‘voice of Scottish independence and reason’ on this site but just what makes your Scottish-based opinion any more valid than his Scottish-based opinion…???

    Absolutely nothing. “Enough said”.

  15. Anthony

    “unfortunately every company polls different groups of marginal seats (and often one company will poll different groups of marginal seats because different clients have requested different things, or because the situation has changed over time.”

    But surely it would be worthwhile to track- as per Neil A’s suggestion- the marginal-national (i.e. UK) differential where one is identifiable ?

  16. I know that “Angus Reid bashing” is quite popular, but consider this:

    Since 1983, of all the final pre-election polls,

    29 polls overestimated Labour,
    1 poll got them spot on,
    1 poll underestimated Labour, but was still the most accurate.

    So, for the past 30 years, if you bet that the most anti-Labour poll was the most accurate, you’d be minting it. You’d have been right every time.

    So you can’t just dismiss the most anti-Labour of the pack as inaccurate, since for 30 years, they have been the most accurate every time.

    No doubt things might change this time. But you must justify saying “I don’t trust the Tory-leaning outlier”, since you’d have been wrong for the last 7 times in a row.

  17. Anthony, do you think AR will have a seat by seat analysis or is that too deep a drill down to expect?

  18. Wolf MacNeill,
    I find it difficult to believe that in the seats they hold, the LD average 2005 vote was 46%. That would mean a lot of seats held with absolute majorities.

    The L-Ds had real majorities (50%+ of the vote) in 2005 in 16 seats, 4 of them Scottish. They had more that 45.5% of the vote in another 19 seats, 2 of them Scottish.

    Their seat with the lowest vote share – 34.88% – is Falmouth & Camborne, unless you count Dunfermline & West Fife which they only won in a by-election.

  19. George Gardner,

    Far more likely is Jim Murphy standing for Holyrood 2011 if he is defeated in his constituency (a distinct possibility). Rumour in Glasgow in both Conservative and Labour circles is that is Murphy were to lose East Ren the plan would be to give McConnell a peerage in the dissolution honours and parachute Murphy into Motherwell & Wishaw for 2011.

    If that were the case he would probably become Labour leader in Scotland after the 2011 election (he already is in all but name) unless Labour won the election (in which case he would have a very high profile position under Iain Gray).

    Of course it is all just rumour, but it sounds very plausible.

  20. Rob Sheffield

    You mean you are both polling geeks? :-)

  21. @ Richard Manns

    “So, for the past 30 years, if you bet that the most anti-Labour poll was the most accurate, you’d be minting it. You’d have been right every time.”

    Or, most likely, ‘less wrong’.

  22. Angus Reid are onto something very important here.

    The national swing against the Tories when they lost power in 1997 was 10.0%.
    There were ample examples of marginal seats where the swing was far greater than that. For a start, Michael Portillo would have held Enfield Southgate on a uniform swing, and might have ended up as Tory leader. Even in 1992, the Tory majority would have been over 70 on a uniform swing (not the 21 it ended up as).
    This proves the point that national opinion polls and the media interpretation of them potentially give such a misleading picture.

  23. @Richard Manns

    You are on very dodgy ground for several reasons when trying to use the ‘overstate labour support’ notion as a counter argument to people asserting that there is something unreliable about AR’s consistently higher Tory leads than all the other organisations.

    This labour bias has been done-to-death in many previous threads.

    Nonetheless I’ll take you through it again.

    The only valid comparison is the 2005 election based on the fact that this is the most recent ‘battle conditions’ example based on updated methods and approaches that the mainstream polling organisations had implemented over the years since the 1992 fiasco.

    So, 2005 eve of election polls:

    Mori was 2% over for Labour (38%) and dead on with Tories (33%)
    ICM was 2% over for Labour (38%) and one under for Tories (32%)
    Populous was 2% over for Labour (38%) and one under for Tories (32%)
    YouGov was 1% over for Labour (37%) and one under for Tories (32%)
    …..and NOP was SPOT ON.

    A miniscule ‘overstatement’ comfortably within the MOE !!!!

    Also check out Ivor Crew in ‘Parliamentary Affairs’ – “The Opinion Polls: The Election They Got (Almost) Right”

  24. Oldnat

    are you teasing about Rob’s brother? You’ve mentioned it about three times now. Are you winding Rob up ?

  25. Anthony

    Would it be possible for us to have a little icon ID each?

  26. @ANDREW

    “Angus Reid are onto something very important here.”

    Really?

    I think you’ll find plenty of posters on here are very sceptical of their results and very worried about their methods compared to the mainstream poling organisations. .

    As has been said thousands of times (literally) before: either AR are wrong or everyone else is.

  27. @ Brownedov ‘The L-Ds had real majorities (50%+ of the vote) in 2005 in 16 seats, 4 of them Scottish. They had more that 45.5% of the vote in another 19 seats, 2 of them Scottish.’

    Many thanks for that information. I’m very grateful to you – as well as surprised that LDs have overall majorities in a quarter of their seats. Clearly AR’s 2005 base data look OK.

  28. Al J

    I only tease people who distort politician’s names to express their hostility to them.

    You know, the sort of people who talk about “Gollum” Cameron

    (our ‘precious’ Union – “We wants it, we needs it. Must have the precious. They stole it from us.”)

    That sort of person :-)

  29. This is exactly what I said was happening in an earlier thread, about a week ago. It’s won or lost in the marginals & I reckon that the Tories are set to take many off labour, which will hand them their majority of over 100 seats.
    I will be extatic if AR is proved to be correct & all the others wrong. That will be embarrassing to all those who have slagged off AR for months now, just because they don’t provide the expected answer!

  30. Anthony

    Is there a chance of having a separate ‘Scottish Polling Report’ site?

    So a certain very sad little individual can go and play all by himself !!!

  31. @ROB

    “I will be extatic if AR is proved to be correct & all the others wrong”

    I’ll remind you of this the day after the election

  32. @Rob Sheffield,

    If there was a seperate Scottish site, half our regular contributors would abandon us! (including most of the more sensible ones…)

  33. (and OldNat)…

  34. oooooh, I’d love to have a little wager……
    *sigh* just have to wait and see. Actually as the geek I am, I think I’ll just nip off and have a little pore over the 92 polls……

  35. Neil A

    :-)

  36. Neil A

    are you sure? Not just the one(s) trying to make a scottish issue out of everything !!

  37. I think the first comment from Neil is the most important one above. Also all other polling evidence we have seen for the marginal has shown much greater swings to the Tories in the marginals than generally. I accept the other marginal polls e.g. ICM last month are not strictly comparable.

    The evidence we have seen including the analysis from Andy Cooke suggests that at the time the AR marginal poll was taken the Tories were on track for a large overall majority.

    It would be naive however to assume that the polls will
    not change significantly before the GE and the result
    may still be subject to a lot of events including the Ist Q’s GDP growth/ decline, the Budget, the Iraq Enquiry etc.

  38. Sometimes they remind me of the father-in-law in “My Big Fat Greek Wedding”…

    “everything is originally Greek”…

  39. I’d be there – put me down as Jock McTT

    (Why is the Captcha code HRH2 ??

  40. @MIKE

    “The evidence we have seen including the analysis from Andy Cooke suggests that at the time the AR marginal poll was taken the Tories were on track for a large overall majority.”

    Only if you take AR ‘results’ at face value.

    Rather than viewing them as outside the range of the mainstream polling organisations by between 6 and 9%.

    One would expect a similar differential in their marginal figures as both are based on their methods of selection, sampling and questioning.

    As said- nothing can be read into this AR marginal poll until polls by mainstream organisations are published on the marginals as well and we can compare them: both between them and within them in terms the UK-Marginal differential.

    Anyone ‘banking on’ or- even worse- ‘betting on’ a 100 seat Tory majority based on their AR poll is simply bonkers !

  41. Surely “Jock” is the sort of name people would never be called on that site? I mean how would you tell them apart?

  42. Rob Sheffield

    You don’t seem to have understood my post. I agree that AR overstate the Tory lead (though the other polling organisations, if anything, do the opposite).
    The important thing they are onto is showing how different the picture looks in marginal constituencies – a phenonemon that applied in 1997 and 1992.

  43. Reading the posts I’m beginning to think AR is wrong about the marginlas. If they give the Tories a bigger lead by 100% to some other national polls , they must be doing the same on the marginals.

    AR can’t be right if any of theothers are!!

    Can anyone tell me -are we having a poll from Yougove every night please?

  44. A little thought experiment. Or an experiment of little thought, depending on how you like my logic.

    Labour polled 36% of the vote in 2005, so even with those Labour supporters who didn’t vote last time coming back to them, it is unlikely they’ll get more than 34% this time.

    The Conservative lead is slipping, it may currently be down to 6 points. For arguements sake let’s say it slips a bit further to 5, giving a result of 39-34 for the election.

    Taking AR’s latest poll, it has a 12% swing in the marginals, but understates Labour’s support compared to other pollsters, so let’s bring that down to say, 8%.

    I believe the above would be a best-case scenario for Labour, but surely the only way the Conservatives wouldn’t get an outright majority is if somehow the return Labour voters was higher in the swing seats than in the heartlands. Is this likely? In northern / Scottish seats perhaps, if the main drive is fear of a Conservative win. In the South however I really can’t see Labour retaining marginals, I just don’t think the majority of floating voters living in swing constituencies down here have as much to lose from cuts in public services.

    In conclusion, I can’t see the Conservatives having less than a 20 seat outright majority. I’m also pretty sure that due to the vagaries of our voting system there is actually only a fairly narrow configuration of Labour, Conservative and Lib Dem results that could cause a hung parliament and only a little variance in either direction pushes it to an outright majority. I’ll leave people with more experience of politics and polling to correct any errors I’ve made in my assumptions ;-)

  45. ROB SHEFFIELD

    No Rob I am not taking the AR results at face value and I referred to Neil’s comments i.e. if “we allow for Angus Reid overstating the Tory lead by about 4 points” ……

  46. Al J

    YouGov polls most nights, I think (for Tues-Fri papers?)

    If nowhere else, they’ll be released on PB at 10 pm.

  47. @Al J,

    I agree that AR are probably overestimating Labour in the marginals poll just as in their general polling, but its the relative levels that are important. If their metholody is uniform, and the marginals show a greater Tory lead, that still tells a story.

  48. @andrew

    “a phenonemon that applied in 1997 and 1992”

    See my response to ‘richardmanns’- on looking at what happened in 2005 and why that (rather than 92 or 97) is where we look. Anthony has posted on this as well.

  49. I have given up setting much store by Angus Reid polls. A swing of 12% seems to me to be much higher than can actually be the case, and is wildly out of line with the national polls now. Some like Roland will hope that AR are right and everyone else is wrong, but even the bookmakers aren’t forecasting that sort of swing in any (semi-) target seats. You never know, they could be right, but it does fly in the face of virtually all other evidence or forecasts.

  50. @Neil A

    “I agree that AR are probably overestimating Labour in the marginals poll just as in their general polling, but its the relative levels that are important. If their metholody is uniform, and the marginals show a greater Tory lead, that still tells a story.”

    He said AR were overstimated THE TORIES in this marginal poll…..

    We’d actually anticipate the differential and Anthony or someone has posted on the ‘exponential marginal diferential’ i.e. that marginals are more volatile and the difference between the Tory lead goes up (and comes down) faster in them than the UK poll.

    Bottom line- its still ARs figures and we all need to wait until we have comparable marginals data from ICM or YG etc etc.

1 2 3