The funny mathematics at ICM
June 28th, 2005
-
Have people really forgotten how they voted already
In their latest poll ICM had Labour with a 7% lead compared with the 3% that real voters gave the party in the election seven weeks earlier. Fine you might conclude - the ructions over the leadership are clearly causing problems. However from the detailed data from the survey, now out, it’s possible to draw a different conclusion.
On May 5th, as we all know, the vote split in Great Britain was LAB 36.2%: CON 33.2%: LD 22.7%. Yet from my mathematics it seems that ICM, was working on a split of LAB 38.1%: CON 31.1%: LD 22.25%. when doing the calculations for its latest survey.
The challenge for any phone pollster is that it takes 5-6,000 unsolicited randomised phone calls to find 1,000 people who are ready to be be surveyed and those that do invariably are more pro-Labour than the population as a whole. This problem is not new - it’s just that those who respond to such calls tend to be this way.
So to ensure that their samples are representative ICM and other pollsters do some complex mathematics - they weight the data in accordance with what happened at the election and then make an assumption that a proportion have forgotten what they did.
-
And for reasons that I cannot fathom ICM seem to think that even after this short time there’s a big difference between the number of people saying they voted Labour and those that actually did do.
The notion of interviewees overstating past vote intention for Labour might have been correct in the aftermath of the 1997 and 2001 landslides but I question whether it’s right in the current climate. It will be interesting to see how the other pollsters deal with this.
Before May 5th I wrote several articles about the phone pollsters’ number crunching and placed a four figure sell spread bet on the Labour vote share which I said was inflated because the markets were believing the pollsters. At the time I said I was “putting my money where my mouth was” and was rewarded with big profits.
Mike Smithson
MessageSpace Advertising
Mike, it’s not ‘less people’ but ‘fewer people’ Fewer is for number and less is for quantity as in ‘less sugar, fewer lumps’. I’m shocked beyond measure that a distinguished servant of the University of Oxford needs to be told that.
What’s that you say?
All right I’ll get on with my work.
The small proportion of people who are able/willing to answer unsolicited phone polls clearly have somewhat different opinions to the whole population. But even if five-sixths of the population would never answer such a poll, at least that still leaves a good 8 million potential respondents. I’m still far less comfortable with the main alternative, a self-selected internet panel a small fraction of this size, holding lots of data on each member but with no guarantees of accuracy or even freedom from blatant manipulation, as Mike S has proven.
Long may ICM keep doing their “funny maths”, as long as they’re transparent about it, because it’s the best we’ve got.
I’m Confused if ICM belie4ved were allowing for more people voting Labour than did, doesn’tr that mean Mike if you are riught and peopler are not being coy that ifa anything ICM are understating Labour Support?
Will there be any Cheadle only Polls? How did those Constituency only POlls make out in the General Electionaq? Anglesey was wel out i recall.
If there are, I wouldn’t set too much store by them. Also in by-elections there can be quite a sharp drift of voting intentions between the poll and the election (cue Mike and me arguing over how inaccurate the initial Hartlepool polls were).
However, I think Cheadle will be about GOTV rather than significant switching between parties.
5. Re:Cheadle.
When will we know the number (and names) of all candidates?
Will UKIP stand? and the Greens (they didn’t stand last month, so I suppose they won’t stand next month too)?
4/5 , David/book value. As Cheadle is an interesting marginal there may be a poll or two . However I’m always cautious about constituency polls. Their track record is mixed . However they are very useful to the challenging party who often use them to get tactical votes to oust the principal party .
Here’s some information about the effect of boundary changes on Tyne & Wear.
The region currently has 13 seats (all Labour) and will lose 1 - the controversial cross-river seat of Tyne Bridge. The 2 Gateshead wards of Tyne Bridge will go into a new “Gateshead” seat, which will incorporate parts of the present Gateshead East + Washington West seat.
Benwell + Scotswood, strong Labour areas of Tyne Bridge, go into Newcastle Central, which will make that seat safer for Labour. It may be that 2005 was the LibDems last and best chance to take Newcastle Central.
Newcastle East begins to look a bit more marginal, as it may take on some of the LibDem-leaning parts of Newcastle Central. Newcastle East is also likely to lose Labour-voting Wallsend to Tynemouth, which will please Alan Campbell, whose majority was halved in 2005. Tynemouth would be a lot safer with Wallsend in the fold.
The most fascinating scenario is that boundary changes in Sunderland could create a Labour - Tory marginal seat. A Conservative MP representing Sunderland? Good grief!
7, “However they are very useful to the challenging party who often use them to get tactical votes to oust the principal party .”
- yes, though I’m not sure this is all that relevant when two opposition parties are facing off against each other and there has already been significant tactical voting in a recent GE.
8 - does anything happen to N Tyneside, John?
We have discussed the witchdoctory that pollsters use to get to their final issued figures before and they seem to be making more and more adjustments every time . What does puzzle me a little is that they still claim their polls are accurate to within x% whereas that is surely only statistically true if the fiddle factors are not subject to a greater chance of error than a pure sampling factor .
As Michael says the alternative of an Internet based sample has drawbacks also though contrary to his preference , in my opinion at this moment in time their track record is better than the Telephone Polls .
An accurate poll can be conducted by consulting the same people at the GE and then some time later as you can compare what they are thinking now with known past behaviour but this has its limits as the more times you consult these people , their behaviour will deviate from that of the population as a whole .
8 - this is interesting John. What about Blaydon? The Lib Dems got closer there last time, and, looking at the map on the boundary commission site, it seems the constituency may have gained some extra bits of Gateshead borough in the south, which I think is where Labour is weakest. Gateshead borough in general is gentrifying as people are priced out of Newcastle - could this be the Lib Dems best bet next time?
David at 3:
As I understand it, ICM make 2 major weightings to their raw figures:
Firstly, they attempt to bring their responses into line with the general electorate. For example, in this latest poll, they got preferences from 262 men and 289 women. They adjusted that to make 273 “virtual” men and 275 “virtual” women - that is, a male preference was weighted higher than a female preference.
They got 45 answers from 18-25 year olds, and weighted them to become 55 “virtual” 18-25 year-olds, etc.
Secondly, they weight the figures to adjust for unrepresentative responses from any of the parties (eg, Labour voters being more likely to respond) by asking how they voted in the last election. This is susceptible to “false recall” and outright lying, unfortunately (people don’t like to admit they didn’t bother to vote, or like to say they supported the winner).
These two weightings changed the response from:
153 Con, 218 Lab, 134 LD, 46 others (unweighted - real people)
to a “more representative”:
173 Con, 208 Lab, 125 LD, 45 Others (weighted - “virtual” people)
Compare these with “how they say they voted”:
152 Con, 236 Lab, 143 LD, 46 others (unweighted - real people),
or
178 Con, 218 Lab, 133 LD, 45 others (weighted - “virtual” people).
They then assign the “don’t know/refused” to the various parties for their final scores, and adjust for their “spiral of silence” - people trying not to say that they voted for an unacceptable party. This tends to have the effect of adding a point or two to the Labour final score (as the ICM theory is that “Bashful Blairites” outweigh “Shy Tories” and “Closet LibDems”)
Thus the final score is heavily processed. ICM have spent a lot of time and effort coming up with these processes, but it is useful to see where the assumptions are, so we can second-guess them.
Mike’s point is that from the raw unweighted data, the number of Tories went up from 152 to 153, the number of Labour voters went down from 236 to 218 and the number of Lib Dems went down from 143 to 134 (comparing how people said they voted to how they say they’d vote today - bear in mind the “false recall” tends to mean that fewer people say they voted Tory and more people tend to claim they voted Labour).
With all that in mind, an increase in Labour lead from 3% at the election to 7% from this poll seems a bit - well - overprocessed
8 Itake it you’re being ironic with Mr Campbell?
Explain on Sunderland?
Any changes worth worrying in Wales? Mr Wells was only reporting the new Seat of Aberconwy may be made a tad more tory friendky by the reomval of Bangor anything beyond that?
“What does puzzle me a little is that they still claim their polls are accurate to within x%”
Actually Nick Sparrow of ICM doesn’t claim that. ICM’s line is that the margin of error does not relate to accuracy, it is merely a measure of variability - polls with gross errors in their sampling techniques would still be within 3% of the “correct” value for 95% of the time, it’s just their faulty methodology would mean their “correct” value would be wrong.
Imagine two pollsters asking the same question, which has a true answer of 50%. Pollster A, who has accurate methodology, would, with sample sizes of 1,000, be within 3% of 50% 95% of the time. Pollster B on the other hand suffers from sampling bias that gives an answer that is two points too low - that said, if he too has sample sizes of 1,000, his answers would be within 3% of his “correct” answer of 48% 95% of the time.
1) Nick - perhaps that is why Mike has moved to York. They dont bother with such niceties in the Morrison queue!
“Less lard, fewer pies”
13.”Any changes worth worrying in Wales?”
I read PC could lose a MP and be reduced only to Elfyn Llwyd and Adam Price.
17 They only have two now don’t they after the loss of Ceredigion and not taking Anglesey?
14 - I follow your logic Anthony , but think the logic of Pollster B faulty as his correct answer of 48 is not the real correct answer which is 50 .
19 - yes. Strictly the margin of error is not accuracy at all, but precision. You can be precise but inaccurate if you get the wrong answer with a high degree of repeatability.
8 - Sunderland South was a Conservative seat up to 1964
18. Now they’re 3 now:
Caernarfon (Hywel Williams)
Meirionnydd Nant Conwy (Elfyn Llwyd)
Carmarthen East & Dinefwr (Adam Price)
21 - ???? So were many seats they now have no chance of winning .
A question about the ICM poll - why was it based back to 2001 GE voting intentions rather than 2005?
John at 13 is incorrect re Tynemouth. Tynemouth is not adjacent to Newcastle East and can’t therefore gain any wards from that seat.
The original Boundary Commission proposals would see Tynemouth gain the Valley ward from North Tyneside. Valley is a Labour ward, but not overwhelmingly so. The effect of this would be to make Tynemouth slightly better for Labour, but still very winnable for the Tories.
Newcastle East would lose Wallsend and Northumberland wards to North Tyneside, and gain one of the Jesmond wards (and possibly others, can’t remember details) from Newcastle Central. That would make Newcastle East a better prospect for the Lib Dems.
Because of that, Nick Brown, MP for Newcastle East, made a counter-proposal which would see Newcastle East retain Wallsend ward and North Tyneside retain Valley ward. Tynemouth would then be broadly unchanged. There was a public enquiry last week.
Because Nick Brown’s proposals make Tynemouth a better seat for the Conservatives, the Tories supported Nick Brown at the enquiry, to the disgust of the other CLPs in Tyne & Wear!
We are now awaiting the Boundary Commission’s final proposals.
Mark - ah, but Pollster B doesn’t know he’s wrong. He thinks Pollster A is wrong
Andy - ICM’s weighting was based on recalled vote from 2005, not from 2001. They have apparantly just forgotten to change the blurb on the website.
26 True but Pollsters A and B cannot both be correct and yet because this a problem to the polling industry as a whole , it is surely damaging to their client’s confidence in the accuracy of the forecasts . Although being cynical perhaps 1 or 2 of their press clients do not mind the inaccuracy as it gives them the opportunity to manufacture false and misleading headlines based on polls taken from faulty samples and then adjusted by well meant but possibly inaccurate formulae to give inaccurate results - LOL
25 Geordie Dave. CLP?
While we’re on about boundary changes, I’ve been looking at South Yorkshire - by my reckoning a net loss of three for Labour - here’s my working:
The new Sheffield Central is a totally different beast from the old one, taking in Walkley (from Hillsborough) and Broomhill and Crookes (which mostly used to be in Hallam), together with Nether Edge and Central from the old Sheffield Central constituency. These were the Lib Dems best areas in the old seats, and this could be a quite safe Lib Dem seat with Labour a distant second. Fifteen years ago this seat might even have voted Conservative – it contains some of the best housing in the city – but tastes in this area have moved away from the Conservative party. This seat will probably be one of the top five student populations in the country – Sheffield’s students were previously split mainly between Hallam, Central and Hillsborough, but these new boundaries should bring them all together.
Losing Broomhill and Crookes to Central is good news for Sheffield Hallam’s Tories, as these were the best areas for the Lib Dems. The centre of this seat has moved slightly outwards, curling slightly more around the city to the south and taking in a few more of the villages like Bradfield in the Peak District to the west, and the tone of the seat has correspondingly got less Bohemian and more suburban and semi-detached. If Conservatism in Sheffield is ever to revive, this seat couldn’t have been drawn much better for them. This will be one of Labour’s worst seats in the country as the other two fight it out.
The other interesting seat is Penistone and Stocksbridge, which takes good chunks of the old Sheffield Hillsborough seat together with the majority of the old Barnsley West and Penistone seat. However, despite the fact that both seats were Labour, very little of this is good Labour territory. The town of Stocksbridge is an old steel town outside Sheffield which historically voted Labour locally but, it is surrounded by rural territory and now sits in a much larger ward of Stocksbridge and Upper Don which voted Lib Dem at the last election, as did the other two Sheffield wards in the seat, Ecclesfield East and Ecclesfield West, which are suburban north Sheffield. These have been joined by the small towns and villages around Penistone from the Barnsley west constituency – this is territory with a much more rural outlook which still often votes Conservative. On balance, though, I would back the Lib Dems to muster the anti-Labour vote here next time.
Sheffield’s other three seats, of course, will remain safely Labour, but this will be of little comfort as they have corralled all their strength together. Sheffield / Barnsley used to be 8 (safe) Labour, 1 LD; my guess now is 5 (safe) Labour, 1 (safe) LD, 1 LD / C, 1 LD / Lab.
The rest of South Yorkshire, I reckon, will be unaffected i.e. 6 Labour.
These changes still haven’t been approved of course, so are subject to changes.
Sunderland south may well have been Tory eons ago. Oldham and Bootle Liverpool and Manchester have all had Tory councils in the dim and distant past. But the tories don’t have a hope in Sunderland (or Oldham, Liverpool, Manchester or Bootle) these days.
John C - is the review of Hallam so radical that Nick Clegg could choose between Hallam and Central without being accused of a chicken run, or is he essentially stuck fighting a weakened Hallam seat?
31 - Probably about 60% of New Hallam will have been in Old Hallam, whereas about 35% of New Central will come from Old Hallam (difficult to be more accurate because ward boundaries have changed significantly too). So he could fairly easily go for Central.
I suppose the problem with Central though is that the electorate there will not be the same as last time - student seats have very transient populations and need to be won afresh each time, whereas at least in Hallam he knows he has a core a Lib Dem voters. He also has a personal vote in Hallam which he would not be able to call on to such an extent in Central.
I still think Central this time would have been a better bet for the Lib Dems; whether he chooses to go or not depends on whether next time around the Lib Dems fancy their chances more against the Tories (in which case he’ll stay) or Labour (in which case he’ll go).
Interesting, John. I had thought until now that Nick Clegg was very well placed as he is spoken of as a future leader and appeared to have a rock solid seat. Now looks as if he will have to make a difficult choice and do some hard work to cultivate his base. If I were him I would choose Hallam as incumbent advantage and the danger of appearing to be running scared (perception in the party for his personal ambitions as much as anything) probably outweigh Central being more attractive on paper.
David @ 28. CLP - Constituency Labour Party.
John C could you just summarise those seats with who you expect to win over the 2nd. I think you wereaying tories for Hallam over LIberal Democrats, but a Summary would be most helpfula.
James
I think sheffield lib dems are pretty bullish about the two sheffield seats, and optimistic about the third. I think it is in the nature of these things that Clegg will be seen as the incumbent in whichever he seat he eventually stands. I assume that that will be Hallam.
31, Re Clegg and the choice of seat
Any other potential running chicken? I read here that Charles Clark could go Norwich North (if the not so young Ian Gibson will retire) if the Libdems will still be powerfull next time
36.”I think sheffield lib dems are pretty bullish about the two sheffield seats, and optimistic about the third. ”
4 years in advance?
It seems everyone is taking for granted that the Libdems will do well again in 2009. After improving their number of seats for 3 elections, they could even start to lose some seats.
I think there’s less complacency than you think, Andrea, but Sheffield is probably not high on the risk list.
35 - Had this election been fought on the propsed new boundaries, I would expect the following Con / Lab / LD percentages to be something like this:
Hallam: 43 / 11 / 45 (i.e. barely LD over Tory)
Central: 18 / 36 / 46 (i.e. strongly LD over Lab)
Penistone and Stocksbridge: 25 / 35 / 38 (i.e. narrowly LD over Lab)
Please bear in mind however that I’m not Robert Waller; this is based on half an hour’s thought with a pencil and paper, and the only thing I have to base my feelings of how the vote in the constituencies at the election was split geographically are 2004 council results and a general feeling based on such unscientific things as where the flags were and which areas feel like Labour areas etc.
39. I was not referring to Sheffield in particular, but in general. I read here talks about ousting Frank Dobson. I’m not saying it’s impossible, but it’s not a sure thing that the Libdems will have a strong performance in Hoborn & St Pancras again.
btw who will be the Labout MPs who could lose their seat in Sheffield.
40 - I would agree with you on that, the new boundaries will help Dobson apart from anything else.
40 - The Lab MPs at risk are Richad Caborn (Central), Angela Smith (Hillsborough) and Michael Clapham (Barnsley West and Penistone)
However the boundaries change in Sheffield , the Conservatives have virtually nil chance of gaining a seat at the next GE . Their sum total of councillors is 2 out of 84 in 1 ward out of 28 .
42 - Mark, the reasons I think the Tories have a chance in Hallam next time (on the proposed boundaries) are as follows:
1) The seat is moving from rather trendy, Bohemian areas where public sector people live and vote Lib Dem out of conviction (like Broomhill) to more suburban, semi-detached areas (like Beauchief) where much of the Lib Dem vote at council level is to get Labour out.
2) They will lose the student vote. In the old Hallam 2 and a half wards out of 4 were student-land, in particular Broomhill. In new Hallam it will be one ward out of five (Ecclesall) – and ward boundary changes have made that less studenty than it was.
3) The last remaining Conservative councillors in Sheffield are from Dore and Totley ward, where not even the ignominy of third-party status on the council can discourage them from voting Conservative.
38. To the Tories perhaps, and i am not predictiong many. But two things stick out the currenmt positioning of the Lib Dems public sees wel to Left of Tories now, used to be seen as the cuddly one nation party of unghappy Tory Wets no longer. Kellner on radio poiinted out that may bode well in Labour SEats but not in Tory ones, that’s why in my opinion the escape of Oliver Letwin may prove a seminal moment in Topry LIb Dem battles, not taking his seat a classic prize at this stage makes it very unlikely do so in future. Conversely they are now veru well placed in many Lab Seats see thread abovea.
Look at Cardiff. Cardiff Centralk fell on a laerge tACTICAL tOREY vOTE TO THEM. Newport East is now vulnerable, Swansea West, as well as Cardiff West and Cardiff South and Penarth etc other SEats up the Country. I think the Lib Dems may surprise people next time still, but at the expense of Labour rather than as has been traditional at the expense of the Tories.
John C p[lease see my post at 35. Cheersa.
re 1. My old English teacher will be turning in his grave! I’ve changed it.
43 - But even in Dore and Totley the Liberals took 1 of the 3 Council seats and were not far off taking the other 2 seats so even with the new boundaries there is 1 ward which is a Con/lib marginal slight bias to Con and 5 where the Lib Dems outvote the Con .
Of course a massive national swing to the Conservatives could make Hallam a possible gain but that is not saying it would have been marginal at the last GE on new boundaries .
44. The Libdems are now seen as on the left, but in 2009 it could be different if the vision of Nick Clegg and the ones like him will prevail.
Now the key point is to understand what type of voters left labour in the various seats where the Libdems got very strong swing to them.
If they’re leftwinged voters disillused by Blair, they will come back to Labour if the Libdems will move to the right.
They could get vote from the tories easly if they move to the right, but they will lose some from the left back to Labour.
47 - But if the 2004 local election results were mirrored exactly in Hallam at the GE, the Tories wouldn’t even have got the 35% or so that they did get. (I think?)
45 – David – every time I try to reply to your post nothing happens. I’ll keep trying.
35 - Had this election been fought on the propsed new boundaries, I would expect the following Con - Lab - LD percentages to be something like this:
Hallam: 40 - 18 - 42
Central: 18 - 36 - 46
Penistone and Stocksbridge: 25 - 35 - 38
Please bear in mind however that I’m not Robert Waller; this is based on half an hour’s thought with a pencil and paper, and the only thing I have to base my feelings of how the vote in the constituencies at the election was split geographically are 2004 council results and a general feeling based on such unscientific things as where the flags were and which areas feel like Labour areas etc.
40 - The Lab MPs at risk are Richad Caborn (Central), Angela Smith (Hillsborough) and Michael Clapham (Barnsley West and Penistone)
37 - I thought that Labour now effectively ban chicken running through internal rules (due to the damage it did the Tories in 1992-7). Even if the boundaries change, so long as there is a new seat with (something like) two-thirds of your old seat in it, you have to go for that one or nothing. On that basis, Clarke can’t chicken run as I believe there are changes to the Norwich seats, but not tremendously radical ones which could justify him moving across to North. I may be misinformed on that internal rule. Does anyone know?
35 - Had this election been fought on the proposed new boundaries, I would expect the following Con - Lab - LD percentages to be something like this:
Hallam: 40 - 18 - 42
Central: 18 - 36 - 46
P. and Stocksbridge: 25 - 35 - 38
Please bear in mind however that I’m not Robert Waller; this is based on half an hour’s thought with a pencil and paper, and the only thing I have to base my feelings of how the vote in the constituencies at the election was split geographically are 2004 council results and a general feeling based on such unscientific things as where the flags were and which areas feel like Labour areas etc.
40 - The Lab MPs at risk are Richad Caborn (Central), Angela Smith (Hillsborough) and Michael Clapham (Barnsley West and )
49. Maybe the infamous Spam Trap. Mr Smithson can you help?
RE Andrea Unless someone other than CK takes the helm now it will be too late to change perceptions which lag always behind poliv=cy by the time of the next Election. A curious mix of anti-war but also people who liker Tories not quite able to leap to Labour even the in the 90’s weren’t able to do the reverse thios time. Look at the Seats i mention and the thread above, i think the Libs regardless of a NEw Leader have by closing to within firing range this time have serios Capacity to hurt Labour in Many Seats Next Timea.
Finally worked out why I couldn’t post! Someone’s computer doesn’t like the full name of the Barnsley West constituency!
50. To the favoured sons there is always flexibility i’m sure, any boundary CXHanges could provide a good “knock on ” argument against that.
44 - if the Tories continue to flatline (and that’s a big if), then there will be desperation to get Labour out in 2009/10. That could see some more marked tactical voting on the “anyone but Labour” basis that, in reverse, did for the Tories in 1997.
49 - They only polled 29.8% in Hallam at the GE John and yes you have to allow for the Lib Dems vote being higher in local elections but even so they would still have been well ahead in new boundary Hallam at the last GE .
48 - Andrea - you must not fall into the trap of thinking the majority of voters in this country think in terms of left and right . It may have been true 30 years ago but not now . The majority of voters do not think about politics as a whole very much between elections except when a certain issue comes along and affects their lives . It is also a mistake as some do to assume as some do that everyone who votes for you at an election belongs to your party , there are quite a number of voters who vote contrariwise at successive GE’s . I in fact gave a personal vote in 2001 to my then Labour MP and so voted Labour for the first and possibly only time . I did not vote at all at the last election as I had moved house but Labour did not lose my vote they never had it really in the first place .
53 - being a rather subtle yet highly rude dig at the PM.
51 - I imagine a fair few Labour voters would switch to the LDs if Hallam became more marginal.
52. I don’t know the boundary changes for the seats you mentioned. But if changes are not significatif and Labour will lose all of them, lab election results will be a disaster.
Mind that in 2009 Blair won’t lead labour again. He’s so divisive: he is able to attract voters,but to lose them too. SO labour could lose some voters (who voted it only for Blair), but gain back some others (who left labour because of Blair).
56. You’re right People don’t think much between Elections, and true again they don’t think necessarily as sharply in Left-Right Terms as thirty years ago, but that makes the general hazy sense of where they judge a Party to be on the Political compass allthe more imortant as it informs their Starting point for that brief period at Election timea. I think you may have to concede Kellner has a point the current Public Perception of the Liberal Democrats is likely to be far more helpful to them in Labour SEats than Conservative ones.
I don’t know about disastrous Andrea it may not be an across the board ewffect, but there are more than a few seats vulnberbale to tthe Cardiff Central effect, where the Libs are now sufficiently close that marooned pockets of Tory Support determoined to inflict maximum damage on Labour can in seats where the Tories have no CHance bridge that gap to put the Liberal Democrats over the top.
I take your point re Blair but i don’t think that even GB would recapture all the losty ground, and once the Lib s start pointing out to trapped Tories in Seats where Labour used to have mountainous Majorities that their votes can make a difference suddenly if they vote tactically form interesting things may Start to happen.
Conversely the Libs Current Political Positioning AS Kellner of YouGov pointed out did them no good in formerly key Tory held target SEats like Letwin’s and May’s in the South.
61. but in Cardiff Central the libdems were 2 points behind, in Newport East they’re more than 20% behind with the tories less than a point behind the Libdems.
“I take your point re Blair but i don’t think that even GB would recapture all the losty ground”
he won’t take back all the lost ground like he won’t lose all the “middle England previous tories voters”.
Re 61 - it’s a fair enough point, but conversely the Lib Dems position did them no harm in Taunton or Solihull.
As far as I remember the last You Gov ‘political compass’ poll showed somthing along the lines of…
New Labour -3
Tony Blair +5
The British people +7
The Liberal Democrats -18
Charles Kennedy -11
Gordon Brown -27
The Conservatives + 38
Michael Howard + 43
(where + 100 is the extreme right wing and - 100 the extreme left wing).
Currently the Lib Dems are positioned closer to the political centre than the Tories - will a non-Clark Leader be able to change that?
On Welsh boundaries.
There are basically no changes except in North West Wales.
In North West Wales Labour is probably the winner.
The new Aberconwy seat (gaining Nant Conwy from Meirionnydd losing Bangor to Arfon). Will be a very tight Conservative/Labour marginal at Westminster and a threeway marginal at the Assembly.
The new Arfon seat (the same as the Caernarfon seat but losing Dwyfor to Meirionnydd and gaining Bangor) is probably Labour at both Westminster and the Assembly but could be close in both with Plaid Cymru.
The new Meirionnydd Dwyfor seat (losing Nant Conwy to Aberconwy and gaining Dwyfor from Caernarfon) Plaid Cymru will rack up a huge majority, but since this was already their safest seat it is basically stacking up wasted voted for them.
There could be a bust up in Plaid Cymru about who fights the Meirionnydd Dwyfor seat - Elfyn Llwyd is the sitting MP but Hywel Williams grew up and receives his strongest support from Dwyfor which is joining the seat. Were this to happen it would doom their attempts to save Arfon (which is unlikely).
60 - You have a valid point there A , but probably more important than whether Lib Dems are viewed as left or right is the fact that after successive defeats the Conservative vote that is left is pretty much near its rock bottom level so very much harder to make more inroads into whilst the Labour vote being the governing party is somewhat easier to make inroads into .
62. Sorry i think i may have been thinking of Newport West i think someone mentioned have a look, but Swansea West was something to Consider tooa.
sIXTY Four. The Tories may find the Vale tougher to take now it loses Sully Ward to cardiff South and Penarth which as a by prod may shave Michael’s or his Successors Majority bgy a couple of K, bringing it into a view maybe on a very goood as opposed to just Nineteen Eighty Three Landslide Yearr.
NW Wales sounds as if PC have been comprehensively stacked and cracked there.
51 et al: Having read Anthony Wells’s excellent blog on boundary changes a few weeks ago, and having been stunned to learn that Sheffield Central might be changing colour, I did some calculations myself, converting the 2004 local figures into 2005 GE figures using Martin Baxter’s formula. The results for the new Hallam and Central seats would have been as follows:
Hallam: LD 18,878 (42.7%); Con 12,330 (27.9%); Lab 10,273 (23.3%); Green 712 (1.6%); Oth 1,980 (4.5%).
Central: LD 12,898 (37.7%); Lab 12,181 (35.6%); Con 5,254 (15.4%); Green 2,091 (6.1%); Oth 1,787 (5.2%).
Also:
Penistone & Stockbridge (ex-Sheffield parts): Lab 12,397 (49.6%); LD 7,198 (28.8%); Con 3,481 (13.9%); Oth 1,935 (7.7%).
Looks to me like Nick Clegg’s got a pretty easy choice to make!
51 et al: I’ve done a bit of extrapolating for Sheffield, too (2004 locals into 2005 generals), using Martin Baxter’s method, and my notional results for Hallam and Central would suggest that Clegg’s best bet is the former, viz:
Hallam: LD 18,878; Con 12,330; Lab 10,273; Green 712; Oth 1,980.
Central: LD 12,898; Lab 12,181; Con 5,254; Green 2,091; Oth 1,787.
There are bits of Hillsborough and Heeley that are being moved into Hallam where the Lib Dems win locally and Labour are the main challengers. That would suggest that the Labour vote in the new Hallam is a bit larger than in your estimate, and the Tory vote considerably smaller.
I did say my analysis wasn’t that scientific! How does Martin Baxter’s formula work?
Still, I can’t see where Labour are going to garner 10,000 votes from in the new Hallam.
UKIP won’t contest Cheadle
http://news.bbc.co.uk/1/hi/uk_politics/4631557.stm
They only got 1% last month, but if it will be close could be helpfull for the tories.,
75. Every Little helps. If the Tories do put up a fight ona low turnout a couple of hundred UKIP Votes could very well make all the difference for them.
Just discovered Aaronovitch Watch via the BBC link: http://aaronovitch.blogspot.com
Excellent stuff. Its spurious relevancy to political betting? The assembled oeuvre is a good reminder for long-term investors to sell New Labour. Like Aaro himself, their gradual yet enormous move to the right must eventually lead to total collapse.
Why wouldn’t UKIP want to fight Cheadle? After all I would expect it to be a more high profile campaign than South Staffs as far as the media is concerned. Deal with the Tories? Does the Tory advocate UK withdrawal from the EU then?
78 - Day wasn’t a Maastricht rebel and I do not believe he is unusually sceptical for a Tory nowadays. I think it is quite unlikely but possible UKIP stood down for the Tories. They have not been coy (to put it mildly) in standing against the Tories in marginal seats in the past and many (most?) UKIP people would argue that it is quite good to see Tories beaten due to UKIP intervention because it encourages them to take a more eurosceptical line (and on the traitors are worse than enemies approach).
I suspect that they just saw the dismal result (1%) there in June and thought that discretion was the better part of valour - plus the cost of running a “serious” (by their standards) by-election campaign is daunting.
78. “Why wouldn’t UKIP want to fight Cheadle?”
To save their deposit money.
80 - hmmm… I propose all Lib Dems hunt behind their sofas, and we ought to be able to stump it up.
76 - As I have said before it is a fallacy to assume UKIP takes its votes all or even mainly from the Conservatives . For evidence see my post on Party day giving some results from 2 seat wards where UKIP had only 1 candidate .
O/T - does anyone think that this Railtrack thing could grow into a massive story and problem for the Government?
83 - I suspect there is probably too much public sentiment against rail privatisation for it to be a huge scandal.
It may be exaggerated but before you call it a fallacy, maybe you want to email Peter Kellner, he may take issue with that STatement as woul;d a few Pollsters i suspect 82.
83 - If the shareholders win the case then it would at the least be very damaging to Stephen Byers personally and the government but probably not have any long term effect by the next GE .
Agreed to some extent but many Government problems have little to do with ‘real’ public opinion - in fact there being no obvious public outcry can often make things worse because the Government gets complacent about its chances of riding things out.
85 - Quite willing to take issue on this point , I can show some evidence for my view whilst the other view may be logical but no hard evidence
Andrea @ 62: You really are becoming good at the language: ‘Mind that’ ia excellent Lowland Scots for ‘remember that’, though not much used south of the border. When you start posting in Welsh I’ll send you a bottle of champagne!
89. ” When you start posting in Welsh I’ll send you a bottle of champagne!”
Not great chances to see that happening. I don’t even know how to pronounce some names of Welsh seats (like Carmarthen East & Dinefwr, Meirionnydd Nant Conwy, Merthyr Tydfil & Rhymney) or some MPs names (Elfyn Llwyd or Ann Clwyd).
83. The Railtrack move took place in the days before and after September 11 2001 when the eyes of the world and the media were focussed elsewhere. It never came under the critical scrutiny that, perhaps, it should.
I tend to agree with Alex - there are problems ahead for the Government here irrespective of the general public view of rail privatisation.
This was the action of a Government that had just been returned in a second landslide and thought it could do not wrong. In today’s environment this could look tawdry.
3. Back to the main thread. The ICM approach is based on their belief that more people will say they voted for Labour than actually did. So to deal with this apparent discrepancy you have to scale up the Labour proportion not scale it down.
72 - Thanks for those useful figures Aidan , may have a dabble with my own formula to see how close we are .
92 - I can follow that logic ok Mike but I cannot see that it is an immutable law that ICM will always be correct in this assumption . At some time in the future , if this does not apply and ICM make the same corrections they will actually end up correcting in the wrong direction .
OT, Betfair now have a market on whether the ID cards bill will gain Royal Assent in this session. Silly odds at the moment, though, and nothing matched.
I probably have personal views on this too strong to bet objectively. Pessimistically I fear it will not have too much trouble in the Commons - though its passing the Lords is not a foregone conclusion.
94. Are the North Ireland parties pro or against ID cards?
95 - Where the DUP voted in the last parliament, they seem to have voted against. The UUP (including Sylvia Hermon) voted for, and the SDLP didn’t vote.
95. Ian Paisley’s main concern in the HoC today was that ID card data should not be shared with the Republic of Ireland. Moderate nationalists will probably argue that the Union Jack should not appear on the card, while republicans will boycott it.
There is an Irish dimension if ID cards are made compulsory in the UK. Constitutionally Irish citizens in the UK could not be compelled to carry one.
94 , Mark . I think we’re in for a long session of parliamentary ping-pong . The Lords will chuck it back with relish and with the Lib Dems abrogating the Salisbury Convention I think ID cards will get very messy.
95 , Andrea . They usually wait to see what the other side is doing and then oppose it “on principle”.
96/97 Andrea . You see !!!
96. Thanks. It should pass at the Commons tonight. The rebels expected are around 20. 21 signed an amendment against ID cards including Marshall-Andrews (what a surprise!), Clare Short, Glenda Jackson and Kate Hoey. I read some could abstain tonight.
The Lords will surely change it (I’m already waiting to hear Jenny Tongue’s arguments). If the government will accept some changes, the number of possible rebels in the commons should go down.
Bob Marshall Andrews is amusing the House now.
29, I’m sure you’re right about the student presence in the new Sheffield Central.
The top 10 student % (population aged 16-74 who were full time students in the 2001 Census) are currently as follows:
1. Leeds NW 25.8%
2. Cardiff Central 25.6
3. Cambridge 25.4
4. Nottingham S 25.3
5. Manchester Gorton 24.1
6. Sheffield Central 22.6
7. Liverpool Riverside 22.5
8. Bristol West 20.9
9. Newcastle Central 20.6
10. Sheffield Hallam 18.4
Notice anything common?!
Average Labour vote share loss 8.5%
Average LD share increase 9.3%
Average C share loss 2.7%
102 - Hi Robert - Yes those figures are very significant and presumably the non student part of those constituencies had a lower swing Lab to Lib Dem than your figures and the student portion a higher swing .
It was noticeable to me that even in the Brighton and Hove constituencies the Lib Dems did better than elsewhere in Sussex even though the seats were not promising ones on 2001 results and I am sure it was the student effect here also .
The question is what were the key policies/issues that made them vote in the way they did ?
Jack W - The Salisbury convention doesnt hold if you only get 35% of the vote.
I would also argue the Salisbury Convention was finished off by the use of the Parliament Act over a bill which the government did not have a manifesto commitment to pass (only to allow a free vote on).
103.”It was noticeable to me that even in the Brighton and Hove constituencies the Lib Dems did better than elsewhere in Sussex even though the seats were not promising ones on 2001 results and I am sure it was the student effect here also ”
Add to the students the gay community (according to Lord Rennard gays were switching from Labour to the Libdems in great numbers, although many probably went to the Greens too). Hove is the constituency with the highest number of same-sex couples in UK.
New Labour doesn’t recognise the right of the Lords to oppose Government legislation, manifesto commitment or no manifesto commitment, so complaining about the Salisbury convention not being respected seems a bit like them wanting to have their cake and eat it.
Andrea at Ninety in trhe absence of Cymru MArk let me teach you a few phonetic pronunciations. So Camarthen East and Dinufweir
I’ll go through the others later but one big tip in Welsh Double D’s are pronounced Th, SO MEIRIONNYTH
A few greetings Yuchy DA- Basically Cheers
Borodar- Good Morning.
Any Italian in returna?
Gov majority of 31 on ID cards
108 , David . What’s Welsh for the British Lions were cr*p last Saturday ??? Lost a few grand on that one !!!!
BTW , excellent fireworks at Southsea , either the Lib Dems have won a local by-election or Anne Widdecombe has been deflowered !!
The Salisbury Convention only applied to a House of Lords which had a massive in-built Conservative Majority. Now it’s as much a part of history as the Buffs and the Blues.
106 - why did the gay community switch en masse from Lab to Lib Dem? Are the Lib Dems doing anything particularly more gay-friendly than before?
103 - presumably much of the student anger is over top-up fees?
I’d be surprised if gay voters vote en bloc. There is a tendency to assume that “communities” all vote one way.
112.”why did the gay community switch en masse from Lab to Lib Dem? Are the Lib Dems doing anything particularly more gay-friendly than before?”
It’s what Lord Rennard said and what gay magazines reported. The Labour campaign for the gay community was designed to stop this (the slogan was “go to bed with Charlie and wake up with Howard”).
At the end I don’t know if it’s true. In Pavilion they probably voted Greens.
113. They don’t vote en bloc. The only thing that they could do en bloc in not voting tories.
To be serious I don’t think they vote en bloc, but I think that some “conditions” led them to be more inclined to vote for some parties than others.
The “unscientific” polls about voting intentions for gay people indicated that they voted en bloc LAbour in the past, but this time they were more divided (labour/libdems/greens).
Charles Hendry ( tory MP for Wealden) said he understand why gay people think that the tories are against them.
20 Labour rebels tonight: 4 former ministers (Clare Short, Glenda Jackson, Mark Fisher and Kate Hoey) and 2 newly elected MPs (Linda Riordan and Katy Clark)
112 - My suspicions agree with your view that Top Up Fees were the major reason rather than Iraq but it would be nice to have some confirmation in a post election survey . Certainly my son who left Uni some 7 years ago is still very unhappy at his repayments of the student loan which is pretty insignificant to the debts todays students are leaving Uni with .
The gay community wemt over in large numbers to the LIb Dems in 2001 because of clause 28 - Labour was seen to back down - combined with a major campaign by the Lib Dem activist group in the gay media. There is a community up to a point - people talk about politics in the run up to elections and those in gay bars reading the gay press may come persuade each other. But other kinds of community I would have thought are much more likely to vote together. The language communities (e.g. Turkish speakers) are far more separate from one another, with their own media and friendship groups within the community, than the gay community. At least I’d assume that’s so because you can have much less of a friendship outside your community if you don’t speak English fluently.
I don’t know of any reason why the gay community should have come over to the LIb dems more in 2005 than in 2001. IN fact, I think the community media campaign was much less sharp, and people had forgotten about section 28. In 2001 the Lib Dems got more gay votes than Labour, let alone the Tories. I don’t know about 2005.
I think the main reason students moved to Lib Dem was tuition fees and top-up fees. Having said that, when out canvassing, if I met anyone under 30 who said they didn’t know how to vote, I just told them the Lib Dems were the only party to vote against the war in the key vote and they almost always agreed with that enthusiastically. I only tried top-up fees first if they were obviously student-aged or younger.
117.”I don’t know of any reason why the gay community should have come over to the LIb dems more in 2005 than in 2001.”
maybe because there probably were disaffected labour voters in the gay community too (like between heterosexual voters) and, if they exclude the tories, the libdems are the only main choice left.
Charles Hendry thinks that by 2009 gays could think to vote tories (http://uk.gay.com/article/3627/p=1).
About political ads in gay press, if you look at the G-Scene (Brighton gay magazine) issue in the run up of the election, Libdems ads were put near an hald naked man. It was easy to catch them even for not interested readers (you look at man and next to him you find a “vote Marina Pepper” ad), while labour ads were put next to an obituary!
I agree that students shifted to the Lib Dems in 2005, as did Muslims, for pretty obvious reasons.
I doubt if homosexual voters did though. The big shift to the Greens in Brighton probably had more to do with the high proportion of people studying and working at Sussex University, who vote in the constituency.
This idea that the gay community votes en bloc is nonsense. The Tory party is stuffed full of gays!
120. It could be both.
I know these 2 surveys are very unscientific, but there’s everything I was able to find: http://uk.gay.com/headlines/8483?REFRESH
http://www.rainbownetwork.com/News/detail.asp?iData=22585&iCat=29&iChannel=2&nChannel=News (according to that one they switched becuase of Iraq and student fees).
121, an appealing image …
121. Charles Hendry doesn’t seem to think so. He said “within five years, sexuality will be taken out of politics and people will vote according to what matters to them. ”
so he pratically admit that he thinks gay people vote according their sexuality.
121. Rik W, without wanting to be offensive, I think that thinking that the majority of gay people voted tories is a nonsense.
btw tonight the government majority was cut to 27 at one point (313 against 286)
I have to agree with Rik. There are a lot of homosexual guys in the Conservative Party, especially amongst the younger generation in CF. However I do not think that gay issues have any traction with these people. They do not concern them.
Misc responses/comments:
Changing seats: yes, not allowed by Labour rules. Only if your seat overlaps with a new seat are you allowed a shot at it. You also can’t switch from being an MP to an MEP unless you stop being an MP first.
ID debate: much less passion than when we debated it last year, with the 31 majority seen as a bit above expectations. Davis turned in another OK-but-not-exciting performance, with a third of the Tory benches empty. The LibDems obviously cared about the issue, but only some of the Tories did. The DUP voted yes on the Bill, no on the timetable motion (hence a reduced majority of 21 on that), I think. I thought Clarke was good (quietly forensic), but I would, of course. Spoke for 12 minutes myself but had acquired a blinding headache after sitting there for 5 hours, so have little idea how it went. Galloway put in a very rare appearance (don’t think he’s spoken since the election in the Chamber, unless I missed it) and had a row with the deputy speaker over the minute time limit. All rather low-key: at root there are not actually that many MPs who are deeply excited by it (the data’s all routinely-recorded stuff like date of birth and address anyway, so the main innovation is the biometric check on whether one’s who one says one is) and for most it’s just a question of whether the cost is reasonable. I think we’ll get it through all right after some concessions.
126. but do you really think that the majority of UK gay people voted for the tories in the last election?
Gwyneth Dunwoody won her battle: she’ll keep the chairmanship of the Commons transport select committee.
They even offered the transport committee to the tories to try to get rid of her! (oh God, to many !. I’m becaming like Rik W).
127.”much less passion than when we debated it last year, with the 31 majority seen as a bit above expectations”
Nick, you said here that you expected something like 20 rebels. So a 31 majority is line with expectations. Don’t spin too much…
No passion? According to the Guardian there were mocking cries of “resign” to Clarke.
btw did you see the last poll for Germany elections with PDS/WASG at 11%?
RE 128: No. I was referring to gay Tories, who do not have much interest in promoting gay rights issues.
On the issue of ID cards looking at some of the stuff being written about it, it appears to have been an issue that exercised the usual suspects on the labour and Tory benches. However I think the problems with ID cards will come with their introduction rather than pushing the legislation through. Personally though I am against them and am glad the Tories are making a stand. Indeed outflanking labour on issues such as Civil liberties, Zimbabwe and Special Schools is a clever way of gradually moving the party back towards the centre without comprising on any strong Tory principles.
125 - I didnt say that the majority of gay people voted Tory - so how can what I said be nonsense? I pointed out that the Tory party has lots of gay members - and I mean LOTS! So the idea that they are all Lib Dem or Labour is nonsense, as is the idea that the “gay community” votes en bloc!
131. I still think that the great majority of the gay community vote for Lab/libdems/greens and not for the tories.
Do you think is a nonsense? I really don’t care. Sorry if not everyone is able to show your great judgement and makes some nonsense posts.
It’s been apparent to me for a long while that politics attracts a lot of gay people. I can believe that most political parties are ’stuffed full of gays’.
132 -sarcasm is not necessary - and you may be right that “the great majority of the gay community vote for Lab/libdems/greens” but I was pointing out that no-one should assume that there is some mythical community that votes en bloc!
I suppose it depends whether you believe 2% of the population are gay or 10% with most in the closet
The “gay community” may overwhelmingly vote anti-Tory, but then many gays may not be members of the “community”.
Interesting points Alex - maybe membership of “the community” is conditional on not voting Tory!
134. I wasn’t sarcastic. I was pretty much offended.
Calm down people! I think there’s been some misunderstanding about what various people are saying.
108 - When I saw “Welsh Double Ds” why was I put in mind of Gavin Henson’s girlfriend?
121 - and hasn’t it always been thus? Just covert rather than overt.
I think that Andrea has some problem with what Rik is saying. Rik is saying that there are lots of gay Tories and that they don’t “all” vote Lab/LD/Grn (come on, the Norwich and UEA Tories couldn’t survive without them). Andrea mistakes this for thinking that Rik thinks that a majority of gays vote Tory.
Can we just agree that:
(a) There are such things are gay Tories
(b) Homosexuals do sometimes vote Conservative
(c) BUT they are likley to be the third choice, behaind Lab&LD?
140 - yes, what’s Welsh for “Posh and Becks”?
142 - “Posh y Becws”
Except, of course, CC is more “genuine” than Victoria …
141. I haven’t problems with what Rik said, but with how he said it. He could have said “I don’t think you’re right” or “I don’t agree with your idea” instead of “your idea is a nonsense”.
144 - “your idea is a nonsense”.
Welcome to Willis-World
http://www.globusz.com/ebooks/Hitchhikers/00000018.htm Scroll down to the first “Resistance is Useless!”
Andrea. To be fair to Rik saying something is “nonesense” in the context he said it is not impolite. Possibly a little inelegant but that’s all.
146. Yesterday I found his comment offensive. I interpreted that “nonsense” comment like he was implying that my comments were stupid (not that it is impossible that they’re stupid, but it’s not nice to remind it). Maybe it’s time to stop this argument.
Btw I just saw the Independent “Pink List”, which criteria did they use to decide that Angela Eagle is more powerfull than Ben Bradshaw? or that Matthew Parris is more powerfull than Michael Brown (ok, maybe becuase more people read the The Times than The Independent)?.
146 Roger you stucking up for Rik!? I think i’ll have to and lie down. The Shock is too much for mea.
148 - “Stucking”: sticking (up for) or sucking (up to)?
“mea” - culpa?