Monday’s polls

We have three polls today, the daily YouGov poll, the weekly Ashcroft poll and the twice-weekly Populus poll. Topline figures are:

Populus – CON 32%, LAB 33%, LDEM 9%, UKIP 15%, GRN 6% (tabs)
Ashcroft – CON 34%, LAB 30%, LDEM 5%, UKIP 15%, GRN 8% (tabs)
YouGov – CON 35%, LAB 31%, LD 8%, UKIP 14%, GRN 6%

Lord Ashcroft’s poll repeats the Tory lead it showed a week ago, YouGov produce a four point Tory lead, their largest since January 2012. Populus continue to show a one point Labour lead.

All the usual caveats apply, we need to be careful not to overreact to polls that could just be a couple of outliers in the same direction – that said, in YouGov’s daily poll we’ve reached the point that Conservative leads are a little more common than Labour ones. Of the last ten YouGov polls there have been four Tory leads and two Labour ones. I don’t think we can confidently say the Tories are ahead… but I’m certainly no longer confident in saying that the underlying average is a small Labour lead either. I think we can fairly say that the Conservatives don’t seem to have suffered any short term damage from the debate debate last week.


340 Responses to “Monday’s polls”

1 4 5 6 7
  1. Graham

    My point is that the Greens are weaker now [in Norwich]– particularly in relation to Labour – than in the period leading up to the 2010 election when they significantly outpolled Labour in the local and Euro-elections.

    It’s worth pointing that the Ashcroft poll we saw in June 2014 had the Greens on second at 20% up from 15% in 2010. This presumably means they will be less prone to Lib Dem squeezing than they were in 2010.

    In June the Greens were only at 5% in YouGov’s monthly averages, while they have been at 7% nationally Dec – Feb. And we know that even in seats with a higher than average Green vote originally, that movement seems to translate into an increase in VI. Between June and December Ashcroft showed an 8 point increase in Brighton Pavilion.

    http://lordashcroftpolls.com/constituency-polls/

  2. @Mike N
    “What is it about the SNP that terrifies the English? Is it the tartan terror?”

    I would have thought it was obvious. The stated intention of the SNP, the very reason for its existence, is to break up my country. I’ll repeat that again: the SNP’s aim is to dissolve the country I call home. I am English but moreover I am British, so I have *every* reason to be worried about the SNP gaining power.

    I’m not a fan of some other political parties either, but the SNP is different, they are committed to ending the UK so I regard them as a bigger threat than any other party.

    To be honest this is no joking matter, and I’m pretty sure I’m not alone in England or any of the Home Nations in being worried about the SNP having influence over the UK government. I expect this possible Lab/SNP collaboration to have some effect on VI because unlike the debate debate, this is very emotive and personal for many people across the UK.

  3. @CMJ

    I’m afraid your group selection process seems too intuitive for me to have any real confidence in this process. I suspect we’ll just have to disagree about whether this constitutes definitive evidence in support of a Tory upturn.
    Using a one way ANOVA – A is significantly different to B, and B is significantly different to C.
    As A is less than B, it must also be less than C.

    I’m afraid this is not true of the main effect, which was the only one listed in your link (the one with two degrees of freedom). As I said, the ANOVA null hypothesis is: Mean A = Mean B = Mean C. Rejecting the null hypothesis merely shows that there is a difference somewhere. It doesn’t establish that pairwise comparisons are individually reliable. That’s why statisticians have developed a wide range of contrast tests to deal with exactly the situation you are facing. You could run a post-hoc contrast using the weights [ 1, 1, -2 ] or perhaps [ 0, -1, 1 ] or you might like trying using Helmert contrasts . Whichever (relatively standard) path you take, it still requires a few more steps beyond reporting the main effect to establish that Mean C is greater than Mean B.

    So, in short I still don’t buy your claim that the Tory VI is indisputably on the rise. still, safely within MoE as far as I can tell..

  4. The rise of the SNP may prevent LAB from being the largest party but – assuming they work together to keep CON out of power – it should make no real difference.

    eg. Lab (&SDLP) 270 + SNP 45 = 315.

    Assuming that Plaid and Green do not vote with the government either, that gives a left wing coalition something like 318 seats. With Sinn Fein refusing to vote, a CON government could fail to get through any/most bills.

    Of course the key to all of that is “assuming they work together” and some folk doubt it. But I just cannot see the SNP supporting CON at any price. It would be political suicide back in Scotland.

  5. @Statgeek

    Six-poll sample total (six, six, and five for the most recent data point):
    9623, 10275, 8492
    Margins of error:
    1.00, 0.97, 1.06

    Sound about right. But your UK-wide graph had the Tories ahead by LESS than 1%, and therefore comfortably within your own stated MoE.

    In other words, all you can say at the moment is that the two graphs have merged, and it will take a bit more data to establish that Tories have actually moved ahead.

  6. @Andy Shadrack

    “I see no classic swingback for the LDs [in England]”

    …and?

    The only seats that a swingback to the LDs (or lack thereof) seem likely to affect are Bermondsey and Watford, and at a stretch Sheffield Hallam if you accept that Clegg is in real danger.

    Other than that, the seats in which Labour are a threat are write-offs anyway, while ‘classic’ swingback is not applicable in inter-coalition contests. Conversely, the Greens still only seem on course for one seat under FPTP, despite their upturn in the past 18 months appearing quite stable.

    Nationally I don’t expect the LDs to squeeze them back down – few of those converts are going tp change their mind about which party they prefer. But in the 15-20 LD seats where a LD–>Green switch truly does increase the chances of a Tory win, I’d stake serious money on the LDs milking that for every vote its worth.

    The Lib Dems may be about as popular as a trip to the dentist, and little will change there in the coming weeks. But the idea that they have suddenly forgotten how to target seats is fanciful, hence continuing projections of 25 seats (+/-5), rather than the almost total wipeout that their VI indicates.

  7. I suspect a huge swathe of the electorate are unaware Miliband is Jewish. His problem has always been that he does not connect with the electorate and does not appear comfortable in his own skin. However underestimate him at your peril because he is fiercely determined…but I don’t think that determination will be quite enough to win the election.

  8. David in France: the SNP won’t support the Tories at any price for the reasons you give. And on issues aside from independence and defence they and Labour seem natural allies.

    But the very fact that SNP need to block or minimise a Labour recovery in the election after May (whether that’s in August 2015, May 2020 or some place in between), makes any Labour-SNP coaction inherently unpredictable, however intuitive a united front might otherwise seem.

    And from Labour’s side, they need to finely balance listening to the voters in Scotland on policy direction, with not putting Con-held marginals out of reach by being seen to be putting England low down their priority list. That balance is possible, but again gives rise to unpredictable behaviour.

    Make no mistake, I think Labour in Scotland are stark raving mad for thinking that “vote SNP, get Tory” will work, because in most constituencies voters know they are choosing between two candidates whose leaders will not work with Cameron. But after the election party posturing will undoubtedly come into play – the mid-to-long term stakes for both parties are too high to expect anything else.

  9. @Unicorn

    Statistical confidence intervals are an arbitrary benchmark used to “prove” a hypothesis, no more. However, in a context where we are trying to predict the most likely outcome and look for emerging patterns, there is little sense in using a 95% confidence interval as a test with which to be certain that something different is afoot. If you hold out for that degree of certainty, you’re most likely to find that you’re last out of the blocks in reaching a conclusion that was pretty obvious some time earlier. (And since polls are a moving feast, something else may have come along to upset the applecart in the meantime.)

    To give a practical example. Nate Silver made predictions as to the outcome of the US presidential election in terms of probabilities, even in the very last days he gave Obama only a slightly greater than 90% chance of winning, which basically represented the possibility that all of the polls could be systematically wrong due to similar errors. Apply a statistical test to that, and you are short of the 95% probabiillty that would typically test a hypothesis that “Obama will win the US presidential election.” Yet the polls led all but the most blinkered observers to reach that conclusion nonetheless, and could still be used by Silver to predict the outcome in every one of the individual US states.

  10. JP

    @”Conspiracy theorists really are a detriment to sensible debate”

    Looks very much like an excuse in case of failure-in order not to have to face the reasons that he has such poor ratings amongst the voting public.

  11. If SNP diminish Labour’s seats and the Cons win more seats then surely Cons win whatever.
    Ignoring all the kingmaker scenarios public democratic outcry will see the party with most seats wins.
    Having said that are we really expecting Greens at 8% UK at 18% in a GENERAL election?
    Doesn’t this always end with small parties way down and LDems up?

  12. Uk = UKIP apologies

  13. Despite the pummelling their likely to receive I actually think the Lib Dems will remain kingmakers after the election. Look at it realistically I think its fair to say that if Labour end up the largest party (in spite of the surge of the SNP) Milliband is heading to Downing street (how stable that government would be is another matter) however say Labour are not the largest party. If labour are less than 10 seats behind I still think Milliband would become prime minister BUT what if the Labour advance on the Tories is pretty feeble and they gain less than 20 seats while the SNP nearly wipe them out in Scotland. Imagine this very plausible scenario.

    Tories=295 seats (+10 seats from Lib Dems, -18 seats to Labour)

    Labour=245 seats (+10 seats from Lib Dems, +18 seats from Tories, -40 seats to SNP)

    SNP=56 seats (+40 from Labour, +10 from Lib Dems)

    Lib Dems=27 seats (-10 each to Tory, Lab and SNP)

    Then no real surprises, 8 for DUP, 3 for SDLP, 3 for Plaid, 2 for UKIP, 1 for Greens etc

    In that situation the Tories are clearly the biggest party but the “right wing bloc” of Tory, DUP and UKIP equals only 305 seats.
    While the “left wing bloc” of Labour, SNP, Plaid, SDLP and Greens equals 308.

    Neither side can form a government without Lib Dem help so where do the Libs go. Support the biggest party but somehow work with parties like the DUP and UKIP of which they disagree so much on.
    Or support the group with arguably the bigger mandate, of which they also agree on many issues but it would literally be a losers coalition.

    I get the impression that the decision would depend on whether Clegg or Farron was leading the negotiations.
    Anywho just my own internal musings.

  14. @Rivers10

    I broadly agree. I actually think the most likely scenario will be another Con-Lib coalition (with possible DUP support). It’s not exactly my dream coalition of half Conservatives, half Liberals, but beggars can’t be choosers…

  15. 2 for UKIP would surprise me…

    The scenario you describe is viable though. I do think the LDs could work with Cameron again, but the other parties you mention are undoubtedly a bridge too far.

  16. Phil Haines

    …and in times like 1992, the statistician can smugly lean back and say “I told you it wasn’t a dead cert”.

    There are times when a 95% degree of confidence is asking too much (particularly when the data has pretty huge errors on it), any statistian worth his salt would say “If you want something intelligible from the polls get me polls with a bigger sample”.

    Unfortunately most statisticians aren’t wealthy enough to commission they own set of supersize polls and so we have to make do.

    If anything the papers that commission them like having a larger MOE so they can make stories around the sampling error which happens a lot more frequently than real movements in the polls. The people commissioning them don’t really care on a day to day basis about whether the polls have really moved or not.

    If you had to make real decisions based on statistics, the first question you should ask is “How expensive would it be to get 10x the amount of data?”

  17. Rivers 10….

    your post was really interesting. I think if the tories finished with 50 more seats than labour, the lib dems would back them …this is essentially a repeat of what happened in 2010, with the tories on 306, or 307 if you count the speaker, and labour on 258 or so.

    I think, and most of the projections even now suggest, that the two main parties will be much closer to each other than 50 seats…maybe as few as 5-10 seats would separate them.

    This would make a calculation as to which side to back much harder….

    The perception thing about who “won” the election is important, as it feeds into narratives which will be spun immediately after the outcome. this is politics not mathematical proof. The SNP’s surge has given the tories the chance to be the largest party which seemed remote before the surge itself happened.

    If labour hold 35 out of their 41 seats in Scotland, they are very likely to be the largest party. The market thinks they will lose a lot more seats, that’s why the odds are now on the tories getting more seats.

  18. RIVERS10

    In that scenario, I think the most likely outcome would be a confidence and supply arrangement between Cons and LD. 322 is probably just enough to survive any confidence vote.

    I can’t see the LDs wanting a) to back a Labour Party which would (in your scenario) have “lost” the election in terms of both votes and seats or b) to deal with the SNP

    Re Clegg vs Farron, I think Farron might well be keener on C&S than a formal coalition. I suspect Clegg would be happy to see the current coalition continue…

    LD activists will be very wary of another coalition, given the difficulties of the last 5 years.

  19. @Phil Haines

    You are perfectly right. There is no reason at all why UKPR contributors should stick to standard scientific levels of stability. People could quote (say) 75% confidence boundaries and everyone else would at least know how much confidence to place in their observations.

    I am not quite sure why as a community we would opt to do this, especially given that pollsters themselves routinely quote 90% intervals. What I have been drawing attention is rather frequent lapses in attention to the endemic problem of distinguishing ‘signal’ from ‘noise’.

    Almost without exception, contributors seem to accept that you can’t be confident about the interpretation of a single poll and that there is a need to wait for more evidence before we can tell exactly what is going on.

    But this sensible caution quite frequently goes overboard when the discussion turns to rolling averages, trends and graphical patterns (such as crossover). Obviously MoE (albeit sometimes reduced in magnitude) also needs to be taken into account in deciding whether any of these effects are real. It doesn’t matter how strict a criterion anyone uses, as long as it is evident that there has been some attempt to tease about what is real from what is random.

    For my own part, I haven’t been able to find any secure evidence of a recent Tory VI increase that is not entirely compatible with mere random variation within the standard MoE.

    I have run a linear regression on Tory VIs against number of days since Jan 1, and the slope is just 0.000147 (with p-values in excess of 50%, if you are interested).

  20. Lot of chat about Ed M here today. I see he’s been given a boost on the BBC news and we shall have to see what that does for him. My suspicion is that, despite the complaints about using wife and children, it is what he needs – pullovers and jeans, the whole schmaltzy nonsense, is what is required of him.

    I still think that even with Con on 285 and Lab on 275, he is home (perhaps not totally dry) with SNP and / or LD. This is based on GB Con 34% Lab 34%.

    So as they say, ‘as it stands’. I don’t invoke predictions whether based on astrology or any other occultish method.

  21. @ Roger Mexico

    I agree with your assessment of Norwich-South. In last years city wide elections Labour attained 36% to Green 30.3%, with Conservative on 16.4% UKIP on 9.9% and LD on 7.2%. I think the Green vote was lower in the European election, but they still came second close on the heels of Labour, but did attain more votes than Labour in the 2009 EU election.

    But based on the city wards you advise me are in Norwich- South the vote was a lot closer:

    Labour 35%
    Green 34.7%
    Conservative 16.6%
    LD 11.9%
    UKIP 1.9%

    Who wins Norwich-South may very well boil down to where the LD vote goes in Eaton ward (that is their last ward with 3 councillors) and who UKIP picks votes up from as they only ran in one ward in the city wide elections in Norwich-South.

    This to me this has all the makings of a re-run of the Brighton-Pavillion election battle in 2010, except voters in Norwich-South now know what a Green MP looks and acts like in the HoC, and the Labour candidate is not the incumbent.

    Where Labour may win back support from students is on their tuition fee policy, but the Green Party has continuously called for tuition fee abolition and has a much more popular position on “trident” than Labour.

    I spent my last winter in England in Norwich in 72-73 and it was quite the hot bed of radialism back then. The students actually organized a strike in support of improving construction workers working conditions, who promptly told them to get back to class as their taxes were paying for their university education.

  22. BRISTOLIAN HOWARD and RIVERS 10.
    I think that Labour are going to gain about 40 Tory seats and 15 Lib Dem seats, and lose about 30 seats in Scotland.

    That takes them to about 282 seats. If the Tories gain about 15 LD seats they would be close to the Labour figure.

    However, it is all conjecture; exciting though.

  23. @Phil Haines

    (Accidentally submitted before completion…)

    This non-significant Tory rise represents a VI of about 1% between now and the election.

    Over the individual months of January and February, Labour were very reliably ahead of the Tories.

    If you take the March polls alone this puts the Tories 0.7% ahead on average, but – with a very high p-value (0.394) – this is a difference that could very easily be ascribed to MoE.

    I would grant you that @CMJ’s CUSUM methods are potentially more sensitive than most of the alternatives. But I get uneasy when claiming real effects depends on dividing the VI series into rather indistinct subgroups. These methods were developed for an industry in which it is much cheaper to handle the effects of an unwarranted claim (false positive) than to miss a real breakdown in stability, and I don’t think the balance is quite what is needed here.

    In sum, beyond the extent to which these effects can occur randomly, I don’t yet see any evidence of a real Tory VI increase, and certainly no evidence yet for crossover.

  24. Candy

    Regarding the Conservative poster of Salmond with Miliband in his pocket – this is not a reference to Scottishness, but to Salmond’s recent boast that he would make Miliband “dance to his tune”. Hence the puppet-master image.

    Well it depends what you mean by ‘recent’, but the most recent reference to when that quote was said by Salmond comes from October 2009

    http://news.stv.tv/politics/133094-david-cameron-says-alex-salmond-is-irrelevant-in-westminster-election/

    There are numerous references to the phrase in the last few days, but they are all to Conservative uses of the phrase to describe a possible post-election situation, Salmond currently appears to be as conciliatory as you would like.

    Oddly enough, when the phrase was used by Salmond over five years ago, the idea was dismissed as ‘irrelevant’ by a certain Mr Cameron (still what does he know, sounds very dubious, what with that Scottish surname).

    Mind you, I suppose it is more recent than Deuteronomy.

  25. @Roger Mexico

    In any case it is extremely sexist. It will be Nicola Sturgeon with Ed in her pocket.

  26. “2 for UKIP would surprise me…”

    Lots of the establishment underestimate UKIP. Lab and Con are in long term decline. 2020 will see lab and Con below 60%, possibly 50%.

  27. 45% want independence in Scotland, and the SNP look like winning 80%+ of the seats. At least 45% want out the EU, so don’t underestimate Ukip, Labour voters want out of the EU,yet Labour are alianaiting their voters.
    Not offering a referendum is suicidal.

  28. @Andy Shadrack

    In the local elections of 2008 and 2009 the Greens were well ahead of Labour but still came a poor 4th. They had a higher profile here at that time with Adriam Ramsay as their candidate and Deputy Leader. He is not fighting the seat again and that fact is in itself meaningful. Green party activits openly admit that they have little chance in May and would be pleased with second place.
    I write as someone who will be voting Green on May 7th – albeit in Norwich North.

  29. @Andy Shadrack

    Fair to say many in Norwich Labour Party would not be unhappy with the Green candidate.

  30. Interesting that Roger Scully’s Ratio Swing (similar to EF Strong Transition I believe) projections for Wales show Ceredigion as a Plaid gain from Lib Dem.

    He points out UNS as flawed in projecting Ceredigion as on current Plaid numbers UNS still doesn’t predict a Plaid gain in Ceredigion even if Lib Dems poll 0%.

    Now this more aggressive model also predicts Lab to gain Arfon but overall that’s still a net loss of -1 to the coalition parties compared to UNS.

    Given Plaid’s significantly higher numbers for Assembly polling it suggests a classic Westminster ABT squeeze of Left of Centre voters in most of Wales.

    Given this won’t apply in Ceredigion where Plaid are the leading anti Coalition party I think this is a seat we should be paying more attention to.

    Perhaps the Others 21 that all the models/projections seem to agree on (as everyone seems determined to count John Bercow under Conservatives) should actually be Others 22?

  31. @Phil Haines

    As everything else it depends on the quality of information put in. Japanese steel yields may be more accurately reported than peoples’ voting intentions given that a steel billet doesn’t have the same capacity for lying

  32. Just to clarify to everyone my earlier post was not a prediction, rather it was just a plausible outcome with interesting implications. For what its worth I’m of the fairly common opinion that Labour and Tories will be VERY close on seats and thus the arithmetic makes the only “workable” government a Labour minority supported by the left of centre parties. But I could be very wrong…

  33. Wolf
    Looking at Labours candidate for Norwich South I think its reasonable to assume they are taking the Green threat there seriously and have selected there candidate accordingly.

  34. “their” rather

  35. I think that whilst possible wins of minor parties in certain constituencies, such as Norwich S, may indeed be of interest, it’s the overall result that counts. A few green upsets against Labour and UKIP against Conservatives is not going to alter the parliamentary arithmetic. The core question (given if Scotland is a given) is ‘will Labour get at least those 40 marginal seats of the Conservatives’.

    Nothing else is of consequence as I see it. Perhaps I am politically naive here. If someone such as Ashcroft just polled 3 Con /Lab marginals in England every week, we should have had a very secure view of the May outcome. I grant this would have missed the Scots polling ‘event’, but from a Conservative party viewpoint, that shift is almost academic.

  36. @Bristolian Howard
    “Lot of chat about Ed M here today. ”

    Mostly far fetched stuff about him being attacked because of being Jewish. Ed is not religiously Jewish. He’s an Atheist. He has said this many times. Indeed, had he been a practising Jew he would not have eaten the infamous bacon sandwich!

    He is culturally Jewish. His parents were/are Jewish and their families suffered greatly during the Holocaust.

    None of the above is relevant to the current/recent attack on EM, which are based on policy (internal and external opponents of his direction of travel for the Labour Party) and what some see as his unsuitability to be PM (mainly from his internal and external opponents for the aforesaid reason).

  37. FWIW, Labour have dropped very suddenly with Election Forecast over the last few days. Prediction now at Con 293, Lab 271, SNP 40, LD 24.

  38. @James

    No surprise. A small uplift in Tory fortunes equals a large upward score on the seat forecasting models. Were the movement to have been the other way, Labour would not have increased so much.

  39. @Unicorn

    I did state “if this most recent poll is part of an underlying trend”, which was to say we have to wait and see.

  40. @ Raf

    As I understood it, Tark was not saying that EM was being attacked for being Jewish. Rather, that internal and external opponents, had an investment in ‘whistling’ up prejudices already existing in some parts of the electorate which might be exploited in a negative manner for EM.

    This is straightforward Negative campaigning and is ‘playing the man not the ball’.

  41. Survation –
    “In the next 2 weeks we have polls out on Scottish, UK & for the first time Israeli politics.”

    See the effect of all this blethering about Jewishness on here?

  42. @Syzygy
    Thanks for clatifying that Sue. I did not read the original message, merely some of the later references to it.

    @Tark – I apologise if I inadvertently misrepresented your position in any way.

  43. BristolianHoward
    I think Lord Ashcroft came up with a decent formulae for predicting the magic number of seats Labour need to gain off the Tories for Mill to become PM. Working on the assumption that any of the changes amongst the smaller parties as you suggested doesn’t effect the arithmetic cos they somewhat cancel each other out then the formulae is 46 plus x divided by 2.
    46 being the current seat gap between the Tories and Labour, x being the number of SNP gains from Labour. The resulting number is what Labour needs to be equal in seats to the Tories.
    I personally see 46 plus 30 divided by 2 thus giving 38 required Labour gains from the Tories so pretty much in line with what you said.

  44. @OldNat
    “In the next 2 weeks we have polls out on Scottish, UK & for the first time Israeli politics.”

    The latest polls for the Israeli General Election have Zionist Union leading with a projected 24/25 seats with Likud second with 21. There are 120 seats in the Knesset. That’s what you call a divided electorate!

  45. RAF

    The trouble is so many there seemingly want others to be divided along the lines of the judgement of Solomon!

  46. @Oldnat

    I must have missed that particular blether.

    Polling, schmolling.

  47. At least 45% want out the EU, so don’t underestimate Ukip, Labour voters want out of the EU,yet Labour are alianaiting their voters.
    Not offering a referendum is suicidal.

    Not sure where you get any of that from?

    The last YouGov poll on EU membership showed 45% in favour of staying IN and only 35% in favour of leaving.

    Also, as far as I am aware, most polls show the majority of LAB votes in favour of staying IN the EU.

  48. @ Chris Hornet

    I would be intererested to know which seats you think LD will save and why?

    I speak from the experience of watching the Progressive Conservatives implode in Canada from being a majority government with 169 seats to having just 2 between 1988 and 1993.

    I would also refer you to the experience of Comhoantas Glas in Ireland and more recently Pasok in Greece, who adopted unpopular positions for which the electorate later made them pay for, by withdrawing voter support.

    There are times when money and capability to get someone elected are simply not enough.

  49. Rivers10
    “I personally see 46 plus 30 divided by 2 thus giving 38 required Labour gains from the Tories….”

    Yes, that is about the sum of it, I agree.

    Taking this analysis further, I think one of the most significant findings of the past fortnight was that the swing in Con/Lab marginals where a large proportion of the electorate are public sector is well above UNS – whereas in Con/Lab contests where low public sector numbers exist, equals a much lower swing.

    I do not have the data on the percentage of the electorate in public services for all the Con/Lab marginals – but the task of Labour gaining 38 from Conservative seems much harder than it first appears.

    Especially if a significant number of narrowly held Con marginals resist the trend because they contain few public sector workers.

    Has anyone worked out a Conservative seat vulnerability list based on Ashcroft’s findings rather than straight UNS? It would be interesting I think?

  50. @ rivers10

    Despite the pummelling their likely to receive I actually think the Lib Dems will remain kingmakers after the election.

    I think most on here have been suspecting that for a long time. And I would imagine that the LD leadership have also suspected it for some time. Probably for some years.

    LD will hold the key unless LAB get 285 seats or more. In which case it will be down to LAB and SNP to come to an agreement.

    But will LD want a coalition with a party committed to a referendum on the EU? I suspect they may do so. But it could be chaotic for the Conservatives if they lose such a referendum (ie. Britain votes to stay in). It could be hellish for the LD if they lose such a referendum (ie. Britian votes to leave).

    And what sort of coalition could have it’s two (main) members so divided on such an important issue? The in-fighting would surely be politically fatal to both sides?

    Certainly some interesting times ahead…

1 4 5 6 7