New swingometers

A brief note – I’ve updated the two graphical swingometers on the site so they are based on the 2017 election results. The basic version is here, and the fancy version that lets you put in separate Welsh and Scottish figures is here (the old version without the map isn’t yet updated).


828 Responses to “New swingometers”

1 2 3 4 17
  1. The main 2 reasons the Tories have better vote to seat ratio after 2015 would be:
    1) South West England: Lib Dems held 57 seats in 2010. A lot of these are now conservative especially in SW England
    2) Labour seats lost in Scotland to the SNP. Although the gains in 2017 would improve the vote to seat ratio in scotland

  2. @AW

    Yep, the data in the LSE article is out of date, but it was useful explaining some of the different factors influencing electoral bias. As you say, it comes down to a variety of factors, what was troubling me was the notion of things being intrinsic to the system.

    Thus, FPTP favours the winner, this is accepted as being an artefact of the system. But tactical voting is also an artefact of the system, but if one party’s voters are better at it, this is not to be accepted. Similarly, choosing not to bother voting in a safe seat is a natural consequence of the system, but if you happen to have retired voters with time to vote and your opponents have younger voters busy with work and parenting who decide not to bother as their party will win the seat anyway, this is to be penalised in favour of the party with older voters.

    A split vote can harm disproportionately under the system, but this is accepted as an artefact of the system, unless the effect is uneven and a particular party suffers whereupon we have to do summat about it. Meanwhile smaller parties are massively disadvantaged but nothing is proposed for them, this is just another artefact of the system to be preserved.

    Of course, micro targeting can also affect things, whereupon maybe the idea of boundary changes suddenly becomes less attractive. This is before the problem people may respond to changes, and if seats become more competitive maybe more younger people will make the effort to vote if seat is no longer as safe. On top of sudden changes like in Scotland it seems rather like playing Whackamole, or Votamole….

    (Indeed, one might even see some of what happened in 2017 in part as a response to Tories suddenly becoming more efficient what with Labour losing Scotland and Tory micro-targeting. Thus the mobilisation of the youth, and the way the “progressives” returned to Labour etc.)

  3. @Trevor Warne

    “CON ran an awful campaign – no sweeties in the manifesto, terrible effort to sell their manifesto, failure to keep the focus on Brexit and maybe not doing enough to motivate their core voters.”

    Yes, up to a point. It was a poor campaign, certainly compared to Labour’s, but by winning their biggest vote share for 30 years, I think it’s pretty obvious that the core Tory support (circa 30%) turned out in spades. I’m defining “core” as voters who wouldn’t contemplate voting for anyone else and habitually turn out for the party they’ve supported all their lives. Wouldn’t have thought many of those types of voters failed to turn out for the Tories last Thursday. More interesting to speculate where their big increase came from. The revival in Scotland, and the more tepid one in Wales obviously helped, as did the UKIP collapse and the appearance of “Anyone-but-Corbyn” older, ex Labour voters. This pool shrank as the election campaign wore on, and I’d be intrigued to know how many 2015 Labour voters actually voted Tory in 2017, and why they did so. Brexit? Corbyn? Economy? Security?

    My hunch is that there is a certain softness in the Tory vote, self evidently so considering they performed historically so well by their recent standards. Must have been many first time Tory voters on board, but how long will this newly assembled coalition of voters last as the Tories incumbency piles up a litany of disappointments? I suspect some of it will melt off in a fit of buyer’s remorse quite quickly. How quickly, and how permanently, will depend on events, and how Labour play their surprisingly acquired good hand, but I think the Tory/DUP administration will quite quickly become mired in deep unpopularity of a type that Cameron and May surprisingly avoided for much of 2010-17. Labour’s woes in that time obviously assisted them, but they’re now facing a resurgent and buoyant Opposition. Different game on now, I suspect. Different parameters completely.

    My last point brings me on to local elections. For a long term incumbent government, the Tories took surprisingly few big hits at local elections between 2010-17. They actually advanced in the most recent ones! Miliband had the odd good night, but nothing spectacular by historical standards and the Tories remain surprisingly strong in local government, certainly for a 7 year governing party. At similar stages in their respective hegemonic central government days, Thatcher and Blair had been hollowed out at local government level and these were key factors in the eventual defeat of their parties. Very few councillors and foot soldiers on the ground. Shrivelled local party organisations and low morale.

    I rather suspect that the current Tory Government, battling headwinds on Brexit, the economy, public services and defending a wafer thin Commons majority, will suffer grievously in the local elections to come. That could, surprisingly, be quite a factor in how our politics pans out over the next few years.

  4. Carfrew – turnout is the other way round actually. A party having low turnout in its safe seats produces a “bias” in *favour* of that party, not against it (imagine, for the sake of it, that there are ten seats, all with 10,000 people. Five vote for party A, five vote for party B, but turnout in party A’s seats is 80% and turnout in party B’s seats is 40%. Overall party A would have 66% of the vote, but only 50% of the seats.

    That said, I’ve never actually seen anyone saying things should be done to iron out “bias” from turnout, tactical voting or the pattern of party contestation – only from uneven seat size (people often think more of the bias is from uneven seat size than actually is, but I think that’s more from ignorance or wishful thinking). It’s why I try to remember to refer to it as “bias” – the system does often skew in favour of one party or another, but that’s often just because of people using (or not using) their votes in a way that maximises their impact, which is their right.

  5. @CrossBat11

    For the Tories to do well they need to feel threatened and have a leader .prepared to meet the public.

  6. Matt126,

    Spot on which is why I feel that a further swing Con-Lab would probably lead to some LD gains off Cons, as I can’t see the swing being due to Con holding and Lab over taking.

    This would mean more ‘waste’ Tory votes being a close second to LDs.

    So whilst the current Cons advantage would not disappear imo it would likely narrow so less than 2% needed by lab for most seats. OM not helped by LD taking seats off Cons of course; although Lab get a decent number of seats for a modest vote share increase in Scotland so in practice 3% Con-Lab swing might just be enough for Lab to squeak a majority.

  7. Wish we had proportional representation. I voted for AV, hoping it might lead to PR at a later date. But people are still wed to FPTP with a local list of candidates.

    If we are going to have more coalitions or C&S agreements between parties in the years ahead, i wonder whether voters might be turned off by the haggling that goes on. Not to mention the feeling of instability and elections taking place more often, if that is the case. The reaction might be to give a party a majority and smaller parties see reduced votes.

    I predict that the DUP will only offer limited support to the Tories, which might just be Queens speech and first budget. After that, any support will be subject to discussions. The DUP will be mindful that they have greater responsibilities in Northern Ireland.

  8. Re the Libdems with a new leader to come they have always in election manifesto appeared to position left of centre going for Labour votes. Would they now be better moving slightly right and target Con Voters.

  9. @CROSSBAT11
    I think the ‘anyone but Corbyn’ vote diminished virtually to nothing during the campaign, in line with the relative standing of JC v TM.
    Here in the People’s now even more Socialist Republic of London, however, there remained a significant Brexit shift. A number of Lab voters lending to Con because they didn’t trust Lab to deliver a good stiff Brexit: more Con voters going the other way because they DID trust Con to deliver a ruinous (in their opinion) Brexit.
    My guess is that these are short term vote loans, though I’m hopeful many will remain beached comfortably on the Labour shore because of the wider tidal flow in favour of Lab.
    What this means in locals is another matter: most of the councils round here are Labour but few are entirely safe and I suspect they might suffer from that ole political pendulum – you’ve been in for the last 10 years and we’re not that happy (and we glaze over when you say it’s all down to government cuts in support grant). On the other hand I’m hoping that the continued struggles of the Tories at national level and the strength of the Lab revival will see us home.

  10. R Huckle – well said. My preferred choice would be elections every 4 years and a 60 seat STV system with 10 MPs in each seat. Largest party gets to try and form a government first.

  11. @Anthony

    ” turnout is the other way round actually. A party having low turnout in its safe seats produces a “bias” in *favour* of that party, not against it”

    ———-

    Yes, that’s what I’m saying, although maybe I could have explained it better. I mentioned that part of the reason for Labour’s better votes/seats ratio is lower turnout in safe seats, hence mentioning younger people busy as opposed to retirees. And if you correct for this, should you really, and won’t you then potentially get more turnout in response? Or targeting…

    I wasn’t arguing Labour were disadvantaged over lower turnout in safe seats, I was questioning whether it’s really fair to consider it a bias in favour of them as opposed to just another artefact of the system. And it’s a game of whackamole anyway. But I’m still getting my head around it all so prolly coulda been clearer.

    However, I wasn’t sure about what mechanism is being used to determine boundaries. If it’s just equalising the size of constituencies that’s one thing, if trying to correct for things like turnout it’s another. I take it from what you’re saying it’s just constituency size. As for correcting constituency size, of course there’s the issue of whether to use the whole electorate or just those registered…

  12. Good morning all from another sunny day in the great metropolis that is the PSRL.

    AW thanks for the updated swingometers – had some fun playing wit. Put in Tory 36 Lab 44 and the only Tory seat left in SW London was Chelsea and Fulham. But at 29 – 48 it went red.

    In terms of the current systemic ‘bias’ in favour of the Tories, given the influence of Brexit on voting patterns and historically rather large swings we have seen in certain areas currently I don’t think it matters as much. In FPTP there tends to be a tipping point where the system starts to by its very nature favour the largest party. So fi Labour does reach that point they will reap the benefits in terms of seats.

    I remember in the early 90’s when all the commentary was about how the boundaries were favouring the Tories and Labour had a mountain to climb.

    Overall I think this election will go down as a cautionary tale (for both sides) in regards to the assumptions that are made by parties in regard the electorate, leadership appeal and the use of data. The Tories adopted a very simplistic and one-dimensional strategy which really did fall apart when in contact with the enemy. Labour as a party probably did miss opportunities for bigger gains as to the very end they thought it was them that was on the defensive – as it was largely using the same assumption and data as the Tories.

    Corbyn took a big gamble in choosing the approach he did and it paid off – the extent to which this was either due to insight or luck will be dependent largely on one how views Corbyn. However the next election is likely to be very different and more focused on the economy, and the Tories will not make the same blunders again. My guess is we will be in for another long night in the not too distant future.

  13. While we’re at it, do people have the explanation/explanations for Labour tending to have smaller constituency sizes? Is it some inner city thing?…

  14. AW,
    What has not been factored in is that people (including the party HQ staffs) believed the polls in April and May (and it seems until the exit poll had been up and first counts declared).
    Postal Voters and even on the day in person voters who wanted to vote for the Conservative brand, and for a vote of confidence in clear comprehensive Leave aka Brexit, were convinced the only question was size of Queen Theresa’s majority.
    They felt they could and should mitigate its size especially as they became concerned about who this absolute ruler potentially was and what she truly believed.
    They felt a much smaller majority than 100 or 50 would be the safest bet for Brexit Conservative brand and not dementia tax, fox hunting, no pension, no fuel, Theresative party.
    They also believed Corbyn had zero chance of victory. Moreover, they believed his seat share was set to reduce heavily.
    Thus, both traditional Labour and Conservative voters voted for Corbyn’s party to tame Theresa May’s autocratic non-Cabinet government rule.
    They did not realise they were so may that they could take it too far and give her no majority.
    This has been seized on by Remainers to discredit Clean Leave Brexit.
    However, the Labour and Conservative Leave voters who wanted to tame Theresa believed that the LibDem clear second referendum call was the true test of Remainer support and the polls suggested it was feeble. Below 10%.
    Corbyn and McDonnel seemed to be either closet Leavers or at least committed to the referendum result f 2016.
    The absence of Labour naysayers in the campaign shored up the view that a vote for Labour was not a signal that Brexit was in jeapordy.
    The belief in the polls and certain Theresa May victory led to an overdose of protest.
    I say this as a pro-Israel, pro-undivided Jerusalem, pro-Bible, pro-Jesus Christ socially conservative with a small c, and Healey-Callaghanesque pension and welfare for vulnerable type of guy.
    The Conservative-DUP tie up is the dog in the fight I did not realise was really there, based on believing the polls !
    DUP are pro-Israel as per vote in October 2014 in House of Commons.
    Theresa May is anti as per UN vote of December 23 last year.
    People believing the polls and less than sold on their first choice policies and conduct led to protest vote which changed the result.
    How can pollsters call that ?

  15. @ CROSSBAT11 – more detail on my “complacent CON” hypothesis.

    EU Ref turnout by age (lots of discrepancy but taking Ipsos-Mori adjusted numbers):
    18-24 48%
    65+ 81%

    GE (some difference between different post vote opinion but taking something middle, change from 2015 in brackets)
    18-24 57% (+14)
    65+ 77% (-1)

    Now looking at changes and changes from EU-2017:

    18-24 (+9)
    65+ (-4)

    Looking at the seat level detail you can see the drop in CON support in the richer/older demographic postcodes and the gain in LAB support in the student seats. I really don’t want to turn this into an ageist debate (I’ll save that for other members of my household!) but the numbers show CON turnout (in % of total turnout terms) was disappointing – certainly not the swallowing of UKIP and capturing of the Leave vote that was expected at the start of the campaign. “Complacent CON” is a scenario term but the evidence looks compelling.

    I agree that CON are entering the 3rd term (depending on how you define that) territory of being blamed for everything bad (and receiving no recognition for anything good) – is it “bad” enough yet to win a full LAB majority (and hence push for a 2nd GE)? Probably not IMHO – the missing link for a large shifts in polls is economic downturn which (again IMHO) looks quite likely.

    @ MATT126 – agree. If we move (back) to a 2party system all this view of CON or LAB having 40-45% ceilings in vote share is incorrect and (based on current seat analysis) an even split slightly favours CON (depending on how you view the residual LDEM/UKIP/Green vote splitting) – as AW has pointed out. I doubt LDEM can make much recovery (IMHO they will lose their champagne socialist seats in S.W.London and wealthy retiree hotspots back to CON (small sample evidence to suggest that occurs if next election looks like it might result in LAB majority) but elsewhere in the country the ABT vote might help LAB). The return/or not of UKIP will be a much larger worry for CON (IMHO)

    Anyway, another beautiful sunny day so chat later.

  16. @Matt126

    The Lib Dems with a new leader (and a young fresh one if it is Swinson) have a big opportunity with Brexit/ the economy to position themselves as pro european. Accept we are out but focus on the new arrangement (with the implication that re-entry is possible if terms are right)

  17. @AW

    Looking back at what I wrote, this might have caused the confusion:

    “Similarly, choosing not to bother voting in a safe seat is a natural consequence of the system, but if you happen to have retired voters with time to vote and your opponents have younger voters busy with work and parenting who decide not to bother as their party will win the seat anyway, this is to be penalised in favour of the party with older voters.”

    In the last bit, by “penalised in favour of older voters”, I meant eradicating the supposed “bias” in favour of Labour, to in turn favour the party with the older voters. But if there’s no correction according to turnout etc. the point is somewhat moot…

  18. Bardin1,
    “Accept we are out but focus on the new arrangement”

    if the libs take your advice they are rather irrelevant. There is no USP to be had in saying ‘we accept the result and will get the best deal we can’, which is what both lab and con will be saying.

    Indeed, if either main party gets round to saying ‘we do not accept the result because the nations view has changed’, they would be well and truly up creek without paddle.

    At the moment everything is still posturing and still still nothing has been done. No one wants to commit further or get off the fence despite the inherent contradiction of ‘we accept the result’ and varous versions of ‘we will get the best result’.

  19. JSB – not sure this vote Labour to reduce the majority thing impacted that much in the end, it may have helped in Labour held marginal a bit where concentrating on the local sitting MP may have worked to a degree but hard to see how it could in Tory held ones that they lost.

    Guymonde – In my NE seat there will still a fair few 2015 Lab-Tory movers based on Corbyn, Trident and his past associations,

    The numbers reduced as the campaign roll out but was still a factor only to be trumped by the new voters and to be a surprising number of con-Lab which early evidence suggests a large number came from 25-45s particular women.

  20. @Danny

    I should have perhaps clarified by saying ‘accept we are out for now’ .

    My point is there is with the election result and a new leader, there is now no need for them to try to justify their old position of we are out but not really because there will be a referendum in 2019 – just vote against the deal in parliament in 2019 (assuming its bad, and therefore ‘not what people thought they were getting when they voted leave’) and campaign in 2022 to go back in

  21. Apparently the Queen’s Speech is now going to take place next Wednesday:

    https://www.theguardian.com/politics/live/2017/jun/15/theresa-may-northern-ireland-talks-dup-deal-delayed-politics-live?page=with:block-59425a89e4b0240ef7614ad9#block-59425a89e4b0240ef7614ad9

    whether or not the DUP deal is agreed. It may be an attempt to stop the DUP drawing out the negotiations endlessly (you can’t say I didn’t warn you) as nothing is decided yet.

    According the Private Eye, Brenda has informed them that she’ll be going to Royal Ascot instead (I’d want to stay away from this mess as much as possible too), so the Lord Chancellor will have to read the speech instead.

  22. I’ve only read the most recent page of posts today (mea culpa!), but I must say the tone seems to have improved a lot. Some very good analysis and non-partisan posts.

    There’s been some discussion of the LibDems position and prospects of a revival. One unusual aspect of this election was the extent to which they were squeezed. Though their number of seats increased, they actually got the smallest share of the vote that they or their predecessors have had since 1959!

  23. @Trevor

    Yes correlation does not necessarily equal causation but it can give you ideas about where to start looking.

    Regarding the plot for Health being similar to Brexit, if this were the case, haven’t checked yet, it might be summat you’d expect, if both Brexit and Health were associated with DE social strata, and this strata in turn were associated with Tories.

    (The interesting thing anyway is how the data correlated with Tories BEYOND the DE strata thing).

    Regarding Labour vote in Uni towns, why is this a problem? It’s not a surprise since there was a lot of latent potential in those towns, lots of potential Labour votes that hitherto hadn’t turned out. But it doesn’t disprove the point that Labour vote didn’t pile up in safe seats. Canterbury is a University city, but was hardly safe Labour. Indeed, if you want to take not just marginals but safer Tory seats, you can use places like Uni towns with a big latent vote!

    It’s possible there’s a fair bit of correlation not equaling causation in your post!!

  24. @ CARFREW – many seats saw increase in an already substantial LAB majority. The demographics by seat info that RICHARD sent out showed where LAB could gain student seats that they did not already hold.

    Yes, it is a correlation but the correlation of LAB GAINS from CON versus youth % of total demographic is very high (as was Remain vote which correlates well with the London gain).

    I’ll be careful to make sure to clearly state correlate in future as I agree my post did indeed push the boundaries of “inferring” causality.

    LDEM analysis with lots of correlation evidence coming soon once GBP market settles after BoE report and can reset short GBP position.

  25. @Sorbus

    No problem, Here is the fuller article I wrote on this:

    https://politicalstatistics.wordpress.com/2017/06/14/a-review-of-green-performance-ge-2017/

    I am aware of the post-election internal debate within the Greens. There have been some articles praising the success of the progressive alliance, but frankly they are devoid of evidence and appear to me to be justifying what essentially is a failed policy (IMHO). It’s a bit like trying to feel better after being run over.

    I think the Green star is waning, and unlikely to recover without PR, or some political earthquake occurring.

    Perhaps the Greens can consider themselves a bit like UKIP – never had real power, but have influenced politics beyond themselves.

    Are smaller parties really a source of alternative policies that get eventually absorbed into the main parties if the idea looks like being a threat to the mainstream?

  26. @Trevor

    Lol, you’re pursuing a straw man. Yes, I don’t think the argument was that Labour wouldn’t increase vote in safe seats too. The question was whether it would be confined to safe seats or whether it would also happen where they needed it. (It did, though clearly still a bit short, but the way it’s currently trending might give Labour further hope…)

    Yes, there’s a correlation with youth vote too, and even Remain. It’s entirely possible to appeal to several demographics at the same time.

    Look forward to the LibDem analysis!!…

  27. Carfrew – the Boundary Commissions’ jobs are solely to draw up seats that are within the quota for electorate size (and pay attention to local govt boundaries, respect local ties, minimise change, etc). They certainly aren’t supposed to try and “correct” for turnout or tactical voting or anything. There would be a very strong case against that for all the reasons you’ve said, and I’ve never heard anyone seriously suggest it.

    And there are more questions than electorate or those registered. It could be those registered (i.e. all British & Commonwealth adults who have actually registered), an estimate of those eligible to register (i.e. an estimate on the number of British & Commonwealth adults living in that area – the theoretical number of the register if it was 100% accurate and complete), all adults (all adults living in that area, regardless of if they would be eligible to register) or all people (counting children too). One could make cases for all of them, and they would probably all have different party partisan effects. The comments here are probably not the place to do so!

    My view is that it’s actually a bit off a red herring – people missing off the register are partially cancelled out by inaccurate entries on the register (often they are the same people who have moved address without re-registering). It would make less difference than people think. Personally the change I would make is to base the review on the electoral register *on the day of the last general election*, rather than on the 1st Dec following the general election. An increasing number of people do not bother to register when they move house or in the annual canvas, but instead register only in the weeks leading up to an election – hence the register on election day probably maximises completeness.

    That’s by-the-by, the most important things are frequency, speed and tolerance. The older the boundaries the more they favour Labour, the newer they are the more they favour the Conservatives (and because the boundary commissions try to minimise change, a more relaxed quota will benefit Labour through inertia, a tighter one will favour the Conservatives). The difficult thing, therefore, is find a happy medium there when any decision is essentially arbitrary but gives a small partisan advantage to one side or the other.

  28. @Catman

    Whether small parties prosper under FPTP depends to some extent on whether people feel they can afford the luxury of voting for a minor party, which in turn depends on how much of an emergency they consider it to be to foil another party, and whether their preferred big party tick enough boxes.

    Miliband didn’t tick enough boxes, and we hadn’t had Brexit in 2015, and we’ve had more years of the current economic policy which prolly isn’t the preferred Green option…

  29. LDEM analysis with heavy correlation and possible explanation (not proof!)

    Lost seats with PC/LAB gain
    Ceredigion (has a large uni but not dominant in demographics, low house prices)
    Sheffield Hallam (has a uni element but not dominant in demographics, lowish house prices)
    Leeds North West (student, lowish house prices)

    Lost seat with CON gain
    Southport (bit weird, retiree town – sticks out as error given the large swing change to CON from LDEM)

    Held seats (all under threat from CON)
    North Norfolk (retiree area, majority dropped)
    Carshalton (lots of offsetting factors resulting in little change in % vote)
    Westmorland (retiree area, lowish house prices and large swing to CON = now held by a small majority so not a great place to have your leader! bit weird should have stayed safer LDEM IMMHO possibly a move against Farron?)

    Gain seats (all from CON)
    Eastbourne (retiree town, high house prices the small swing masks the drop in UKIP vote)
    Oxford West (has a large uni but not dominant in demographics, high house prices)
    Twickenham, Kingston (no obvious demographics, high house prices, high remain votes)
    Bath (student, high house prices)

    Haven’t used this analysis for Scottish seat gains – mentioned that on HIRETON comment earlier.

    Back testing the results using the following parameters:
    1/ student/retiree
    2/ average house price v national average
    3/ Brexit vote (London)

    My model and hence In My Model’s humble Opinion (IMMHO) suggest (but does not prove!) LDEMs performance in most seats was due to:
    A/ losing the student vote to PC/LAB and losing areas with low/lowish house prices
    B/ gaining support in retiree towns with high house prices
    at the expense of CON
    C/ Remain possibly the same thing as London bias (London like Scotland seems to be its own nation!)

    If someone (YouGov!!) has enough sample data they could see if this can be seen in polling info. If my model is correct then LDEMs vote share change should be very obvious in regional and demographic cross breaks.

    The next GE might have different levers. Will CON drop the “attack” on pensioners and will this cut away LDEM’s gains and hand the seats they took from CON back to CON? Could CON also take Westmorland (v.small majority)?I suspect LAB will keep the student vote and LDEM stay squeezed out of Wales and most areas of England but how much more tactical vote can they take from LDEM?

    Next GE LDEM prediction for E+W
    Lose 4-8 seats to CON (possibly wiped out)
    Make no gains from LAB
    Tactical vote tough to call. ABTory or ABLabour (ABC is too confusing as people using the C for Conservative or Corbyn so moving to ABT and ABL)

    P.S. More good news for mortgage free people with bank deposits (highly correlated to older people) – looks like BoE’s patience at looking through the “temporary” rise in inflation is over and rates will go up soon. Doubt the snap recovery in GBP will last though…

  30. @AW

    Yes, I had somehow gotten the impression somehow they might try and go beyond just constituency size. But in any event, the other effects, turnout etc., are often mentioned as biases in favour of a party and I was exploring to what extent this is really the case, and whether it’s worth worrying about given votamole etc.

    Take your point there are further complications regarding who you include in the constituency. Some of course argue for as many as possible on the basis the MP represents all, not just those who vote, but we’ve had that debate before. (It is at least a salient debate!!)

    Thanks for filling in on the numerous bits of the puzzle I was missing. I shall retire to ponder further…

  31. CMJ
    “Perhaps the Greens can consider themselves a bit like UKIP – never had real power, but have influenced politics beyond themselves.
    Are smaller parties really a source of alternative policies that get eventually absorbed into the main parties if the idea looks like being a threat to the mainstream?”

    I think that’s true under our system. However, is that necessarily a bad thing? For instance all the main parties now have some kind of ‘green’ policy. The Green Party themselves might not think that those do not go far enough, but if their ideas are beginning to be adopted, surely that’s good? It would only be people using the Greens as a route to power who would be disappointed. Same applies to UKIP of course.

  32. @CARFREW

    The main impetus for voters to move to smaller parties is the differential between the two main ones. If voters think they are very close then they will bleed voters from the edges from the more engaged ideologically partisan voters. So lab to greens and cons to UKIP. Of course other things can outweigh such effects where the election becomes too important to protest belief vote, brexit being a prime example.

    I think in general we are slowly moving to a viewpoint among the electorate that FPTP is past its prime and that we need some sort of proportional system. This viewpoint is particularly prevalent among the younger generations who have less loyalty to a particular party but rather vote more on short term concerns, though that’s not to say they are equally willing to vote for all parties. Some like UKIP are Kryptonite to them so would never get young peoples votes no matter what they do. Its that lower level of party loyalty which makes the young much more open to electoral reform than their elders, so given demographic replacement I think its only a matter of time until it happens. Thus we see Corbyn making electoral reform a plank of the labour manifesto, which aided in attracting the under 40’s vote to the party.

  33. TREVOR WARNE

    LDEM analysis with heavy correlation and possible explanation (not proof!)

    Lib Dem seats tend to show the importance of candidates and campaigning. In Southport John Pugh, the previous Lib Dem MP, retired unexpectedly when the election was called, giving them little time to find another candidate, never mind promote them and tactically squeeze Labour. They held 93% of their vote but ended up third. A collapsed UKIP vote put the Tories in and there wasn’t time to respond by taking votes from elsewhere. Indeed the seat is now a three way marginal (though Pugh only had 31% to start with.

    On the other hand in (demographically-similar?) Eastbourne the losing Lib Dem MP from 2015 was standing again. Despite no UKIP candidate and their 6k votes going mostly to the sitting Tory, the Lib Dems had two years to campaign and were able to transfer any Corbyn surge to themselves tactically.

    Similarly the surprise victory in Oxford West was due to an active candidate who fought last time, not much UKIP vote for the Tories to benefit from and the ability to keep the Labour vote to previous levels. There was also a decent Green vote (from Bernie Sanders’ brother) to pick up. University presence in a very strong Remain area must have helped as well.

    Twickenham and Kingston also show the same pattern helped by previous, high-profile MPs standing and were able to keep Labour to their previous levels. And in Bath the Lib Dems were able to benefit from an unpopular Tory MP. But in all three of these the Tories retained their votes and the Lib Dems won from increased turnout, keeping Labour static and squeezing Greens.

    So looking at Lib Dem results it’s probably better to look at mechanisms rather than demographics for an explanation – though those were probably needed as well.

  34. @TREVOR

    If interest rates are going to start rising it will utterly nuke our housing market. The main reason they have not as yet is because of that fear at the BoE, and I don’t think older people will see that outcome as “good news”. Though a large downward correction in house prices is needed, all those pensioners sitting on unearned gains who were hoping to use those gains as their pension are going to see them disappear. Not to mention all the older voters who have jumped feet first into buy to let. There will also be lots of newly unemployed as the wider economic effects are felt. Its going to be economic carnage if/when it happens.

  35. The Advanced Swingometer gives a useful insight on possible changes in Scotland if another election comes soon, and the parties are doubtless taking note.

    Were the % vote for SNP to fall by 2% from 37% to 35%, they would lose a quarter of their seats (33 down to 25, 6 to SLab and 2 to SCon), on the AS.

    It is possible on the device to change % vote manually for more than 1 party, which approaches the present situation whereby the SNP vote loss goes to both SCon and SLab, if not also to SLD. But I wonder if greater predictive power would result if we could have 2 regions of Scotland, the Central Belt and the Rest (including both North and Borders).

  36. Back to Scotland… move (back) to 1D politics (left = SNP+SLAB, right = SCON+SLIB)

    7 of the SCON seats (1 hold, 6 gain are self-supporting at current polling levels)
    6 of the SCON gains relied on tactical voting and have almost exhausted the SLIB assistance (ie vulnerable)
    Argyl & Bute the only “easy” SCON gain from SLIB tactical vote
    Fife North East the v.obvious SLIB gain from SCON tactical vote or just three more votes!
    Caithness Sutherland and Edinburgh West majority cushion could be helped by SCON tactical vote
    SLIB other two seats don’t need tactical help and as “safe” as anything is in Scotland these days (IMHO why Jo Swinson is bookie fav to take over as LDEM leader)

    Unless the polls move much SCON+SLIB in all but two cases have the tactical “right” vote pretty well fully exhausted after several elections homing in on the best way to split the seats out. At 17 seats combined they might gain 2 (1 each) with a little mutual help but SCON might also lose 6.

    As mentioned before the “left” side hasn’t shown much/any tactical vote. SCON and SLIB voters both unwilling to help SLAB in 2017 result analysis.
    Of the 42 “left” seats SNP currently hold 35 but 11 of those are within a 5% swing to SLAB and conversely 5 of SLABs 6 seats are within 5%ish swing of returning to SNP.

    Summing up changes predicted for next GE (with no change in poll % for any party)

    SCON -6 / +1
    SLIB 0/+1
    SLAB -5/+14
    SNP -11/+8

    SLAB need to work harder to regain the “left” vote. Looking at 2010 and previous elections (ie prior to SNP dominance) the left/right vote is going to be harder to break and generate seat gains – the battle they need to win is taking out SNP and reuniting the left under the red banner. SNP will obviously be trying to do the opposite and reunite the left vote under a yellow banner. SNP also have incentive to sow discord between the right/union parties tactical voting successes (given their opposing views on Brexit the SCON+SLIB grassroots tactical alliance is astounding and shows how little impact Brexit had in Scotland for this GE)
    Conversely SCON should be very happy to keep the left vote split (and should make sure they keep SLIB happy!)

  37. @ ROGERMEXICO – thank you for the excellent input on specific seat issues for LDEM – I’ve made a note of all of those. I tend to look at result purely through the “eyes” of my model so I miss specific seat input.

    @ ALEXW – since the biggest losers from a PR system are the two dominant parties it would be very unlikely that either would agree to a change.

    P.S. I agree on your view on the housing market – instead of the budget deficit being helped by “taxing” old people’s property gains house prices may well become more affordable for the younger generation – a silver lining perhaps!?!?

  38. @CatManJeff

    “I think the Green star is waning, and unlikely to recover without PR, or some political earthquake occurring.”

    Agree. PR would be nice for groups like this, but sadly I don’t see it happening soon.

    It may well be that many genuine Green voters lent their vote to labour this time, but it could just be that Labour also stole their ground to some extent. If you believe the Political Compass assessments of manifestos (and I have to as I haven’t time to read them all), then in 2015, the Green stance was the only option in England for people vaguely on the left or libertarian side:

    https://www.politicalcompass.org/uk2015

    Outside England their are other choices, but if you look at the ballot paper choice for many people in 2015, it’s varying degrees of authoritarian right, or the Green party.

    But 2017 was rather different:

    https://www.politicalcompass.org/uk2017

    Now Green and Labour are very similar by their assessment. So in the days of Greens getting a heady 4%, it wasn’t necessarily people who thought of themselves as naturally ‘Green’ voters, but just that they were the only party that catered for their views.

  39. @AW

    “Personally the change I would make is to base the review on the electoral register *on the day of the last general election*”

    There’s quite a bit to be said for this, including…

    An interesting side effect would be to increase activity in safe seats during the relevant election campaign. It would be important for parites to maximise voter registration where they are strong, to prevent a reduction in the number of those seats.

  40. scotland

    For those who thought my posts hysterical. Vindication.Alistair Weeks on the PB site.
    You read it here first:-)

  41. Hello all from sunny Bournemouth East.
    There is an interesting YouGov tweet about The PM’s popularity falling quickly.
    The Cons must be hoping it will recover.

  42. THE GREEN VOTE
    @ Catmanjeff
    Thanks for these data, which show the size of the Green vote in the seats which Lab just failed to win. One can also look at the other side of the coin: the seats where Lab just won; in nearly all of the 15 Lab wins with majorites under a 1,000, the fall in the Green vote from 2015 was greater than the Lab majority.
    As @ Triggguy said, the unusually high Green vote i 2015 probably reflected disillusion with both main opposition parties.
    These martginal shifts don’t matter in a big win for either side, but the last three elections have been close, when every seat counts, & maybe the next one will be as well.

  43. O/T

    Someone has put the UK govt up for sale on Ebay:

    http://www.ebay.co.uk/itm/1-Government-USED-Quick-Disposal-SW1A-United-Kingdom-Circa-2017-/252989301531

    They’ve had 26 bids so far, but have only raised 48 quid (poor Mrs May!)

  44. Chris Lane

    Surely the last time there was such a dramatic reversal of fortunes was with Neville Chamberlain and Winston Churchill

  45. A purely proportional system would have produced:

    Labour 260
    Conservative 275

    However, with a Greek-style top-up of 50 for the winner:

    Labour 260
    Conservatives 325

    Not vastly different to FPTP. Of course, with an actual PR system, the smaller parties would probably have gained higher percentages and the larger 2 lower, so some form of coalition would be needed.

  46. Public opinion towards…

    May:
    Favourable: 29% (-13)
    Unfavourable: 63% (+16)

    Corbyn:
    Fav: 46% (+10)
    Unfav: 46% (-4)

    (via @YouGov, 11-12)

  47. CANDY

    “Someone has put the UK govt up for sale on Ebay”

    ——-

    Have they put any of the pollsters up for sale? (Not Yougov, obvs…)

  48. Do we know anything else about the YG leader polls? Are they weighted like a normal VI poll would be?

  49. Some interesting crosstabs from YouGov’s favourability poll:

    https://d25d2506sfb94s.cloudfront.net/cumulus_uploads/document/ovy5dhg86p/InternalResults_170612_Favourability_W.pdf

    In Scotland:
    Lab: +27 (58:31)
    Con: -40 (25:65)
    SNP: -10 (42:52)

    Might see quite a bit of Lab swing there if there is an election soon.

    Corbyn moves to net 0 (was -14), compared to May 29-30 this is mostly people with no opinion starting to like him.

    May crashes to -34 (was -5) but is still +42 with CON voters in GE17.

    Con is +72 with GE17 Con voters and Lab is +84 with GE17 Lab voters. I don’t think you’d see that much CONLAB movement in England if there is a new election soon.

    But I do estimate that seats that got a lot closer in GE17 compared to GE15 will see increased tactical voting and that a new election would be decided on enthusiasm driving turnout. The party whos’ voters care the least will lose (nothing unusual with that).

    UKIP is liked by 34% of GE17 con voters, but only by 8% of Lab voters. So a warning about a resurgent UKIP (if Tories water down Brexit too much) could steal a lot of voters from them.

1 2 3 4 17