Which polling methodology are you putting your money on?
January 25th, 2005- Are you more likely to be Labour in public than in private
Today’s ICM poll in the Guardian shows no movement on the pollster’s survey for the News of the World earlier in the month but has Labour 2% down on the last Guardian poll in December. The figures are LAB 38: CON 31: LD 21
With just 100 days to go before a May 5th General Election, there’s a remarkable degree of unanimity amongst the pollsters. The telephone-based surveys - ICM, NOP and Populus - all have Labour on the same figure - 38%; the Tories on 31-33%; and the Lib Dems on 20-21%.
There’s quite a lot of alignment with the internet-based pollster, YouGov, except it has the Lib Dems 4 points higher and Labour 4 points lower.
So with the Tories static at just under their 2001 actual share the big question is the scale of the “toggle” between Labour and the Lib Dems. The higher it is the less the chance that Tony Blair has got of achieving a third landslide victory.
- For gamblers the big General Election bet is about which polling methodology is getting it right - the telephone firms with their higher Labour shares or the internet with it’s bigger figures for the Lib Dems.
For a long time we have been arguing that interviewer-free polls produce bigger Tory figures because supporters are more likely to admit their allegiance to a computer or a robot than another human being. Based on the latest round of polls we might have been looking at this the wrong way round.
- Could it be that supporting Labour is something you are more likely to be in public than in the privacy of an internet survey or a polling booth?
Who knows? We’ll find out on May 6th.
Mike Smithson
MessageSpace Advertising
Not so much ’shy Tories’ as ‘demonstrative Labour’
Remind me - have ICM changed their methodology at all since 2001?
Maybe its a part of the tactical vote unwinding? Or a demonstration of most undecided voters just saying Labour, i.e. the party in government of the day.
ICM’s methodology has sort of changed since 2001 - because their isn’t a sure fire way of working out target figures to weight past vote to, essentially every Parliament the pollsters are starting with a clean slate in terms of past vote weighting.
ICM may or may not have technically changed their methodology - in 1997 they were using a different method (taking an average of past voting recall in their polls) of obtaining a target for their past vote weighting as they do today (taking a mid point between that average and the real vote at the last election). Exactly when they changed, whether it was before or after 2001, I don’t know.
Tom - Anecdotal perhaps, but a friend of mine found strong evidence that in the Con/LD marginal he was working in over the weekend that Labour supporters were still planning to vote tactically to oust the sitting Tory MP.
But will LD’s vote tatically to keep labour in?
Dean - in such seats it gives Labour voters a “bigger bang for their buck”. They can express dsenchantment with Blair and give the Tories a kicking at the same time.
There might be an argument to say that residual Labour voters in such seats may well be the ones more likely to switch this time (disenchanted die-hards) - one such voter wrote recently to the Independent saying she was doing exactly that in Totnes.
However, OTOH there may well be other former Labour tactical voters put off by the increased level of “dialogue” between Labour and the Lib Dems (as others including Nick P have argued), so any effect may be offsetting.
Was this seat in the SW?
Andrew - some will, some won’t. I see the parties occupying a kind of policy Venn Diagram, with areas of overlap between all three and some overlap between each two.
I think there are some LD voters by inclination who would rather see Labour than Conservative, but others who now feel that Blair is not much different and will revert to their original preference (”tactical unwind”).
Steve
It was Harborough where Labour picked up 20% last time. Its one of a number of Con/LD marginals like Bournemouth East where there is still a significant Labour vote.
Andrew
Good point which I think Steve has covered. Thanks for the backgound info on Edinburgh in the previous thread by the way.
Dean - I did a posting a few weeks ago on “squeezeability”, using the criteria of taking Labour down to 10% and adding that onto the LD vote (Con remaining the same). It gave quite interesting results IIRC!
I couldn;t find it so repeated the exercise. +ve means that you can squeeze the Lab vote down to 10% and get a majority, -ve means you need to squeeze below 10% (harder to do).
Bridgwater 6.3
Taunton 4.5
New Forest East 2.6
Norfolk South 2.2
Haltemprice & Howden 1.4
Westbury 1.0
Isle of Wight .8
Dorset West .7
Aldershot .7
Orpington .3
Bournemouth East .3
Wells .0
Cambridgeshire South East -.9
Harborough -1.3
Eastbourne -1.5
Maidenhead -2.4
Folkestone & Hythe -2.7
Surrey South West -3.0
Wiltshire North -3.1
Cambridgeshire South -3.1
Solihull -3.9
Totnes -5.1
Steve - I think the LDs may actually do better in some of these seats than those where the majorities are smaller and they have pretty much exhausted the possibility of squeezing the third party vote further. Having said that there are some seats in the former category such as Bridgwater where there is a significant traditional Labour vote which seems almost impossible to squeeze.
And this is the list of Tory held, LD 2nd seats in order of % Labour vote > 10%:
SEATNAME LAB01
Bridgwater 26.8
Cambridgeshire South East 26.4
Solihull 25.6
Aldershot 25.1
Norfolk South 24.5
Cambridgeshire South 24.3
Tunbridge Wells 23.2
Aylesbury 23.2
Huntingdon 22.8
Saffron Walden 22.6
Cotswold 22.6
New Forest East 21.7
Chichester 21.5
Worthing West 21.5
Westbury 21.4
Surrey Heath 21.4
Henley 21.1
Arundel & South Downs 20.7
Woking 20.3
Wealden 20.3
Horsham 20.2
Folkestone & Hythe 20.2
Harborough 20.0
Ribble Valley 19.9
Hampshire North East 19.9
Bournemouth East 19.9
Hampshire East 19.6
Bexhill & Battle 19.4
Surrey East 19.1
Sussex Mid 19.0
Chesham & Amersham 18.8
Penrith & The Border 18.5
Salisbury 17.5
Wokingham 17.4
Skipton & Ripon 17.4
Leominster 16.8
Stratford on Avon 16.7
Devon East 16.7
Mole Valley 16.6
Haltemprice & Howden 15.7
Wells 15.4
Isle of Wight 15.2
Maidenhead 15.2
Christchurch 15.1
Taunton 14.9
Ryedale 14.7
New Forest West 14.7
Wiltshire North 14.3
Worcestershire West 14.0
Dorset West 13.6
Eastbourne 13.3
Totnes 12.2
Tiverton & Honiton 11.9
Dorset North 11.2
Westmorland & Lonsdale 10.9
Orpington 10.8
Dean - presumably that’s the residue of the Somerset coal field. Of those on the list I only feel confident about Taunton.
Steve - I think so. Labour support is mainly confined to the town of Bridgwater itself.
Like you the only one I’m confident of is Taunton, although I think the LDs will win a handful or so of the others.
For interest, what are the LibDem seats that would fall to the Tories if there is tactical unwind of say 5% from LibDem to Lab (as suggested in the NoW poll of that type of seat, if I recall correctly)?
There are two types of Labour voter who might lend support to LibDems in this sort of seat: those who are unhappy with some of Labour’s policies (e.g. the war) and those who aren’t. Some of both groups will have lent support last time. The sharper attacks on Labour by the LibDems will tend to encourage the first group and discourage the second.
Two interesting points about the Guardian poll. It shows the situation essentially unchanged from that of recent months, with a moderately comfortable Labour lead, and despite a 5% impact of the prisoner abuse allegations it shows opinion on the war still fairly evenly divided (47 against, 38 for, 15 unsure). This is different (narrower) compared with other polls, almost certainly because ICM asks a slightly different question, though I forget the exact wording.
It’s useful to have this poll taken just before the Tory immigration proposals. The next two should show what favourable impact they’ve had and how far it’s lasting. If the answer to the second question is “not very much”, I think that the Tories really will start privately to give up.
Nick
Nick, if my maths is correct (and as you’re the one with the Maths PhD you can put me straight if I’m not :)) I’m guessing that would be LD seats with majorities of <5%, which are:
SEATNAME MAJ01
Cheadle .1
Weston-Super-Mare .7
Norfolk North .9
Dorset Mid & North Poole .9
Guildford 1.1
Somerton & Frome 1.3
Brecon & Radnorshire 2.0
Hereford 2.2
Ludlow 3.8
Newbury 4.8
Romsey 4.9
With the exception of Romsey, all have been discussed here as at risk. All have Labour vote shares >10% with the exception of the last two, which suggests that they’re the ones that have had a significant amount of Labour TV support.
With regards to the attacks, there’s now much more two-way traffic between the parties at the national level than in previous years. My guess is that both parties will play harder where they’re fighting each other than where they’re fighting the Tories (simply via targetting).
Interestingly, in Tabman Steve’s list, the bottom two seats (i.e. the ones with the lowest Labour vote) both had by-elections, as did Brecon & Radnor (5th from bottom), although that was a few years ago now. Perhaps it is only during concentrated by-elections when Lib Dems can educate Labour voters about tactical voting?
Nick 16. The ICM Iraq question is very “Blair-friendly” compared to YouGov and Popoulus.
These are the precise wordings:-
* ICM: From what you have seen or heard, do you think the war against Iraq to remove Saddam Hussein was justified or unjustified?
* YouGov: Do you think the United States and Britain were right or wrong to take military action against Iraq?
* Populus: Thinking about the build-up to the war in Iraq and everything that has happened since, was taking military action the right thing to do, or the wrong thing to do?
In their latrest polls ICM - with its reference to Saddam’s removal - had -9%; YouGov had -21% and Populus -24%.
I have looked at some seats where the Lib Dems are second (eg.the so called ‘decapitation’ seats of Maidenhead and Dorset South and it seems to me that the Labour vote has been squeezed as far as it can be - we are down to the Labour hard core now.
I suspect there might not be many Lib Dem gains from the Tories at all.
The tactical unwind will help the Tories as well, as they are the greatest sufferers from tactical voting.
Augustus - I think there’s some substance to your point. Becasue LD core support is much less than the other two parties, on average “their” voters would be more likely to make tactical voting decisions as a matter of course (and perhaps is the reason behind the “If you thought the LDs could win here” 2001 MORI survey results). Although, again, its debateable whether they’re *anyone’s* voters really.
On the converse I would expect Labour to have more diehard tribal types and because FPTP is less of a problem for Labour (at the moment), there is less incentive to vote tactically unless they *really hate the Tories.
However, as you rightly point out, under the glare of an election campaign many people may find that, like jumping into a cold swimming pool, after the initial shock of changing allegiance wears off they actually quite like their new MP and decide to keep them.
There is a new “decapitation” story in the on-line Scotsman - lifted straight from the Press Association, I think. The link is
http://news.scotsman.com/latest.cfm?id=4044170
By the way there is a very interesting column in the Indy today about the Tories by Steve Richard, wondering if they could spring a suprise at the next election.
The gist of which is “the country is close to their policies but doesn’t think much of them as individuals. They needed a big figure like Clarke or Portillo to lead them out of the mire, but they all deserted.”
As it also says, exactly the opposite to Labour in the 80s (big figures, unpopular policies).
I hope no-one’s bet on an English test match victory
I’m surprised that none of the Lib Dems on here have noted that it is 24 years today since the Gang of Four deserted Labour. An opportune moment to remember that increased representation in the polls, and indeed the popular vote, didn’t translate in seat victories at that time. Could the same happen again or have times changed. Obviously they will do better than ‘83 but in general I think the result will be the same.
Re the Scotsman article: Howard talks about campaigning in every target seat - How many are the Tories seriously targetting?
Philip Stevens in the FT (not able to get it on line unless a subscriber) seems to have been reading these posts - ends by saying “…my guess is Mr Howard will have a lot more to say about keeping out immigrants”
Somebody should ask Howard if he minds that anyone with a Euro passport is free to come (as we are free to go there) - free movement is one of the major benefits of being in Europe.
Steve 2 - you’d better get a bet on the Tories then, 1983 result was 46% Tories , 27% Lab, 26% Alliance. Seriously the Lib dem vote is more targetted now, but it shows that with no chance of forming a Gov substantial numbers did vote Alliance - Wonder what will happen if Charlie gets a bit of momentum going? (Shame ICM didnt agree with You Gov)
Steve 2 - Times have changed to some degree. In 1983 the Alliance got in excess of 25% of the vote and 23 seats. In 1997 and 2001 the LDs got 46 and 52 seats respectively with vote shares of 18-19%.
Clearly the FPTP system works against the LDs, as it does the Tories to some extent, but better targeting of resources and tactical voting has allowed some mitigation of the impact of fairly evenly spread support across the country.
In normal times the main opposition party can be expected to receive the vote of the disillusioned electorate however since the Tory party is perceived by the majority as still being worthy of a kicking [in my case I’m still kicking because of the botched privatisation of British Rail in a scorched earth policy] it means the protest vote is going to the Lib Dems
Policies mean nothing to the protest vote – the Lib Dems have something the other two haven’t – they aren’t either of them. So everything should be smelling of roses for the Lib Dems except it isn’t. Look at the results of the local by-elections shown below:
1. Hackney LBC, Queensbridge [21/01/05]
Con 696 (36.3; +26.1), Lab 595 (31.1; -11.4), Lib Dem 334 (17.4; -5.1), Respect 291 (15.2; +15.2)
The protest vote which you would have assumed would have gone to the main anti-war party went to Respect.
2. Beeston Central Ward – Broxtowe Borough Council. [2003]
Broxtowe B.C consists of 21 wards and ignoring those not in the constituency only 1 ward –Beeston Central gave the chance for some “others” voters to register their votes (1 green also stood in another ward otherwise choices were only the 3 main parties) – the results were
Conservative 359 & 332
Labour 510 & 532
Lib Dem 183 & 179
Green Party 272 & 175
The protest vote went to the Greens.In Beeston generally the Lib Dems polled very well.
My point is that the Lib Dem vote is very soft and vulnerable to Respect in the cities and to the Greens elsewhere, the LD vote went down in over 20 seats at the 2001 GE where the Green Party stood in opposition.
Tabman Steve - I assume you are VinoS - Off message I also went to Folley’s(?) today,first time I’ve caught the tram -impressive & cheap - book shop as well.
Icarus (27) - yes, the Philip Stevens’ piece is interesting. Mike will be pleased to note it even uses the phrase “tactical unwind”:
‘Mr Blair’s concern, though, is what may happen to the Liberal Democrat vote in places where Labour has a small majority over a second-placed Conservative. Quite a few of these constituencies were won by the government in the last two elections because of tactical voting by Liberal Democrats aimed at locking the Conservatives out. But the courtship between Labour and the Liberal Democrats is long over. … By some accounts, and these psephological exercises are not an exact science, up to two dozen Labour-held seats could be lost to the Conservatives if the tactical vote were to unwind.”
When Philip Stevens, the sharpest British political columnist, says it, it’s time to start taking it seriously!
Vino (30) - interesting point. 2 questions/points:
1) Were the LDs targeting either of these wards? If not, you’d expect the LD protest vote to get squeezed by other protest votes. LDs are good at squeezing 3rd/4th party votes in well-run campaigns when it matters.
2) I’m interested that in 21 seats the LD vote declined where Greens stood. In Oxford East, where I live, it was very noticeable in 2001 that LDs attarcted a lot of ’split tickets’: ie, local Green voters voted LD nationally. Assuming a 05/05/05 poll combining local and national elections, I’d expect to see this repeated. My *guess* would be that those 21 seats were not LD targets; so again vulnerable to a split protest vote.
Vino - I enjoy our E Mids sparring far too much to pose as someone else. I have no idea who VinoS is - s/he certainly isn’t me!
This is like Bill & Ted’s Bogus journey, with Good Vino and Evil Vino
If I’d have known you were going to go to Foley’s (although if you spend too much, as I’ve been known to do in bookshops, it could be more accurately described as Folley’s) I could have popped in to say hello in person.
33 -Stephen Tall - I’m not sure - to be honest I don’t whether they were specifically being targeted by the Lib Dems.I quite sure that there is a lot more seats than 21 where the Lib Dem vote went down in the 2001 GE when the Green Party stood against them BUT in a lot of seats I would have thought the Lib Dem vote went up despite the Green Party standing.I still stand by my main point which is the Lib Dems are very vulnerable to Respect and the Green Party.
34 - Tabman Steve - sorry I got you wrong on VinoS - but it is the Midland sense of humour i.e very dry - I’m off on leave and got willingly dragged via tram to Station[London tomorrow]-Foley’s since you mentioned it-only 2 books both sci-fantasy £1.00 each-Broadmarsh Centre-Victoria Centre then home on bus.Nottingham’s public transport is first rate.
It is excellent - I have a choice of several buses to get to work. I’m not actually from the Midlands so hence my lack of dry humour! I’ll check out Foley’s again tomorrow.
The volatility of this upcoming election is largely down to two major factors:
Opinion Poll accuracy (or lack) - will they be closer than in 01?
Tactical Voting - will it unwind, and if so, how far?
On the first, the widespread belief is that the opinion polls err in Labour’s favour. The “failure of the polls” in 92 did lead to major investigations by the polling companies and changes of strategy by many. Nevertheless, in 97 and 01, they still overestimated the Labour lead. Have they got it right this time?
In 97, the polling booths produced 103 Tory voters, 94 Labour voters and 112 Lib Dems for every 100 of each that the pollsters found (from the Martin Baxter historical poll of polls at the time)
In 01, there were 108 Tories, 91 Labour and 107 Lib Dems for every 100 of each that the poll of polls predicted. (In 92, the figures were 116 Tories, 90 Labour and only 92 Lib Dems per hundred of each predicted).
On Baxter’s current poll-of-polls, we have Con 32.07, Lab 36.97, LD 21.04. If that is as accurate as in 97, it would reflect a “polling booth” value of Con 33.03, Lab 34.75, LD 23.56. If it is as accurate as the 01 polls, it would be a “true” value of Con 34.64, Lab 33.64, LD 22.51.
The latter figures give a wafer-thin Labour majority of 8 (with no TV unwind): Con 231, Lab 327, LD 57.
So we have (without Tactical Voting addressed) a range of Labour majority of 8 to 116 (the “8″ assuming accuracy equivalent to the 2001 polls, “116″ assuming the poll-of-polls to be spot on).
How about Tactical Voting?
The accumulated Tactical Voting squeeze, from the 92, 97 and 01 elections, equates to a total of +3.6 Lab to Lib and +10.2 Lib to Lab (that is, to get the current status quo on the 2001 vote shares from a hypothetical baseline of no tactical voting, you’d need to input +3.6 and +10.2 into the relevant boxes).
This means that if anti-Tory Tactical Voting were to disappear completely (which is unfeasibly unlikely - voter inertia should lead some to continue), the values for these boxes would be -3.6 and -10.2 respectively. Putting these in should give the absolute limits of Tactical Voting unwind.
For the “97-class accuracy” projected vote shares of Con 33.03, Lab 34.75, LD 23.56, it gives Con 285, Lab 277, LD 54 [Highly non-intuitive - the Con seat share is higher than Labour on less of the vote! Then again, with vote shares between the top two parties this low, it all depends on which seats are lost where]
For the “01-class accuracy” projected vote shares of Con 34.64, Lab 33.64, LD 22.51, it gives Con 315, Lab 248, LD 52.
Please note that a full tactical Vote unwind in one fell swoop such as this is really unlikely - I’m trying to show the limits of the possible.
So, on the poll-of-polls as of 9th January, we can have Con at 180-315 seats, Lab at 248-381 seats an LDs at 52-58 seats. This is a range from Labour Majority of 116, to Conservatives minority of 9 seats. Now that’s volatile.
Also note that the polls could be inaccurate the other way - the pollsters might have overcorrected. In which case, that same poll-of-polls could reflect a Labour majority of 130+
As a PS, I put in “Mean” values - projecting vote share using derived accuracy midway between the 97-class and 01-class accuracy polls, with TV unwind at 50% (-1.8 and -5.1). This gives a hung Parliament: Con 262, Lab 295 and LD 58.
38 Andy Cooke - very impressive and well written - I must be quite sad for I think I understood it!
Vino,
You must have been hanging around here too long
Thanks for the kind words
Andy
A very interesting analysis. WHat would it be if the ‘92 class of accuracy prevailed. For what it is worth I think that your projections on this basis are far more likely than the wild talk we hear from Lib Dems on this site. I am looking forward to some very disappointed Lib Dems on 6th May.
I remember how in Reading year after year the Lib Dem group leader would confidently predict how they would gain the one ward we then held - and every time he failed. Last year I was there when we took one of the three seats in his ward!!!!
Rik,
Thank you. To use a ‘92-class prediction, just multiply the Con vote share by 1.16, The Lab vote share by 0.9 and the LD vote share by 0.92. I must warn you, though, I regard the ‘92-class accuracy level to be in the “freak poll inaccuracy that should never be repeated” category - the pollsters would probably have heart failure if that kind of result was to be repeated.
Interesting though just to put in the straight results from 92 and 97 into the electoral caculator. Taking just the % figures for the parties in 92, under today’s electoral boundaries the Tories would be 16 short of a majority on 308 seats. Whereas in 92 they had a majority of 21 and 336 seats. Interestingly using those figures labour only has one more seat but the Lib Dems have 15 more (on 35) As for the numbers 97 the results are broadly similar. It goes to show though the effect tactical voting has on the Tory party today, which means it could not win a majority on the same shares as in 92 (when they also suffered from more limited TV)
Andrew Milne at 43,
The effect, when the boundary changes is taken into account, is even greater:
At http://www.election.demon.co.uk/bcsum.html there is a summary of the partisan effects of the 97 boundary changes - essentially +7 to Conservative, +2 to Labour and -2 from LD, taking the notional 92 result to Con 343, Lab 273, LD 18.
Using Martin Baxter’s calculation of the effects of the new Scottish boundaries (which shouldn’t really be directly added to the 92 result as it is based on the 01 vote share, but what the hell - it will still give a pretty close summary of the partisan effect): -1 Con, -10 Lab, -1 LD. Would give 342 Con, 263 Lab, 17 LD - a majority of 34 for the Tories.
Compare with, as you say, Con 308, Lab 272, LD 35. A majority of 34 taken to minority of 16. As you say, the Tactical Voting that already existed in 92 is ignored here; only that which has been squeezed further since then. It’s a definite issue.
If, as Mike suggests on his next article, the new Tory election guru is going to use the “send a message to Tony Blair” theme, if all he does is to prevent a large proportion of the Tactical Voting, it could have the same effect as a large swing in the polls.
I personally thought that this election would see the rise of the “shy Labour” phenomenon, especially among GMW, to complement (or possibly even overtake) “shy Tories”, as the Tories steadily become more “respectable” and Blair becomes more tarnished. Not much evidence of it yet.
26 What you forget is that then the Alliance vote was spread evenly across the country. Now we have 40% plus in getting on for 200s eats and 15% minus for almost an equal number.
35 Vino - Like Steve T, I suspect most of those seats if not all were not target seats.
Oh Rik - Go on, let’s have that £50 bet on S&C ?
45, I don’t know about that one, I think there are a few shy labour out there, their……… well just to shy to say so
47 is that spam?
slots In practice all men are atheists; they deny their faith by their actions.
forex currency…
forex currency They precede the terrasse by three or four intimists, with Tschaikowsky Benedict. At last I destroyed Madonna-stamp, our giustizia, to take him as a two-daughters-in-law, for I cared for confisi somet…