A new poll by ComRes in Tuesday’s Indy tells a completely different story to ICM and MORI’s recent figures. The topline figures, with changes from ComRes’s last poll, are CON 37%(-6), LAB 36%(+4), LDEM 17%(+5). The poll was conduced between Friday and Sunday.

In recent weeks we’ve become used to polls showing differing results – methodological differences have resulted in some pollsters showing much larger Tory leads than other ones. However, until now polls have been consistent in showing the same trends – until last week the pollsters, despite differences in extent, were all showing the Conservative poll shrinking. The two polls published since the press reaction to the PBR, conducted by ICM and Ipsos MORI, both showed that trend reversing, with the Conservatives again gaining support. This poll is not just out of line with the figures other companies have been presenting, but also it’s the exact opposite of the trend ICM and MORI are showing.

Normally when polls disagree I look at the methodologies and try to explain the difference. There is no obvious explanation here. I’ve looked through the ComRes tables and there is nothing obviously freaky or wrong – the Lib Dems appear to have been weighted to a much higher figure than their previous poll – 12% of the sample said they voted Lib Dem in 2005, as opposed to 8% of the sample in ComRes’ previous poll – which goes a long way towards explaining the jump in the level of Lib Dem support, but not that of the two main parties: the overall shares of recalled past vote are pretty similar to those ICM use to weight their polls.

This is the first poll conducted since Damian Green was arrested, so theoretically it could be possible that the public reaction to it has been the polar opposite to the media’s, but that would be unusual. It could also be that the the increased Conservative lead we saw from ICM and MORI was just a reflection of the bad press coverage the PBR was receiving at the time those two polls were carried out, and the public reaction to the PBR now it has sunk in properly is actually more positive. The alternative possiblity is, of course, that either MORI & ICM, or ComRes are “rogue polls”.

I am cautious about the term “rogue poll”. It tends to get thrown around willy-nilly against any poll people don’t like and sometimes at companies with methodology people see as faulty. What is actually refers to is that the 3% margin of error commonly quoted for polls is at the 95% level of confidence. In layman’s terms it means that 19 out of 20 times the “real” figure will be within 3 points of the figure quoted in the poll. A rogue poll is that 1 in 20 where the figure is more than 3 points out. It is inevitable that these things happen, and happen to all companies – but realistically it is impossible to ever be certain whether a poll that looks out of line is a rogue poll, or the start of a new trend. We won’t know for sure until we see some other polls that confirm or contradict this one, but until then I will urge my normal caution against polls that show large changes in support, or sudden reverses in the trend.

195 Responses to “ComRes show Tory lead down to a single point”

1 2 3 4
  1. I am off to look for my passport

  2. At first sight looks like very good news for Brown.

    The key question is when did the field work take place. If after the Green affair broke then a very interesting result, if before it may already be old news.

    Looking more closely the apparent sharp drop in Tory figures looks less convincing, it is more likely that the previous Com Res poll was an outlier because disregarding that poll, Tory vote has been consistent around 40% with Com Res since the summer.

    More surprising perhaps is the sharp increase in Labour vote, which is out of line with both ComRes past results as well as more recent surveys, I would view that one with caution until more confirmation or otherwise is recieved from further polls

  3. Now this will set a cat amongst the pigeons.

    There now seriously needs some looking at the methodology for the pollsters as some are showing a small lead and other are showing a big lead.I cannot imagine, and I may be wrong, the public is changing its mind every day.Without looking to see if any economic questions were asked it does beg the question was the PBR as badly recieved as the press have had us believe.As Peter said on the other thread has Greengate(I hate all this gate stuff,Greengate is actually a place in Oldham) benefited Labour.Maybe the public have got a bit worn down with the negative overkill on the PBR and the governments part/non-part in Green and is getting some sympathy.

    Maybe the lack of any economic alternative by Osbourne is filtering through,?

    As I said in the last thread, that if after the 15% lead poll the next one showed a fall for a pollster that was then showing a Conservative lead in double figures, this would make it even harder to fathom the current trend. I didn’t think that the lead would more or less disappear all together.

    Until the next poll, which on current polling random form will show the Conservatives about 25% in front,
    the game is back wide open again.

    Meanwhile over in Oracle World they are putting the coffing labelled Labour back into the cupboard.

  4. I’m not going to be down-hearted by this one poll. I think getting hung up on one poll to the next is risking ‘wood for the trees’ syndrome. Will see what the polls over next 7 days or so and if then it’s back to single figures generally, then I’ll get my bags too.

    PB posters are reporting Lab ahead of Con in the SE in this poll and surely that raises questions on the accuracy of this poll (for me to cling to anyway)…

  5. If you missed it on the other thread Guido Fawkes a right wing forum posted the above results as:

    Conservatives – 48%
    Labour – 25%
    LDems – 15%

    Nice to see that they are on the ball.

  6. Yes that figure for the South East is startling, and I’d feel the same way about any poll which showed the Tories in the lead in Scotland or the North East. We just have to wait and see the next batch of polls to see who’s right and who has the egg on their face!

  7. Scrapheap – there are somehow more Conservative posters on PB than here and many of them post on Guido as well.

    Must of been one of them that was wishful thinking the ComRes result on Guido.

  8. If Guido himself posted it well go over and laugh and point fingers. If one of his commenters did – so what? Every open posting forums attracts some people who are idiots.

  9. Anthony – I wouldn’t let my eyes venture over there, I’m staying loyal to here.

    I only picked up the Guido trash result off PB.

  10. Scrapheap & Impartial Observer – strange things in regional breaks aren’t necessarily a sign of something wrong.

    Polls are weighted to be representative of the country as a whole. They aren’t weighted to be representative within regions. Therefore, in theory a sample could have too few supporters of one party in one region and too many supporters of that party in a different region, and still be fine overall. As I said in the main article, the proportion of 2005 Tory voters in the sample seems to be fine.

  11. Anthony. Is it possible for any poll to show – for Wales and SW – Cons 41% LD 7%.? Have I misunderstood something about the way polls work in the regions? the figures look totally corrupt.

  12. Wow!

    I notice that this poll was taken over the weekend and when the weather was particularly cold – could such things have an impact on the kind of people who would respond? (Some Labour supporters may scoff at such a question, please feel free to do so.)

    In any case, I think it gives added significance to Yougov’s next poll. It’s certainly going to be a nail biting time.

    For the Lib Dems this is another relatively good poll. And the Lib Dems are likely to gain most from a Hung Parliment.

  13. Anthony. Correction – the Con figure is apparently 51%! Looks even more absurd.

  14. oh well, there’s that hope dashed by Anthony. Thanks for the explanation – why do they bother with breakdowns like that then if it’s only the whole that is representative and not the parts?

    I’m still in Victor Meldrew mode on this poll – at least for now. I don’t believe it.

  15. Comres figures for Scotland are;

    Labour 37%, Tories 17%, SNP 34%, Libdems 4%, Other 8%

    Those figures don’t seem that unusual for Scotland and In line with what i’d currently expect labour back ahead, the SNP a strong second, with the Tories falling back a bit. The only odd one is the Libdems who at 4% have struck a new low in a long line of lows for the first time falling below the vote for others.

    As I have said before these figures are only usefull for general trends so leaving aside the Libdems score if there is a problem with this poll it doesn’t seem to be in Scotland.


  16. The real question?
    When is the next poll?
    I just don’t buy that people are changing their minds so often.

    The fact is the Tories have been above 40 for ages, except once before with com res and I just don’t buy into the idea that they have forgotten the debt figures because they got a penny off a Mars bar.

    Nor do I think Green will help Labour, hurt the Tories.
    None of the narrative fits with a major change.

    The figures and the trend look off.
    When’s the next one?

  17. Collin – if random chance gave ComRes a really strange and freaky raw sample – if, for example, they found loads of 2005 Tory voters in the South-West and hardly any in the South-East, it couldn’t necessarily produce dodgy figures overall.

    Just because there are odd regional breaks the overall GB figures wouldn’t *necessarily* be out as long as the overall political make up of the sample was sound. That said, you’re much better off with a raw sample that isn’t strange and freaky.

  18. ‘That said, you’re much better off with a raw sample that isn’t strange and freaky.

    I feel a potentially partisan comment comment coming on.

    Is it the case then that we can expect a rogue 1 in 20 times [not pointing a finger at the moment, well waving it about a bit perhaps]?

    It will make for some strained nerves for all concerned in the run up to an election.

  19. Any chance that Conservative voters are trying to tempt Labour into an early election?

  20. At least this poll puts a bit of fun back into it. PB forum seems to have exploded and the regional votes in the actual poll look a bit freaky, however with the closeness of the poll there should be plenty of posts on here tomorrow. Would love to see the face on some of the posters when they read this in the morning.

  21. Anthony

    If we look at 99% confidence intervals, do the margins of error change significantly? I’m asking about polls in general as well as this one.

    For instance, if there’s no overlap between the 99% intervals for current polls showing 10+ points for the Tories and this one, at the 99% confidence level, then surely there’s some methodological problems for one of the polling companies?

    Or am I wrong? My statistics knowledge is pretty poor, I admit.

  22. These results are completely baffling.

  23. this poll isn’t out of line with the last yougov poll though – and don’t yougov do the fieldwork for comres?

    there may be some issues with the yougov panel. they may select an appropriate group with an eye on demographics but mori’s finding over the behaviour of public sector workers was interesting – there may be hidden imbalances in the yougov panel

  24. Of course Anthony is right to say that we can’t read much into 1 poll. The Weighted Moving Average is now 41:34:16, that is a reasonably “safe” figure and unlikely to be out by more than 2.5%.

    Where I must disagree slightly with Anthony is in this “3% MoE” figure, which depends on simplistic statistical assumptions which are not the case in real life. Of the 32 ComRes results that have been published since I started tracking this, 4 (12.5%) have been >5% from the WMA and 2 >6%. The empirical Standard Deviation of the ComRes results is +/-3.2% which is slightly higher than the other pollsters. An error in ComRes of just over 6% is not unusual, and indeed their poll on 19/9/08 turned out to be under-estimating the CLead by 6% as well.

  25. @ Tony Jones – “Would love to see the face on some of the posters when they read this in the morning.”

    My face looks like this :O and :(O)


  26. As I have said many times the regional subsamples in these polls are too small and not individually weighted and therefore do not give meaningful results .
    The Comres data tables are particularly useful in that they break down how people say they voted in 2005 by the regional subsamples .. From this in this poll it can be seen that LibDems in Scotland Midlands and especially the South West were considerably under sampled in this poll and LibDems in the South East oversampled compared to the real votes in 2005 .

  27. Anthony, can you suggest any reasons why both polls (say the Guardian’s and this one) could be right?

    I can think of two, not unrelated. One is that public opinion really is this fickle, so even a correctly sampled poll with slightly different research dates can give hugely difference answers. Another is that public opinion is quite fickle, as is those saying they are likely to vote (is that filtered in both), and together this means the choices of those who actually intend to vote are very changeable.

    I suppose a third explanation is that both polls are at the extremes of their margins of error, which would just about give you a swing of 12 points in the lead, wouldn’t it?

  28. Amazing-
    Big movements from last ComRes Poll 12/13 November are (+/- Con/Lab)

    Females -9/+4
    18/34s -10/+19
    25/34s -16/+5
    ABs -12/+6……..eh???
    DEs -13/+17……these aren’t typical Public Sector are they?

    South East -9/+1….??

    ….so young women; the upper middle class plus the working class-and people in the South East.

    Cameron had better pack it in then!

  29. This s a bit of a stunner. Reading the infromation..is it me or does it go what was stated in the last two polls. I think I will wait until the next poll out (isn’t there one this sunday?) before we can either describe this as a complete crashing of Tories (as a liberal I am happy that we had a slight rise), support or this poll is indeed a rogue. However may I argue that the ‘green’ afair should have no factor. Both Nick Clegg as well as Tony Benn all argued for Mr Green, so if the tories went in for overkill on the affair..then as Nick Clegg and the liberals are then guilty by association and sghould also be showing a decline in support…

  30. James,

    “My face looks like this :O and :(O)”

    Have you considered surgery…..


  31. Also, although it does seem unlikely the pubilc’s view of DG’s arrest is so much different from the media, if the public believe MPs have cushy lives and are not subject to the same rules as the rest of us (which they do believe) then I suppose a lot of influential people moaning about MPs being subject to police action might not go down well.

  32. These wild fluctuations obviously have their reasons. Not just the polling sampling method used. Is there any chance for instance that support for Labour/Conservatives et al varies radically from one part of the Country to another but are not being adequately incorporated into the polling methods?

    As we know Scotland and Wales for instance are not typical of England: and northern England may be radically different from the south. Other reasons are obvious including local strength of say a BNP candidate.

  33. @ Matthew – the ComRes poll was conducted “between Friday and Sunday”. Green was arrested last Thursday and the story didn’t really hit the media until Friday. The BBC initially played it down until Sunday (when it became clear that it wasn’t going to go away, and the Mumbai story was fading, so they started paying the Green Affair a bit more attention). Ditto The Guardian/Observer. So I suspect that the timing and a slow, understated BBC response (particularly) meant that the story hadn’t gathered a full head of steam yet when the ComRes poll was taken.

    If there is a Green effect, my feeling is that it’s more likely to show itself towards this coming weekend.

  34. There’s only one way to know for sure, call a general election. It may be easy to discount this poll, but not long ago the Tories were down to a 3% lead. It seems that they are either 10 points up or about dead even, depending on .. well, whatever really.

    On a different note, wasn’t it nice to see Ed Balls coming down hard on the officials at Haringey? I’m not a great supporter of this government but he showed much needed decisive action.

    Doesn’t make me want to vote for them, but it’s nice to see some people in government are capable of action.

  35. A bit radical, Mark! The polls seem at odds with one another….tell you what, let’s have a GE – just to settle an argument!

    Seriously, i sense some logic in a rise for the LibDems and a drop for the Tories.

    The public response to the PBR seems to be “not enough!” It’s hardly going to endear them to Osborne if he’s saying “irresponsible, all have to be paid back”, which can be interpreted as “too much!”.

  36. A rogue poll is one so they say where one does not like the results. Well I don’t like the result of this one but I am reluctant to call it a rogue poll as others of my persuasion are too ready to do.
    For the last month or so there has been a clear dividing line between one set of polls ( not necessarily the same polls) saying that the Tory lead is rapidly shrinking and another set which says it is pretty solid. Maybe all this conflicting evidence tells us not so much that the public are fickle but that due to the the financial turmoil this autumn voters are not thinking ahead to the next election . Instead they are focused on the short term and how to get through the week. Perhaps the pollsters should agree not to release any polls until February next year in the hope that things settle down one way or the other.
    All this up and down stuff is not good for my heart!

  37. Wow, not sure what to say… someone is going to be looking at their systems once the election is over. It was once pointed out that there seem to be a large number of people who just vote according to what they have read in the papers that day…. are people really that volatile?

    This has made it all very exciting though!!

  38. Morgan – at the 99% confidence level the margin of error on a sample of 1000 is 4%. Though I wouldn’t get too hung up on the specifics – margin of error calculations are from statistical formulas based on a genuinely random sample with a 100% response rate, they don’t really reflect real life.

    Johnny H – no, YouGov don’t. ComRes is a phone poll, and the fieldwork is normally done by ICM.

    John Charlesworth – it’s unlikely. At a regional level polls will be representative. They will always contain the right proportions of people from Scotland, the North, the Midlands and so on. At a lower level than that they could theoretically have something like too many urban and not enough rural people, but there is no reason at all to think they have, or obvious reason why they should.

  39. The poll is the poll.

    I for one would really appreciate the chance to vote sooner rather than later, so hope that this does (trick?) the current encumbent into an early election.

    I would like to see (and please point me in the right direction if it already exists) what the strength of feeling is behind the answers that people give. I think the results would be startling.

  40. Nick: ‘A rogue poll is one so they say where one does not like the results’

    I would suggest it is a rogue poll because it shows large swings for no apparent reason and is out of step with most of the other pollsters. This poll is both.We should wait and see what other polls suggest before we use the ‘r’ word but it does look it.

    One thing I’m certain of though is that Labour’s friends in the media will be spinning this poll for all it’s worth, then more.

  41. Anthony .

    Some people are discrediting this poll on the grounds that some of the regional subsets look out way of whack. For instance the South East figures show a Labour lead, I believe.

    I thought that the pollsters don’t pretend to have representative samples by Region and provided the national sample is representative then the poll is perfectly valid in its own terms. So there may be too few Tories in the South East but too many in the North. The margin of error on these subsets must be very high. I saw one poll which found only 6 Tories in Scotland for a 9% vote share whilst another one found a 21% vote share within a couple of days.

    I note that the percentage of people in the poll who say they voted LD in 2005 is only 12%. Higher than 8% certainly but still incredibly low relative to their 22% vote share at the last GE. I thought MORI were the only major pollster not to vote on past voting record. However, if Com Res are doing so they’re not making a very good job of it.

    Personally I think this is one of those 5% of the polls which is outside the margin of error but then I’m a Tory! The only thing that occurs to me as a possible boost for Labour is that voters are giving Labour’s VAT cut the entire credit for the massive price cuts which are being heavily advertised everywhere. Add these to the cuts in tracker mortgage rates and to the cuts in petrol prices and if you’re in work there might be a little bit of feel good factor running up to Xmas.

    By contrast the Damian Green affair may be a bit inside the beltway for a lot of people especially given the Mumbai, Baby P stories running alongside it.

    As ever we await the next polls!

  42. “there might be a little bit of feel good factor running up to Xmas. ”

    Indeed, as I myself will be driving 800 miles or so and watching my mortgage go through the floor.

  43. I just noticed that on the 26th October Comres produced a poll which showed the Conservatives on 39 while all others around that time were showing the Conservatives on either 42 or 45.

    This poll is contrary to the trend of the last two polls it has been pointed out. And the fact the Comres has had a recent history of being out of step with others should cause us to treat this one with some skepticism.

  44. This poll fits in with the trend over the last few months but not the trent (or blip) in the last two. Could it just be that there was a lot of churn over the last week that has now leveled off? But the first poll for a long time showing the Conservatives below 40%, if this continues will we have an election early next year?

  45. that should have been “…but not the trend (or blip) shown in the last two polls.”

    Interesting point made in the Independent:

    “There are other signs that natural Labour supporters are returning to the fold. The number of Labour “identifiers” who say they will vote for the party has risen from 81 per cent to 87 per cent since last month.”

    Could this explain the big change that is taking place?

  46. Glad I’m not a politician trying to work out for each poll if it meant I still had a job if it was the real thing. I’d be giddy.

    Seriously I think the way the polls have moved over the last year suggest something about the electorate-perhaps less voter loyalty than ever before which is leading to all parties coming out with increasingly attention grabbing policies rather than well thought out policies.

    However the rapid spin of policies is making the electorate even less committed (or more cynical) and so increases their willingness to alter their minds because they don’t actually really believe in politicians any more. which then cause politicians…

  47. The public love firm governments who show their firmness by arresting the badies! And who better to arrest than the Tories? Finally, they’ve got the right people! What a shame they can’t hold Green for 42 days without charge, that would be even better!

  48. It may well be that the last three polls are all indicating actual voter shifts that are not incompatible but have been confused by sampling differences. It is clear that Labour has picked up significantly in Scotland, and I suspect also among traditional supporters in England. The PBR tax changes played better to these groups I suspect. The Tories are probably strengthening among southern voters, where the PBR debt issue and Green affair played badly. If opinion is getting more polarised then presumably sampling he polls could simply be accentuating these different but valid trends?

  49. I’m not really sure this poll is unbelievable, but I think we need to adjust our conclusions of what the polls are showing.

    The recent tax debate in the PBR was described as opening up clear water between the three parties and various different issues (security, social care, civil liberties, economics etc) are clearly affecting different interest groups in different ways, so I would therefore say that the different polling methodologies are identifying the different trends which are at play and that no current poll shows the true state of the parties.

    I think the polling companies better start pulling thier fingers out to account for these differences because the next general election is definitely shaping up to mark a once-in-a-generation choice about the direction for the country.

    I wish to repeat my advocacy for a plotted graph of poll results and my opposition to the WMA.

  50. This is a very eye-catching result, no question. The headline figure for the Conservatives is the most surprising since over the last two-and-a-half months and 25 polls only 2 have shown them under the 40% level (both interestingly were ComRes polls).

    I do find it hard to believe that the Conservatives have fallen 6% when the two preceding polls have suggested that their support has actually risen. I very much doubt that public opinion has changed that dramatically in a single week when press coverage has been fairly minimal.

    However, we must wait for the YouGov poll before labelling this as an outlier. If the YouGov poll shows a Conservative gain then I think we can conclude there has been a gain of some kind. If not then I have no idea where we are.

1 2 3 4