Tories take 11% lead with ComRes
February 25th, 2008
Did Cameron benefit from the Auschwitz row exposure?
A new poll by ComRes for tomorrow’s Independent shows a three point move back to the Tories since last month. The shares are CON 41% (+3): LAB 30% (nc): LD 17% (nc)
If the pollster was following its normal pattern then the fieldwork would have taken place from Friday until Sunday when one of the big political stories was the row of David Cameron’s Auschwitz trip press release.
My theory, as regular visitors will know, is that the Tories do better in polls if their leader is in the news at the start of the fieldwork period irrespective of whether it is positive or negative coverage.
The survey follows other polls last week which showed that the Tory lead was falling and the shares of 41% equals the highest ever level recorded by the pollster.
After last week’s polls which seemed to suggest that Labour was coming out of the Northern Rock crisis unscathed this new survey will come as a blow.
One factor which probably hurts Labour with the firm is the way the turn-out is calculated. ICM filters so it only includes those saying 7/10 or more and they are all given the same value. ComRes filters out those below 5/10 but then weights responses in line with the expressed certainty. YouGov does not filter or weight on the certainty to vote.
So the extraordinary diverse range of polling results at the moment continues.
MessageSpace Advertising
YEE HAR Con overall maj 200 guaranteed at next GE
Lab/LDs fighting for 2nd
Not bad, I still think we need to see a fair few more polls though!
Mike S, does the “more follows” include the appropriate number of smiling faces?
If band width becomes a problem, do what the Romans did…..
and substitute five DCs’ for a grumpy Gordon.
Its amazing how polls are varying. I wonder how much the (alleged) initial response to Northern Rock was that it had to be done, and now people are beginning to realise what a millstone its going to be. After all, the ability to borrow money is gone for good, the PSBR and the National Debt cannot stand it . Only alternative would be to do an Anthony Barber and print it, but that’s not prudent and highly inflationary. After all the recent criticism of DC and GO , I suspect this poll will be quite satisfying.
Good news for DC and bad news for GB.
However two things bother me;
1) it bucks the recent trend of the tories lead falling slightly. I’m not going to call this a rogue poll, it would make as much sense to call the recent ICM and YouGov polls rogue, and this poll is closer to the yougov than the ICM was. Instead i will say, the picture remains murky and it’s hard to tell what’s going on or find any distinctive trend.
2) no change in both Labour and LD figures, which would suggest that DC is picking up more votes because of certainty to vote increasing, which isn’t usually an issue for the tories. Or from nationalist parties, which I would expect to be experiencing a bit of a boom in light of the EU treaty.
Do we know further details, e.g. supplementary questions?
Obviouslt a good poll for the Conservatives, though it’s interesting that their share increase comes from neither the LDs nor labour.
Increases my suspicion that at least one polling company has got it’s techniques wrong. We could be heading for a SUSA versus Zogby situation at the next election!
re 6. Hopi remember Smithson’s law - “A rogue poll is one where you don’t agree with the numbers”.
Oh, absolutely - I’ve no idea which of the pollsters might be wrong, only that if it’s their techniques that are producing widely diverging results, we shouldn’t expect their results to stop diverging as we get closer to an election…
AND to add - a I say in the article the Tory polling figure is closely correlated to the amount of coverage that Cameron has been getting.
9
Yes Mike, and when it comes to the election, Three weeks of DC v Gordon on the Telly …….
Slightly off thread England v Ireland is on (beware) the ides of March (15th). I hope the RFU have had the sense not to invite Gordon to the match…. Punters should check before placing any bets. Gordons record in that respect isnt too clever…
9. lol. Keep at it Mike! Unfortunately the tenuous link will sooner or later not work I fear. Cameron’s mugshot will appear and they won’t do well in a poll. What then?!
Good poll for the Tories, but the increase only in terms of vote certainty takes the gilt off it.
Can we have these repeated images of DC in bright colours to make them Warholesque?
Well done Tories.
1 - “YEE HAR Con overall maj 200 guaranteed at next GE. Lab/LDs fighting for 2nd.”
If the Tories did have a maj of 200, they would have about 425 seats. Allowing 30 for NI and Nats, if the LDs and Labour were fighting for second, each would get about 95 - 100 seats. You have just made a LD happy, Ave It!
11- We don’t know that the Tory increase comes from certainty to vote- it could easilty come from smaller parties (or even i assume, decreased certainty vote by Labour voters, thou presumably that would lower labour’s share.)
6. The figure of 15% for others in the last ComRes poll was way out of line with what anyone else has been finding and probably was a bit of an outlier. That the drop to a still substantial 12% has benefitted the Conservatives alone is interesting, but probably masks more complex movements.
On a personal level, the overall numbers do chime with what I’m finding on the doorstep in canvassing in advance the local elections this May, though as I’m in what is predominantly a Tory-LD ward, it’s not really a microcosm of the national political picture, and although there is a lot of anti-Labour feeling, there has always been a lot of non-Labour voters.
NR will continue to run and run.
The govt should have appointed an administrator. They are well organised at selling financial assets in bite-sized chunks. And yes, they are used to selling all asssets at the bottom of the market.
NR is a bank. A bank needs customer confidence—it’ll never get any while owned by the govt. Nationisation will run it down much more slowly and inefficiently and expensively than an administrator. It will cost much more public money this way.
GB shouldn’t have talked to Branson–that’s the job of the chief administrator. It would have saved the govt money to have done the deal with him at any price. GB was worried that Branson would make a profit out of NR. Daft. GB#s resonsibilty was to minimise the loss to the tax-payer, not to worry about somebody else’s profit. GB failed again, and it wil be a running sore until it is unloaded.
The detailed data is already on the Comres website , it is interesting to compare the raw data with that from the recent ICM/Guardian poll last week which of course had a substantially different published result .
There is very little diference in the raw data , most of the difference in the published results comes from the different weighting techniques used by the two polling companies .
ComRes raw data Con 26.4% Lab 20.6% LD 10.5% Others 8.3%
ICM/GU raw data Con 24.6% Lab 21.1% LD 9.9% Others 5.8%
A Labour difference of 0.5% in the sample ends up after weighting as a 4% difference in the published Labour figure and a 0.6% higher LibDem % in the Comres sample ends up with a 4% lower published figure . Clearly unless we know whose weighting is correct we cannot decide whose published figure is the more correct .
16, This is really interesting. Mike any chance of an artcle on raw data comparisons and impact of different wieghting techniques? Obviously each of the polling companies have good reasons for doing what they’re doing, but it might be helpful to explain these in some detail (again?- this may have been done before).
12 - its all about the Cons encouraging the little parties - we love democracy!
17. Yes agreed. It’s probably been done before but I’d be really interested in something on raw data comparisons, and if at all possible how they compared to the actual GE last time round?
Re 7 - Feb Polls for comparison (OK I know you should compare like with like because of methodologies):
ComRes 25th C 41 Lab 30 LD 16 Con Lead 11
YouGov 21st C 40 Lab 34 LD 16 Con Lead 6
ICM 17th C 37 Lab 34 LD 21 Con Lead 3
YouGov 15th C 41 Lab 32 LD 16 Con Lead 9
Populus 3rd C 40 Lab 31 LD 17 Con Lead 9
ICM stands out with higher LD share & lower Conservative one. Otherwise Conservatives & LDs pretty stable, Labour moving around a bit within 30%-34% - doesn’t really bear out “extraordinary diverse range of polling results”
16. the other 40% of the population, are these non voters? or undecided?
Mike, i don’t think the Cameron thing is true, as leader of the opposition he is always in the papers and on the news, he was there on monday with northern rock as well, and that didn’t help him. However we would need to find an objective way of measuring his exposure, to prove either of us right.
Good figures, and more support for Mike’s theory too. How long can Gordon survive with Labour figures of 30%? - they booted out Blair when Labour slumped below 35%.
OT, but has anyone been watching the Portillo on Thatcher documentary that’s just finishing on BBC4? Thoroughly absorbing, and with most of the big hitters from the last 20 years of Toryism, bar Hezza and Major sadly. Even Portillo hasn’t come across as the irritating git he has of late. DC on good form too.
A good watch. Worth a catch on iPlayer if you missed it.
I can’t believe the Auschwitz row had much to do with it, it didn’t really have that much exposure.
I’m not one to normally challenge Smithson’s law. but this poll does seem to be out of kilter, we’ll see what the next one brings.
re 16. I think Mark is wrong to draw conclusions from the raw data. The main adjustment that pollsters make is to ensure that there’s a demographic balance and Mark does not take that into account.
The second adjustments that the main phone pollsters -ICM, Populus and ComRes - make are based on recalled past vote at the last election. This is to ensure that the samples are politically balanced.
The third adjustment is on certainty to vote as I note above.
In this poll there a nearly twice as many Tories saying they are 100% certain as there are Labour supporters. If ComRes adopted the ICM approach to voting certainty then my rough calculation is that the margin would drop to 8%-9%.
21 the other 30% are won’t votes/don’t knows/refused
21 G, How do you know for certain that it didnt help him? Thats only journo speak saying that… I suspect what politicians and the media think doesnt always transpose into what floating voters think ….
25. Thinking about how undecideds have caught us out before, in America, is this a problem?
The polls are wild at the moment. I sense the public trying to decide which lot they distrust least. I do find this result bizarre after what has been the least well-managed week for the Tories in months. Unlike Richard, I don’t find the Smithson theorem tenuous. Only Mark Senior tonight has so far identified another possibility.
22. Thanks for the reminder, I shall make a point of watching it. Thanks again.
26. YouGov recorded a slight dip in the tories lead. They may have been wrong, but by the same token this poll might be wrong.
It is worth pointing out that since March less year, when the firm had a major methodology change, ComRes has been the least Labour friendly pollster. This is based on the way they do the certainty calculation and the past vote weighting.
It is also worth noting that the fieldwork for many ComRes polls is carried out by ICM using the same facility (here in Bedford actually) that Populus use.
I don’t really think the polls are wild at the moment. Based on Ted’s figures, Tories are at about 40, Labour about 32 and Lib Dems 16 or 17%. I’d say these figures with a tory lead of 7-8% is what feels right.
re 21. I keep a daily tally based on a screenshot of the BBC politics page and give three points to the main story, two to the other highlighted stories and one to the one-liners. It is pretty good system of predicting polls.
At the start of last year IG Index ran a weekly spread market on YouGov’s BrandIndex ratings of politicians. I think it went on for twelve weeks and my scoring approach for Cameron, as outlined above, worked in all but one of the weeks.
This became quite a nice source of extra pocket money until IG pulled the market.
I don’t think the polls are all over the place at all, they mostly put the tories around 40%, Labour around 33% and Lib Dems at around 17%.
We don’t appear to be in the middle of any great change, just a frantic spinning of microchanges.
32 - Snap!
28. As Ted shows at [20], the polls do have a degree of stability in them, with the odd figure out of line with the others (not that that makes it wrong). Even so, a ‘poll of polls’ based on the median figures of the five Ted quotes would give: Con 40%, Lab 32%, LD 16%. I can’t believe that would be far off the mark.
35. Great minds. Shall we split labour at 32.5?
24 Looking at the raw demographic data , the 2 samples are not very dissimilar . It is strange though that Comres do always finf many more responses to Other Parties in their samples than the other pollsters ( this was even more so in their January poll ) presumably their questioning is slightly different to the other pollsters .
32/5. Double-snap!
33. Fair enough. Seems sensible.
Do you keep a record of the results? Could i see them?
Replying to comments on previous thread
Thanks everybody for all your comments, which are duly noted.
It seems to be a general thumbs up so I guess we just go ahead and organise things.
Richard (19) - The Podcast idea is very interesting but you’ll have to explain to me exactly how it works, and also if possible elaborate on your other ideas. Can you drop me an email? arklebar@talktalk.net
StJohn/David Kendrick/Jan from Norway/Paul M - Your interest in the syndicate is noted and we will definitely go ahead soon. There is another related matter in the pipeline which will be announced very soon and after that’s published, I’ll start putting the syndicate together.
Paul M - I think I have your email address but just in case I am confusing you with somebody else, please drop me a line to confirm.
I note your comments about the Brandindex problems but this is different. We would for the most part be investing in the main markets, where liquidity is fine.
More anon and thanks again everybody.
36 The trouble with a poll of polls is that it is averaging polls some of which must be wrong . ICM’s weighting techniques may be wrong in which case the poll of polls will have a LibDem figure too high or alternatively they may be using the correct weighting in which case the poll of polls will include lots of polls with a too low LibDem figure .
42. Do you feel that poll of polls result is right Mark?
20 Ted thanks for that interesting comparison.
ICM does look like an outlier. Is Com Res also an outlier at the other extreme and the real position a Conservative 9 point lead?
The Lib Dems look to be in the 16/17% level, which is 1% to 2% below last Feb. No Clegg bounce.
We are just 9 1/2 weeks to the local election day.
A couple of comments.
The Lib Dems must be disappointed with the small Clegg effect after 2-3 months. In December, some commentators were suggesting that Clegg would take a lot of support from the Conservatives. Has Dave seen off Nick?
I suspect Northern Rock is contributing significantly to Labour’s poor showing: the sums of public money at risk are enormous and the Government’s response has been consistently highly political.
41 PtP email sent
43 I think that figure is about right for a poll of unweighted polls . The real position hinges on whether weighting is a correct technique and which is the correct weighting .
44 HF as I have pointed out neither Comres nor ICM are outliers , it is the weighting that makes the published figures so different .
42. That’s why I used median rather than mean - to avoid rogue results having too much impact. The only thing with using that technique with such a small sample is that the Lib Dems end up on their lowest recorded score of the five, as they received 16% three times. At the risk of introducing a ‘fiddle factor’, there’s possibly a case to raise it to 17% in light of the other results.
46 Thanks PaulM.
re 42. Mark is absolutely right about so called “polls of polls”. Each of the pollsters ask their questions in a different way, apply different weightings and in many cases have different sample sizes. The fieldwork also took place on different dates.
Look at each poll and compare it with the last survey from the same pollster; then see if the trend has been spotted by other pollsters.
I want really really want to believe it was ‘Auschwitz’ - just to p*** off Balls. Anything to wipe the smirk of the face of the little….
New theory… everytime Balls makes a politcally strategic move, the Tories reach +11% in the polls!
PtP - Have left you a message on previous thread.
43/49 If you look at the Comres data for how people voted in 2005 LibDems lose 6 voters to Labour but gain 17 a net +11 , and lose 12 to Conservatives but gain 6 a net -6 and 6 to Others but gain 2 net -4 . On those figures alone a LibDem overall figure of 16/17% is much too low .
What could have happened is that when the initial news of about Northern Rock broke, people’s initial reaction was to give some sympathy to the governments positions. Now, after a few days of quiet refelection, a differant view is emerging. I’m a gig believer that you have to wait a few weeks for big events to filter through to the public mood - It could be that ComRes is going to the first of a new run of very bad polls for Labour. I think we’ve got MORI and another ICM to come bfore the end of Feb? It’ll be interesting to see what way they go.
51. Mike - that’s fine for finding trends, and of course trends are important both in themselves and in identifying what has caused votes to shift. What you can’t do in that way is find what the current position of the parties is. That said, you are certainly right to imply that pollsters past figures should be used to sense-check the current ones, or to put them into context.
3 “do what the Romans did…..
and substitute five DCs’ for a grumpy Gordon.”
Mike, this is a good idea.
When Labour starts hitting the low 20s, 20point leads will become a regular occurance. Dave is going to start dominating the pages - and would really upset Zanu NuLab. You need to be ‘impartial’
What could have happened is that when the initial news of about Northern Rock broke, people’s initial reaction was to give some sympathy to the governments positions. Now, after a few days of quiet refelection, a differant view is emerging.
I’m a big believer that you have to wait a few weeks for big events to filter through to the public mood - It could be that ComRes is going to the first of a new run of very bad polls for Labour. I think we’ve got MORI and another ICM to come bfore the end of Feb? It’ll be interesting to see which way they go.
I think that this poll is right. Comres have the right methedology. Labour party support is not going to crawl over 30%, not with Brown. The guy is a loser with a very large capital L. He is also going to help Livingstone lose in London against a prize joker.
Well done Gordon.
58 I doubt whether there will be another ICM poll this month , probably one in the Sunday Telegraph on the 6th March .
What odds 15 Daves this year from a major firm?
I’d go 5/2.
“I’m a big believer that you have to wait a few weeks for big events to filter through to the public mood”
A couple of examples;
ERM - The true damage wasn’t seen until late 1992 and early 1993, I believe.
The bottled election - Brown called off the election in early October, yet Labour didn’t reach their absolute low point until mid to late November, I think.
Initial polling, following some major event, is a waste of time, IMO. Much better to wait a week or two and see how people truely feel after reflection, before you try and draw too many conclussions about how something has played.
51 Agree that you need to compare each pollster with the last to get a trend, but looked back over the polls since October 10th - after the non-election. 38 polls, Conservatives 40% or more in 29 of them. Of the 9 with less than 40%, 5 were ICM, 2 Populus, 1 ComRes and 1 Ipsos-Mori.
I’d say Conservatives look pretty stable over last 5 months, volatility has been in Labour & Lib Dems.
52. Yes, I imagine part of this bounce was Balls’ use of the horror of the Holocaust to bash Cameron. It seems like the remnants of New Labour know the tactics of Blairism, without having Blair & Campbell’s insight of when to deploy them appropriately.
New AP/Ipsos Nomination and Presidential Polls -
Clinton 43% - Obama 46%.
McCain 53% - Huckabee 27%.
McCain 43% - Clinton 48%.
McCain 41% - Obama 51%.
http://apnews.myway.com/article/20080225/D8V1JDOG0.html
58. Thnaks Mark? We are due a MORI poll this month though aren’t we?
Northern Rock is only part of the picture.
Nothing in Britain is getting better. Today we read house prices are down 5th month running.
Teenagers are committing suicide - and its not difficult to understand why. Who would be a teenager today? Or the Social Services who kidnap newborn babies to satisfy adoption targets. Who would be a parent?
Or the British woman who spent the first 3 years of her life in the USA denied the right to stay in Britain to look after her elderly mother.
Or the British servicemen died because of Labour’s defence cuts. Who would sacrifice for Gordon Brown while he dispises the sacrifice?
Today I read the grim news about the Car Clamper murderer Levi Bellfield - another reminder that Police targets treat the crime of climbing a tree as equal to murder.
This is just a short list. It is the reality of life in Britain under Labour. Labour think they get away with it because each story barely registers any change in the polls…… But they add up. There really are a lot of bad stories.
62 - In the case of Northern Rock it’s not going to be a two week wait, that holds only for events that have an immediate effect/impact.
If Northern Rock starts to cause concerns it will be a longer term and more hardy depression of the government’s vote share, that this may happen is the fear that must haunt them.
I know I am repeating myself, but Brown is just such a loser, a sad risible, pathetic, excuse for a leader. Bad judgment, weak, vacillitating, bullying, cowardly, unprincipled, and just plain humourless, uncharismatic and sad. Nuff said.
67- bolted horse- I am afraid life would be no different under the Tories, and for the poor, and disenfrachised, much worse I would suggest.
Central probabilistic forecast (ComRes)
Con 331
Lab 239
LD 41
Nats 18 (assumes 13% SNP rise)
Oth 3
NI 13 (SF abstain)
Con majority 17
6-poll moving average
Con 298
Lab 269
LD 46
Nats 16
Oth 3
NI 13
Remembering that there is only a 1 in 4 chance that a particular poll will get a party’s score correct (to the nearest whole number), the picture since the Tory peak in mid-December 2007 looks as if the Tories have fallen about 2.5%, with the LDs rising by a similar amount, Labour remaining fairly static on about 32-33%.
http://www.titanictown.plus.com/bigpic.jpg
We should not be surprised if the next poll shows the Tory lead down to 3 or 4% It’s very nip and tuck in seats at the moment, similar to the pattern which prevailed during mid-2006.
Talk of a Clegg bounce was only ever talk. Ashdown and Kennedy got no bounce. I doubt Steel did either. Why would Clegg have got one? Does it really matter that he has not?
66 Should be a Mori poll this week and a Yougov Daily Telegraph poll later this week with a Populus Times poll next week .
68. Yes, I agree. NR could get a lot, lot worse for Labour. All along I get the feeling Labour have been complacent about the damage NR may do to them. That complacency only increased last week, when initial polling seemed to show the public mood was to offer Labour support. And thats really my point. Those initial polls are not always a reliable reflection of how the oublic truely feel.
73. Many thanks, as always.
I think Ted’s run down of the Feb polls at 20. above is most telling..There really is no signifiant divergence…in fact these polls look very stable with Cons just over 40% and Labour in very low 30’s. ICM seems to give Lib Dems some of Dave’s votes for some reason..
One other point on the recent ICM and you gov poll..as Mike pointed out they were both conducted over the two weeks when lots of the schools (certainly private and grammar schools) had half term week. Many Tory voters (like me) were away…this would explain the small drop in Tory support in those polls. Factor this in and in fact Feb polls are all pretty much the same with NR maybe having little impact at all..
Personally hung parliament must remain the favourite even at these numbers, unless there is a significant unwind of the anti-tory tactical vote or as I have mentioned previously a significant change in the ‘efficient’ distribution of Labour’s votes…Of course its probably a good chance that both of these things will happen to some extent at least..
An interesting report from the ground in Texas. Early voting indicates that turnout in some Clinton favoured areas is up between 50-100%. Turnout in more Obama favoured areas is up between 400-800% on 2004.
http://www.huffingtonpost.com/mayhill-fowler/ground-report-clinton-ca_b_88313.html
PtP - had a little good news today, so will be certainly contributing to the syndicate, and am as always in favour of jolity and social occasions at the NLC.
Mike - any news on the rest of that polling data that wasn’t being released? Are BPC taking strong action?
71. Do we seriously believe that if the Tories polled 11% more than Labour, on 41-30 at the next general election, they would only end up with a majority of 17? I know all the arguements about how the seats favour Labour and how the Tories are still way behind in the north, but honestly, I cannot believe thats right. Surely they would get a bigger majority than 17?
70 bolted horse- I am afraid life would be no different under the Tories, and for the poor, and disenfrachised, much worse I would suggest.
you could be right. The damage caused by Labour is so severe, Britain may never recover.
However, I remember things were better under the Conservatives and it is Labour policy that has caused the current decay. Open door immigration, “Positive” discrimination and “Positive” racism, Political Correctness, Dumbing down education, Employing unsuitable staff in the NHS, Cronyism, Institutionalised High level Corruption .
The repair needed is as a mountain to climb but the will to put things right is a start.
I think that like the teenager suicides, British people are facing dispair. Eventually they will blame Labour.
Martin is not going to help Labour. Mr Speaker isn’t going to get any better and so far, is still in charge of the expenses enquiry.
It’s another drip drip, making Gordon look weak.
If the public perception is that Martin is a bar to a cleaner Parliament and Brown won’t/daren’t push him out [for his own advantage/protection at PMQs…whatever]it will look like party interest before public. The harder it is at home, the more they will resent the ermin trimmings at the HoC.
79. I am sure others like Rod will explain why, but given the level of Labour support in Scotland, Wales and the North it does seem counter-intutitive that a 30% share would be ‘efficient’ enough to prevent the loss of huge swathes of southern and midland marginals..
71 The majority of 17 is due to the distribution of MPs. Labour Scotland and Labour inner city areas get more votes.
It is a classical pattern. Crap area votes Labour. Labour drives out the self sufficient classes to Conservative areas.
Conservative areas end up with more voters. Crappy Labour areas end up with fewer voters.
With a majority of 17, the Conservatives can redress the imbalance.
79. Yes, they would. It would be astonishing if that sort of movement in ‘positive’ votes did not lead to a similar change in the way people cast ‘negative’ votes i.e. anti-Tory or anti-Labour.
LA Times reporting that Mitt Romney MAY re-enter the GOP race!!
http://latimesblogs.latimes.com/washington/2008/02/romneyback.html
Didn’t the timing of the ICM poll take everyone by surprise? I think that Mike pointed out that it was brought forward by a week which brought it into the period of the half term break, it was also before the NR nationalisation?
82 and 84. Thanks guys.
I’m still learning about divying up seat allocation. Its certainly a complex issue….
69. Crikey Tyson, say what you think…. But you are a Labour supporter, are you not? I cannot see how it is either possible or likely that Labour will do away with Gordon before the election. His position at the top of the party is far stronger then Blairs.
80 - To bring two things together, people need hope, the idea that things can change. It was there in 1979 and 1997 but it got blown away because the ideas held out were only ever divisive, it was just that one side of the divide suddenly gained much more power over the other.
Empowerment of people to affect their own lives, power at a closer and more local level, fairness in representation, pathways to improvement and freedom from those who try and feed off the law abiding and hard working. That’s what people crave and maybe if it was posited away from an ‘us and them’ view of the world we might start to feel better as a nation.
82.There are Labour heartlands in Scotland that will remain solid for them come the next GE and under a future Conservative government, in much the same way as area’s down South did for the Conservatives throughout the last 10 years.
But Labour’s support has been slowly seeping away in other parts of the country in recent years. What will be interesting at the next GE will be the way those area’s vote, IF the result is not deemed a foregone conclusion and that increases voter turnout as I expect it too. I think with 4 party politics up here, we might just see more than a few surprise wins for various parties against the overall poll findings.
85. Certainly worth a punt, seeing as McCain’s contradictory statements could be the sign of a bigger hole.
But irregardless of betting, a Romney comeback would be wonderful for the USA! GOP backstabbing, a fix at the convention after losing the popular vote, and an overwhelming mandate for Obama and the Democrats winning seat after seat in congress.
Alas, it’s too much to hope for!
79. First of all, as I’ve said many times, you should never put all your eggs on one poll. There is a 3 out of 4 chance it will be wrong. There is scant evidence (at the moment) that the Tories are actually 11 points ahead. More like 6-7% in my view.
However, let us assume for the sake of argument they are 11% ahead. That would amount to a 7% swing. There are all kinds of fanciful extrapolations that could see the Tories winning X amount of seats for such a lead.
However, two historical patterns need to be considered.
i) Regional swings have NEVER made a significant impact on overall seat totals. The largest impact was a minuscule 5 seats, in 1992, and there is no evidence that I have seen which suggests the next election will be anywhere near as regionally heterogenous as 1992.
ii) The LibDems and their precessors have ALWAYS performed a bit better than UNS predicts when their vote share declines.
That leaves us with
i) systematic tactical voting against Labour.
ii) less efficient distribution of Labour’s vote.
I’ve seen no evidence for the first. If anyone can point me to such evidence, I’ll gladly look at it. ii) would imply a differential *increase* in Labour’s turnout. Again I’ve seen no evidence of this.
No-one can say for certain what the overall result will be next time. For the moment, it is not unreasonable to continue to forecast using UNS. Anything else is just wishful thinking unsupported by evidence.
92. Thanks for such a extensive reply.
I just find it amazinf that a party could poll 11% more than their leading rival, yet only get a majority of 17 seats. But if thats genuinely where we are, than thats where we are.
92. is by RodCrosby…
64. Could also be a sign that people actually want to think the best of Cameron. If something sounds dodgy, they listen and give him the benefit of the doubt.
Tonight Portillo said something along the lines of…..policies are almost irrelevant, people want to know you and think you are an OK bloke and so can trust you.
They will never love another after Blair, but perhaps even liking a politian is a big deal in today’s climate.
85 I suggested the other day that people might want to get themselves some Romney insurance. It may be straying into tin-foil hat territory, but if I were spending as many of my own millions on getting elected as Romney has, I would probably have spent a few more digging the dirt on my opponents - and then releasing it at the most opportune time…
93. Yes, I agree. It is amazing. As amazing as the 66 seat majority on a 3 point lead and 36% share Labour got in 2005, not to mention their results in 2001 and 1997.
Face it, the system is significantly skewed against the Tories, and has been more or less for the past 35 years. Labour’s general weakness in the 1980s concealed this trend….
93. It doesn’t end there though does it? How does a Govt continue without the popular vote?
It has been done before, but only on very small margins and in different times [when people didn’t back chat their doctors and repected politicans].
I can’t imagine a situation where a Govt have a working majority and are 5 to 9 points behind in the polls. I suppose it depends how you define ‘working’.
93 - if the GE % shares were as in this poll then IMHO then the Tories would have a much larger majority than 17, more likely in the 60-80 range.
Watched Portillo’s program on BBC4 - a lot of old footage but somehow worked very well in the way it was all put together - well done MDX!
Almost feel that the Tories are like a grand prix rookie driver leading a race for the first time, and have a hold what we have approach, the only problem is that it’s as though it’s 1-0 with nearly all of the second half left - I think the tentativeness that the opposition front bench showed over Northern Rock has highlighted this.
What I want to see from Cameron is more of an exposition of the decentralisation narrative that he has hinted at but hasn’t in my opinion really put the flesh on the bones to yet plus an emphasis back on some policies in pursuit of social justice ie reinforcing the narrative that the Tories have changed and at the same time concreting this view in the minds of the public.
80 correct. labour will lose because they have failed on every level. and they will lose big.
Re 69, Tyson, “I know I am repeating myself, but Brown is just such a loser, a sad risible, pathetic, excuse for a leader. Bad judgment, weak, vacillitating, bullying, cowardly, unprincipled, and just plain humourless, uncharismatic and sad.”
I hear he speaks highly of you
Hi guys!
98. Personally, I think the next election will be the most interesting since 1923, and the cockeyed result will be the death-knell for FPTP.
Tories ahead by a country mile in England, narrowly ahead in votes overall. Labour comfortably ahead in seats in a hung parliament….
Surely this has to be Hillary’s last throw of the dice. Pretty desperate stuff:
http://www.timesonline.co.uk/tol/news/world/us_and_americas/us_elections/article3433929.ece
Hang about. It’s not so much Tory constituencies having a higher population that means the seats won maths is so far out of line with the national vote share maths, as I understand it. It’s Tory constituencies having a higher turnout.
New CBS poll:
A new CBS News/New York Times poll finds Barack Obama with a 16-point lead over rival Hillary Clinton among Democratic primary voters nationwide.
Obama, coming off 11 straight primary and caucus victories, had the support of 54 percent of Democratic primary voters nationally. Clinton had 38 percent support.
In a CBS News poll taken three weeks ago, shortly before Super Tuesday, Obama and Clinton were tied at 41 percent. Clinton led by 15 points nationally in January.
http://www.cbsnews.com/stories/2008/02/25/opinion/polls/main3874915.shtml
The figures for 106 are -
Clinton 38% - Obama 54%.
McCain 46% - Clinton 46%.
McCain 38% - Obama 50%.
http://www.cbsnews.com/stories/2008/02/25/opinion/polls/main3874915.shtml
I LIKE MIKE…
http://www.youtube.com/watch?v=xvSXpM5qGmg
What a nice guy…
106/107. MM and PJ. I think March 4th will see game over for Team Clinton.
104. PfP. I agree. I think we are up against it now with our Clinton VP bet! You can lay off to small stakes on Betfair at 22/1.
If Roger looks in, thanks for your excellent Oscar tips. I managed to back the few losers you identified because they were the longer, more speculative outcomes and only to small stakes. But I had a decent bet on Marion Cotillard for best actress. Great spot by you! You clearly felt she should have been the odds on favourite so 5/2 against proved great value. Well done to Roger!
108. RodCrosby. LOL!
How’s your Titanic chart looking with the ComRes poll? Iceberg problems?
110. The Tories are still afloat, but the pumps are losing the battle against the sea…
Re 108, Rod Crosby, “I LIKE MIKE…”
He does come across very very well, and as a likable chap. He also clearly has a sense of humour and does not mind getting a laugh out of being the but of the joke. If he were not a nutter I would have backed him.
112. Could he run again?
111. Very good!
If I were you I would sneak on to one of those lifeboats while there’s still time.
Re 113, Rod Crosby “Could he run again?”
I suspect he could.
I am pleased Ralph Nader is running again, I am hoping to see the Republicans tighten a bit.
For the record, the Anthony Wells swingometer (which I understand also uses UNS) gives a Con majority of 38 on the ComRes figures.
Even with a very small amount of tactical unwind it’s easy to see there could well be a Con majority of at least 60 on these figures.
Rod - I know we’ll never agree on this so I won’t waste time going through points made previously. But I would be interested in the answer to one question - why does your forecast give a different result to Anthony Wells when you are both using UNS?
Tom Bower in the Daily Mail, As usual the Left is utterly wrong. This isn’t about class, but simple corruption Ouch!
113. He could, and may well do. But his support will always be limited to evangelicals. The rest of the Republican party simply does not believe in having government services any more, let alone extending them to those lacking a decent education or healthcare coverage.
116. Three reasons, essentially.
i) I employ the Curtice and Firth “probabilistic” UNS that was used to good effect in 2005 to forecast the 66 majority for Labour. Straight UNS assumes the distribution of the marginals is uniform. 2005 showed they were significantly non-uniformly distributed, which accounted for most of the Tories perceived “bonus” over straight UNS. Of course we now have new boundaries, but the probabilistic methodology preserves the “even-handedness” to the parties in considering the marginals. Using Wells’ notional figures, straight UNS gives the Tories 342 seats for an 11 point lead. Probabilistic UNS reduces that to 337.
ii) Simple UNS (either probabilistic or straight) takes no account of the Nats performance, while as we know the SNP are roaring ahead in Scotland. In other words, under simple UNS the Tories are forecast to win seats like Angus and Perth & N Perthshire from the SNP, and Argyll & Bute from the LibDems, whereas with any significant SNP rise such forecasts become silly. Factoring in the estimatated SNP performance (+13% on 2005, conservatively) reduces the probabilistic Tory seats from 337 to 334.
iii) The LibDems have historically always done better than straight UNS indicates when their vote share declines. “What we have, we hold” could broadly sum that up. Of course the LibDems will lose a lot seats if their vote share went as low as 17% and the Tories as high as 41%, but my best estimate is they would end up on 41 as opposed to the 36 forecast by straight UNS. Of the 5 extra holds, three would be held against the Tories, two against Labour. Those three extra LibDem holds would reduce the forecasted Tory seats from 334 to 331.
Actually, a simplistcic calculation would imply a Tory majority of 12, but since the 5 SF members will likely continue to abstain, bringing the size of the effectve new House down to 645 (instead of 650), the Tories 331 members would constitute a 17 seat majority over all other parties…
But like I said, I don’t believe the Tories are currently ahead by 11% anyhow, and it’s daft to single out a favourable poll to the exclusion of all others…..
119. Thanks Rod. I wasn’t saying I thought they were 11% ahead either - I was just interested to know the differences between your forecast and Anthony’s and thanks for such a comprehensive explanation.
Points ii) and iii) are easy to understand. But could you clarify point i). Re the Con “bonus” in 2005 - are you basically saying the Tories “got lucky” in 2005 - ie they got what look like randomly higher swings in “the right places” to win a few extra seats? And that next time the “risk” is that they’ll do a bit worse than UNS?
For the record, inputting these numbers into Wells gives Con seats of 344, as opposed to 342 in your post. Obviously only a trivial difference so not worth pursuing.
120. On the last point, I’ll follow that up. Anthony has made minor adjustments to his notionals over the past year or so, so it’s possible my figures are out of date. Or more likely there are some rounding errors going on somewhere. I notice that if I increase the Tory lead fractionally to 11.1 instead of 11.0, I also get 344 Tory seats on simple straight UNS.
Re point i) John Curtice explains what may have happened in 2005 in “From Votes to Seats”
http://www.
britishpollingcouncil.
org/
john-curtice.pdf
(heck knows, but the spambot does not like that URL!!!)
The relevant bit starts at page 4, final paragraph “the third possibile reason…”
Yes you are correct.. The Tories got lucky in 2005, in that the swing obtained coincided with a point on the uneven marginal distribution that was likely to deliver more seats than straight UNS forecast. Likewise on the new boundaries an 11% lead would see them do slightly worse than UNS predicts. Ironically, there are other places on the distribution where they would do slightly better - around a 5-6% lead for example. However, none of the current differences are as extreme as existed in 2005.
If you are a math nut, Curtice and his co-forecaster for 2005, David Firth, explain the “Route to 66″ in the following paper
http://www2.warwick.ac.uk/fac/sci/statistics/staff/academic/firth/papers/curticefirth.pdf
Page 9, “Modelling votes and seats”. Specifically, I use the formula shown on page 12 to arrive at a probabilistic forecast for each party.
Of course the final, and probably most significant part of their overall forecast were the “regression equations” based on exit polls on 05/05/05.
Naturally, no-one can estimate those until the next exit poll, and therefore ALL simple forecasts are a bit moot. Nonetheless, for good or ill I try to chart a course that is as “central” as possible, based on the currently available data.
92 Rod - Interesting stuff. Couple of Qs
i) re “systematic tactical voting against Labour”
Wouldn’t a reduction in anti Tory tactical voting have a similar effect? (especially with a higher overall Tory vote)
ii) re: probablistic UNS. Imagine this scenario
2 seats (A,B) Tories need 4% swing
6 seats (C,D,E,F,G,H)Tories need 6% swing
UNS 5% says all seats swing by 5% and hence two Tory gains (A, B)
Probabilistic says average swing is 5% so if following
A 3% B 6% C 4% D 4% E 5% F 5% G 6% H 7% then average and median swing would still be 5% but B, G and H would fall so three Tory gains. Is that it ?
119 Rod
I assume that any betting markets on the majorities are computed on the basis that Sinn Fein MPs can vote against the leading party ?
Yokel etc Do Sinn Fein have any prospect of winning additional Westminster seats (North Belfast?) Is Eddie McGrady of the SDLP standing again in South Down? I assume the Foyle SDLP man (Durkan) is unlikely to lose to SF.
122.
Taking ii) first. Essentially correct in your simple example, but the calculation would in fact be probabilistic, so the winning probabilities might look like this..
P(A) = 0.75
P(B) = 0.75
P(C) = 0.25
P(D) = 0.25
P(E) = 0.25
P(F) = 0.25
P(G) = 0.25
P(H) = 0.25
The expected seat total would be the sum of these probabilties, i.e. 3.0, but of course no particular seat would be forecasted as won with certainty, and the distribution would be a lot smoother than in this example.
See the URL references at 121 for real examples.
Re i) Yes, tactical “unwind” broadly equates to new “anti-Labour” tactical voting. It can be considered one-dimensional, although strictly on 3 distinct axes [Con-LD, Lab-LD and LD-Lab]. Problem is, the great tactical-voting manifestation occurred bewteen 1992-1997, and to a lesser degree 1987-1992…
In a nutshell, if it was going to “unwind”, the best opportunity was in 2001, but it never happened…. I think since 1992 about 10 million people have died-off the register and another 10 million new people joined, so it appears the scope for serious unwind has passed. Throw in another set of significant boundary changes, dwindling turnouts, people moving around, immigration and the potential for false memory, and the waters are muddied further. I mean, of those who are still alive, howe many actually remember how they voted in ‘87 or ‘92, to be able to recall what they are supposed to be “unwinding” from or to? I just can’t see any reversal being on the same scale as the original phenomenon, and in any case I’ve seen little evidence in either by-elections or locals that it may be happening at all (unlike the period 1987-1997.)
123. I’ve asked this question a number of times, but no one seems to have offered an answer.
What constitutes a hung parliament?
Conventional wisdom says…blah blah… if Labour wins 326 seats it’s a majority of two.
But.. if SF abstain it’s a majority of 7.
But.. if the SDLP hold two seats it’s a majority of 11…
In other words Labour could still be in a technical majority on 321 seats, not 326.
There are even more complex scenarios, such as if Dai Davies holds Blaenau Gwent but decides to take the Labour whip. Then Labour could survive on 320…
Thanks very much Rod.
I tend to agree with you on the tactical unwind. My view has been that in the marginals the choice is do you want a Tory government or not. If you do, you’ll actively vote Tory, not move between LD and Labour and let it happen by default. I appreciate that last time there were seats like St Albans where the Tory won due to a Lab-Lib swing, but in an election where there is a realistic chance of Tory Prime Minister, I don’t think that will happen so much. That’s why I focus on the absolute Tory vote % and the increase from last time, rather than the leads.
127. The big unknown about ‘tactical unwind’ is that we don’t know if it has, in effect, already taken place and been accounted for within the polling figures. Or not.
With the introduction of STV in local government elections in Scotland last year, though, we now have some real data of voters’ 2nd preferences.
The evidence from Scotland is that the Tories are the least transfer-friendly of the main parties. The Lib Dems and the Greens, where they stood, being the most transfer-friendly.
125. With Plaid Cymru in coalition with Labour in Wales, presumably they can be added to the Labour column too?
109. Stjohn. Well done on Cotillard but I’m disappointed you didn’t splash out on the final eight suggestions I gave you on Sunday. I was hoping you’d have swapped your Villa season ticket for a corporate box by now.
124 - As the original inventor of the phrase “Tactical unwind” I would dispute your reasons for rejecting it. The phrase was never meant to literally describe a voter who in 1997 etc would have voted, say, Lib Dem but voted Labour to keep the Tories out and has stuck with them ever since making their vote still eligible to be unwound.
On the contrary it was merely intended as a device to moderate the inadequacies of UNS, based on its effective assumptions of a population that stays in the same place, will vote consistently, and transfer their vote uniformly (or at least that any such changes outside of these assumptions will cancel themselves out). As such the possibility of “tactical unwind” is no more challenged by, for example, people dying off, than is UNS itself. Tactical unwind is merely a way of modelling movements in vote-seat conversions that UNS is incapable of picking up.
Tactical voting, IMO, has always been best established by the “two party preferred” question that some pollsters use. With necessary allowance for actual Labour->Tory transfer I don’t think this has changed much in general pattern in the run up to any of the last three elections, heavily favoring Labour. Hence there remains scope for “tactical unwind” as long as there remains the possibility of the two party preferred question moving in the Tories favour (as i said with necessary allowance for real Labour->Tory shifts).
Much as I’d like this to be accurate I don’t think it is (take that, Smithson’s Law!:P).
The polls are far too diverse right now. I reckon the actual Tory lead is what it has been for a while now, 5-7%.
PTP,
I have left a response on the previous thread.
YOU ARE KIDDING YOURSELVES.
WHEN CAMERON HAS A BAD WEEK — COMRES ALWAYS COMES TO HIS RESCUE. HOT LINE TO ANDY CAULSON I HAVE BEEN INFORMED.
THE NEXT ELECTION IS 2 YEARS AWAY.
So shouts the Brown bunker. Are you also implying that COMRES are breaking the rules. If so, where is your evidence?