The monthly online ComRes poll for the Sunday Indy and the Sunday Mirror is out tonight and has topline figures of CON 28%(nc), LAB 36%(-1), LDEM 10%(+2), UKIP 17%(-2). The fieldwork was conducted on Wednesday and Thursday, so at just the same time as the YouGov polls showing the Labour lead almost vanishing, but doesn’t show any significant narrowing. Populus’s poll yesterday also showed things rumbling along as usual.

In the weekend papers we should also have the fortnightly Opinium poll for the Observer and the YouGov poll in the Sunday Times.

UPDATE: The Opinium poll in the Observer is out, and has topline figures of CON 29%(-1), LAB 36%(+1), LDEM 7%(nc), UKIP 17%(nc). No narrowing (or Lib Dem conference boost) there.


108 Responses to “ComRes/Sunday Indy – CON 28, LAB 36, LD 10, UKIP 17”

1 2 3
  1. Second?

  2. Get it together Comres.

    Also first :3

  3. Another wrong prediction.

  4. Oh damn it Phil

  5. So those last 2 yougov’s definitely looking like outliers?

  6. That’s certainly an unexpected one. After the very unpopular Brown government, has anyone considered the possibility of Shy Labourites?

  7. And more clear support for scrapping the bedroom tax:

    Ed Miliband should promise to scrap the so-called ‘Bedroom Tax’ if Labour wins the next General Election

    Agree 51% Disagree 26%

    And more bad news for conservatives on the economy:
    I trust David Cameron and George Osborne to make the right decisions about the economy

    Agree 28% Disagree 49%

    But even worse for Labour:
    I trust Ed Miliband and Ed Balls to make the right decisions about the economy

    Agree 23% Disagree 50%

    Lots of negative responses to Ed Miliband, it would be interesting to see how those change after the conference now that Labour actually have some policies.

  8. Similar results from Opinium

    Toby Helm [email protected] 3m
    Op/Obs poll. Lab 7 point lead. Lab 36, Tories 29, Ukip 17 (taken before chaos conf) Lib Dems 7 (no bounce). E Mil ratings pretty dire.

  9. The two YouGovs (plus the July ICM) get considerable weighting from BBC R4’s political editors. Their commentary this evening was that ‘generally speaking’ Labour and Conservative are level pegging in the polls.

  10. I must say, the fact that our two tracking polls repeatedly show trends that bear no discernible relationship to each other is rather depressing. Forget the monthlies, we don’t really have good baselines for them, but you’d hope YouGov and Populus would pick up the same movement or lack of it.

    Maybe we’d better off just reading tea leaves, or asking Rosie and Dasie.

  11. billybob

    “Their commentary this evening was that ‘generally speaking’ Labour and Conservative are level pegging in the polls.”

    It really does smack of the US media insistence that the presidential election was “too close to call” and “neck and neck”, right up to the actual poll where they could no longer agree the narrative amongst themselves and Obama wiped the floor with Romney…exactly as the polls indicated all along.

  12. The Labour share is around 35-37% pretty consistently it’s the Tory share that is all over the place

  13. The interesting things about YouGov as compared to these two tonight is that the only difference is the weighting of UKIP- add 7 points onto Tory and 7 off UKIP (not quite fair I know) from these two polls and you get very similar results to YouGov.

  14. Am I the only one who thinks it was a tad strange to have two opinion polls showing a diminshed Labour lead ahead of their conference. Why this was the case I am not sure. Why would there be two days when a sample is giving similar results, which are not in line with other polls ?

    We will see from Tomorrow onwards whether YG are showing something different to the other polling companies.

  15. Nick P

    It really does smack of the US media insistence that the presidential election was “too close to call” and “neck and neck”, right up to the actual poll where they could no longer agree the narrative amongst themselves and Obama wiped the floor with Romney…exactly as the polls indicated all along.

    Please, please, please can they get Dan Hodges on television on election night to play the part of Karl Rove. That meltdown was a thing of beauty.

  16. Hmmm, two polls both showing the Tories in the 20s. After the recent signs of economic recovery, and a very choppy period for Miliband and Labour, isn’t that a bit worrying for them? Taking the last two YouGovs aside, there’s absolutely no sign of a Tory revival in any of these polls. The only consolation for them is that Labour are now regularly polling in the mid to high 30s, but to still have some pollsters scoring you in semi-meltdown territory must be quite disconcerting at this stage of the Parliament and with the political narrative, supposedly, starting to track in a favourable direction.

    I wonder when this Tory revival is going to actually materialise.

  17. @Richard

    Not sure if they are outliers, or if YouGov’s shaping is more sensitive to letting party-conference distortion through than ComRes’s shaping. Or which would be the “better” report, to report the distorted result, or a reported result that smooths out the wrinkles a party-conference generates.

    Again, my advice is to ignore polling conducted during and immediately after any major party conference.

  18. There’s no way ukip is that high in reality nor the libs that low …increasingly cynical/sceptical about polls, their timings and their methods – oh and what gets promoted from a poll and what doesn’t. Agenda aplenty!

  19. Nick P

    The pro-Republican media were focussing solely on the national popular vote in the run up to last November. This was either deliberate or stupid (not sure which) but anyone following the key battleground states would have been pretty damn certain that there was no path for Romney to win the Electoral College.

    In this country we focus almost entirely on the national vote apart from when there is a by-election. I am not sure how good a guide these will be to 2015 – I would like to see more polls along the lines of Ashcroft’s marginals polling, which I think are going to be a lot more useful.

  20. Of limited use yes, but interesting also. These are real people voting, so info is worth something.

    Election results from 20 by elections in the last 3 weeks:
    Conservatives 7,849 votes 34.02%
    Labour 6,149 votes 26.65%
    UKIP 4,713 votes 20.43%
    Liberal Democrats 2,053 votes 8.9%
    Independents 1,192 votes 5.17%
    Green Party 977 votes 4.24%
    BNP 120 votes 0.52%

    Difference from last time these seats were fought.
    Conservative down 3.13%
    Labour down 1.67%
    UKIP up 12.86%
    Liberal Democrats down 4.89%
    Independents down 2.2%
    Green Party up 1.14%

    Previous 3 weeks:
    CON -3.2% LAB -3.7% UKIP + 12.7% LD -2% IND -2.9% GREEN + 1%

  21. @ HOOFHEARTED

    How many of the 20 council seats were Tory before these recent elections ? Sometimes when you see stats like these, you have to look behind them.

  22. Real people answer polls.

  23. BILL PATRICK

    “Real people answer polls.”

    With some (non-BPC. of course) pollsters, that assertion may lack accuracy. :-)

  24. It looks like Nigel Farage is ruling out a Ukip/Tory pact:

    http://www.telegraph.co.uk/news/politics/ukip/10325142/Nigel-Farage-rules-out-pact-with-Tories-before-next-election.html

    Opposition pledges are written in water, but he could have evaded the question and he was making a big point in that interview of Ukip’s Northern ex-Labour voters, who would be driven away by any collusion with the Tories.

    So do we think it’s off the table?

  25. Real people vote for real councillors in by-elections, and shows which parties are growing/shrinking their vote.

    Yes Tories only lead the share of the vote because more are Tory held, it also shows despite the rise in UKIP’s vote, Tory votes surprisingly stayed strong. It looks like UKIP are taking support from LD’s, Lab and Tory.

  26. OllyT thinks apparently as I do. The national polling is useful for identifying trends but the reality lies in the marginals, just as it does in the US. Win Ohio and you win.

  27. Ashley

    Well it depends on how you weight the Ukip vote and here I suspect it’s weighted kindly, but even with the most advantageous weighting the Ukip score goes down from the raw figures, so it is entirely possible that Ukip are actually more popular than the polls indicate

  28. @Billy Bob
    “The two YouGovs (plus the July ICM) get considerable weighting from BBC R4?s political editors. Their commentary this evening was that ‘generally speaking’ Labour and Conservative are level pegging in the polls.”

    That really takes the biscuit, even for BBC R4. Back in July, ICM were the only game in town for the BBC when they had the parties level. Now its YouGov, for so long ignored when it had double-digit leads, but suddenly the only polling company worth taking note of.

  29. @Howard

    I agree. It would be nice to have a column added to the current polling tables to break out voting intention by marginal vs non marginal seats. It should be easy enough to do – just ask people which constituency they are in, then split those into relevant groupings.(eg Tory/Lab marginal; Tory/LD Marginal; Tory/LD/Lab marginal)

  30. That’s why Lord Ashcroft is a hero. I may not agree with his politics, but he’s providing the public with the best polling we have.

  31. HOOFHEARTED

    Yes, better to look at real results from by-elections to gauge public support. Because YouGov publishes five polls a week many people here seem to take its results as a benchmark. Whereas if it has a flawed methodology to start with, those results only have a significance relative to each other.

    My view, for what it’s worth, is that YouGov is consistently underestimating UKIP support by about 4% and overestimating the other main parties by perhaps Conservative 2%, Labour 1% and Lib Dems 1%.

  32. AOL poll:

    Do you agree with Labour MP Rachel Reeves that people earning £60k are ‘not rich’?

    Yes34%

    No66%

  33. Prima facie YouGov seems to be rather out of line with other pollsters for the last week or so. Comres shows no real narrowing – Opinium gives Labour its biggest lead for six weeks – even ICM showed Labour’s lead edging higher over the last two months. Only MORI seem to be moving in the same direction as YouGov – but its reliance on 100% certainty to vote figures makes its polls volatile.. A bit puzzling really!

  34. @Spearmint – ” …he could have evaded the question.”

    That sounded very much what he was doing (to me at least) when John Humphries ambushed him with the question of a Con/Ukip pact this morning.

    Judge for yourself (at 1:20.33):

    h
    ttp://www.bbc.co.uk/iplayer/episode/b03bd23t/Today_21_09_2013/

    Jacob Rees-Mogg had a piece in the Telegraph some time ago where he argued that the “not with David Cameron as leader” clause is but a bargaining chip, which Farage will drop in return for something unspecified.

    Earlier in the interview with Humphies, Farage indicated that UKIP will draw up their list of target seats after the EU elections/locals… until then the agenda is erode Labour’s VI to some extent, by crowding Labour out of the media narrative.

  35. @OLLYT
    Nick P
    The pro-Republican media were focussing solely on the national popular vote in the run up to last November. This was either deliberate or stupid (not sure which) but anyone following the key battleground states would have been pretty damn certain that there was no path for Romney to win the Electoral College.

    This isn’t quite true but there was certainly a lot of over optimism about Republican polling. That said the most striking feature of the US election was the size of Obama’s popular vote victory. It was not a close election by historic comparison with say 2000, 1960, 1968, 1892 or even 1976 etc.

  36. Hoofhearted

    I think the 60k figure is to reassure folk that are earning a lot less than that but have hopes of earning close to that amount even if their hopes might be unrealistic. I can imagine someone on 30k being reassured that even if they do get that promotion which they have been waiting 5 years to get they still might be better off with a Labour govt

  37. ” …by crowding Labour out of the media narrative.”

    or

    ” …by keeping immigration at the top the list of media talking points.”

  38. @Richard @ Howard
    “It would be nice to have a column added to the current polling tables to break out voting intention by marginal vs non marginal seats … ask people which constituency they are in, then split those into relevant groupings.”
    but with even as many as 60 relevant marginals out of 600 seats, and say an overall sample of 1800, those results will be based on a sample of about 180, with a consequent large random error.

    @Graham
    Do remember that all individual polls have quite large random errors, and two successive polls may quite easily err in the same direction. Average the last 6 or 10 YouGov polls and then the next 6 or 10 and sequences of 6 or 10 moving up the starting poll between those limits and you will perhaps get sufficient accuracy to draw valid conclusions.
    Do not believe that such a result can be achieved more quickly by averaging polls from different sources. Published result show that there are clearly systematic differences between pollsters which make that invalid – a bit like measuring a length with a steel rule and a stretched tape, when the ruler groups its results around 20cm and the stretched tape its results around 19cm. The ‘true’ length is not 19.5cm.

    The desire to “see” and comment on possible changes usually outweighs patiently extracting valid information from limited data.

  39. “Maybe we’d better off just reading tea leaves, or asking Rosie and Dasie.”

    It’s DAISIE and I am WON tomorrow. But yes, me and Rose know wotswot and there is no chance of a Tory majority in 2015 plus the Lib Dems are dooooooooooooooooooooooooooooooooooooooomed.

  40. @Spearmint at 19.20 (above)
    ” Please, please, please can they get Dan Hodges on television on election night to play the part of Karl Rove. That meltdown was a thing of beauty.”

    I suspect Dan Hodges might do “a Don King”. Don, no stranger to controversy, once famously promoted a fight for one of his top fighters with whom he had a long standing relationship. In the instant his, fighter was vanquished, Don King jumped into the ring into the, crowd congratulating the, winner, holding the winner’s arms aloft.

  41. Dave
    It was Richard’s suggestion and clearly the ‘cross breaks’ would not mean anything. Ashcroft’s polls however *are* significant and we heading for a Labour landslide (as things stand).

  42. @RosieandDaisie
    “…..plus the Lib Dems are dooooooooooooooooooooooooooooooooooooooomed.”

    Expect a response from @Ashley along the lines of:
    “…increasingly cynical/sceptical about Rosie and Daisie, their timings and their methods – oh and what gets promoted from a wuff and what doesn’t. Agenda aplenty!”

    This time he might be on slightly firmer ground.

  43. The difference does seem to be mostly UKIP. Amazing still to see sudden change in tones by Lab/Cons from one thread to the next :)

    From last thread:
    @RiN, I wasn’t trying to say anything about any economic systems, just to cheer up someone being all doom & gloom & looming disaster. Yeah, in this country housing costs have gone all dodgy, but most everything else (including food) is better, we (everyone) do have a higher standard of living than in the 1970s.

    @HBTone, you say a bit less than a million, I say ~900,000 in 2012…sounds like agreement to me. :)

  44. phil

    wuffwuffwuffwuffwuffwuffwuffwuffwuffwuffwuff

  45. @Dave

    but with even as many as 60 relevant marginals out of 600 seats, and say an overall sample of 1800, those results will be based on a sample of about 180, with a consequent large random error
    —————————————————————-
    I guess it would be like the regional crossbreaks. We see wide fluctuations there, but once you start looking at trends like Statgeek does here I think it still provides useful information

    http://www.statgeek.co.uk/polling/polling-trends/regional-trends/http://www.statgeek.co.uk/polling/polling-trends/regional-trends/

    But of course you get the danger that by providing the data someone takes a single poll and writes a national news story like ‘Tories winning with the 18-24yr age group’ like we saw this week, when we know that is not true once looking at that cross break over a period of time and across polling companies….

  46. @nickp, Phil Haines

    The commentary centered around the “two clouds over the Brighton conference”… the Damien McBride one, and… “the bigger cloud is Labour’s standing, roughly level with Conservatives in recent opinion polls.”

    An interesting footnote… this chief political correspondent is the son of an ex-Labour MP, and the husband of “Osborne’s fixer and gatekeeper since he first became shadow Chancellor”. Until late last year that is, when she was replaced by someone described as “the BBC’s lead political producer.” Quite a merry-go-round.

  47. @Howard, @Dave, @Richard

    I consider that Richard’s suggestion has some merit, but in terms of YouGov using the tactical situation in seats for weighting purposes. That would purely be to eliminate random m.o.e. rather than to generate some crossbreaks which, although interesting, might still be of limited value.

    My point is that Gloucester and Cheltenham are both urban constituencies with reasonably similar demographics. At the last GE, Labour polled 35% in Gloucester and 5% in Cheltenham. The only factor that can explain that disparity was the tactical situation in the seats: Gloucester was a Lab/Con contest, Cheltenham a Lib Dem/Con contest. Gender, age, social class and newspaper readership would be hopeless at predicting such disparities. Given the clear tactical voting at play, even party ID could be expected to fall well short.

    So I suggest that weighting to get the right % of respondents for each of the different types of tactical seat would improve the accuracy of YouGov polling in terms of national VI.

  48. The issue for Labour isn’t so much the problem of Conservatives narrowing the polls, it’s more that they need to boost their own ratings. That is, they need to push their own ratings to 40%+ again, while also convincing the electorate that their Leader is capable of winning the election.

    The Tories (and the rest) will fluctuate their own scores, but the central issue for all parties is really about maximising their own support.

  49. @ John Murphy

    “This isn’t quite true but there was certainly a lot of over optimism about Republican polling”

    Possibly, however even the notorious Rasmussen Polls never had Romney ahead in enough battleground states to get an EC majority.

  50. @Phil
    …”At the last GE, Labour polled 35% in Gloucester and 5% in Cheltenham. The only factor that can explain that disparity was the tactical situation in the seats: Gloucester was a Lab/Con contest, Cheltenham a Lib Dem/Con contest….”

    The implication of that is that many anti-Tories voted tactically for LD in Cheltenham. These will now be disillusioned. The interesting question is what they do next time? Reverting to Labour would presumably not work, else they’d have voted that way in the first place. So what else will they do? Vote UKIP perhaps?

1 2 3