Looking at the Welsh constituency betting

November 17th, 2019

This looks like being a volatile election in many areas, but nowhere is that truer than in Wales. Five different parties are currently polling in double digits and none is yet polling above 30%. Current polls suggest that the current distribution of constituencies could be upended. Polling, of course, could still change dramatically before the election actually arrives.

Historically, Wales has been dominated by Labour. They have won the most seats in every election since December 1918. At the last election, Labour won 28 out of the 40 seats and 49% of the vote.

Perhaps Labour will recover to that vote share but right now that looks unlikely. They last polled at 29%, having mislaid fully 40% of their support since 2017. When parties suffer such dramatic falls, uniform national swing becomes a very dangerous rule of thumb to be using.  In all likelihood the distribution of that loss of support will vary considerably between constituencies.

In 2015 in Scotland, Labour lost 42% of their support from the previous election. But that loss of support differed dramatically between constituencies. In Edinburgh South, Labour actually gained vote share. In Glasgow North East, Labour lost over 50% of its vote share. In some seats in which Labour were not in contention, the loss of vote share was still higher as Labour got squeezed (they lost more than two thirds of their vote share in Ross Skye & Lochaber, for example). 

So those betting on Welsh constituencies need to consider not just what the parties might eventually poll but how that might be redistributed. How efficient each party’s vote is going to be is going to be critical.

Unlike in Scotland in 2015, no one party has taken advantage of Labour’s travails. Relative to the 2017 election, the Conservatives are polling just one point up at 28%. The Brexit party have taken a sizeable chunk – 15%. The Lib Dems are up 7.5% to 12% and Plaid have inched forward 1.5% to 12%.  

Appropriately for a country whose national flag bears a dragon, this polling suggests that the Welsh constituencies are going to be a battle of the five armies (you can decide for yourself who are the eagles and who are the orcs).  Anyone making confident predictions is a lot braver than I am. Too much looks to depend on the distribution of the vote as well as the vote shares.

Anyway, let’s take a butcher’s at the constituencies. Here they are alphabetically. I present these for betting purposes from the perspective of Labour, since they are in contention in most constituencies.  

It is, however, more useful to sort these by the price on Labour, as here.  Immediately, as you would expect, the red mostly rises to the top. Labour are priced at 5/6 (the bookies’ evens) or better in 20 seats. This is actually ahead of what uniform national swing would suggest, which is for them to win 18 seats.   If Labour’s vote were to crash as hard as the most recent polls indicate, you would expect them actually to underperform uniform national swing: the floor in their vote in their weakest constituencies means that the loss of vote share would need to happen elsewhere ie in constituencies with more Labour voters at the last election.

So bettors appear to expect Labour to do better than recent Welsh polling suggests.

As it happens, I agree with the consensus. Labour have already been doing a bit better in UK-wide polling since the election was called and there is no reason to assume that anything different is happening in Wales. And that long muscle memory of voting Labour in Wales is likely to help them when it comes to the crunch.

Even if you disagree, in some of the seats the main challengers are preposterously short. In Ogmore, Labour are priced at 1/5 to retain a 37% majority. In Newport East, they are at just 4/6 to defend a 21% majority.  Perhaps some bettors have inside information. I’m dubious. In both of these seats I’d rather be backing Labour at those prices than the Conservatives at 3/1 and 11/10 respectively. The prices are just too short.

For the same reason, Labour look good value at 3/1 in Vale of Clwyd. They hold the seat with a very experienced MP in Chris Ruane, who is very used to scrapping for votes. The seat is relatively close to Merseyside, which has a Labour party that in recent elections has demonstrated a ferocious ability to work neighbouring constituencies to secure outsized swings as compared with the national figure. He’ll be harder to unhorse than that price suggests.

If you’re looking to bet on the Conservatives, unless you have compelling information, I’d be looking at the size of swing required. The 5/4 in Bridgend looks tempting, given that the Conservatives only require a 5.5% swing and the polls suggest that they’re doing twice as well as that in Wales. With 40% of the vote last time, they might take the seat just by standing still.

There’s no evidence as yet of the Conservatives significantly gaining vote share in Wales, so I would not bet on them in seats where they had a low vote share in 2017 – the votes would have to fall in a very precise way for them to take such seats unless they actually gain vote share. I’ve mentioned Ogmore once. The Conservatives tallied 25% of the vote in 2017. If they’re not going to increase their vote share, 3/1 is ludicrously short for them to win it.

Plaid Cymru haven’t advanced much in the polls so you wouldn’t expect them to be particularly well-placed to make gains.  Ynys Mon is their best prospect, but that is murky, given its long history of voting for the person rather than the party. You’d need more local knowledge than I have to risk your money.

The Lib Dems have their tails up. However, they have few obvious targets in Wales – unsurprisingly since they were wiped out in 2017. If they tally 12% in the polls, that vote will have to go somewhere, but it’s not at all clear where.  They will fight hard to retain Brecon & Radnorshire, which they won earlier this year at a by-election, and they will scrap with Plaid Cymru over Ceredigion (like Ynys Mon, a very personality-driven seat). Elsewhere their best hope in reality is to establish themselves as being back in contention. They held Cardiff Central till 2015 but 11/2 looks very mean for them to get a 24.5% swing from third place. Not a bet for me.

This brings me to the newest party on the block, the Brexit party. So far they’ve had a woeful election, withdrawing candidates from every Conservative-held seat after pressure from the newspapers. They are as a result unlikely to obtain anything like the 15% in Wales that they polled in that last poll. But if they do, that vote is likely to be concentrated in specific seats. The Brexit party performed very well in the Welsh valleys in the Euro elections and UKIP scored fairly well in some of these seats in 2015, demonstrating a continuing strident Euroscepticism in these areas. The bold might fancy a flutter on some of the long shot bets on them in such seats – 80/1 in Llanelli or 50/1 in Torfaen for example. This is 2019. Stranger things have happened.

Alastair Meeks


The latest Ipsos MORI government satisfaction ratings are worse for the incumbent than Major faced just before Blair’s GE1997 landslide

November 16th, 2019

I was so taken by David Herdson’s observation in the previous thread header about how poor the current Ipsos MORI government satisfaction ratings that I thought I would dig into the pollster’s huge archive to see if there were historical precdents.

What’s great about the firm which has been polling UK politics since the 1970s is that it has been asking the same questions in the same format over the decades. You can therefore make comparisons.

Going through the the archive poor negative numbers were indicative of a government that was about to lose power.

The big difference between now and 1997, of course, is that the government ratings then were showing a broad picture that was similar to the voting intention ones. That is not happening now.

What it says about what will happen on December 12th I don’t know but it certainly adds to the overall uncertainty.

  • Just in case you think that the current ratings are an outlier they are the best ratings for the Government since Johnson became PM.

    Mike Smithson

  • h1

    For how long can Johnson continue to defy gravity?

    November 16th, 2019

    He needs to keep running and not look down for four weeks

    Wile E Coyote has enjoyed so many lives that even a cat would feel embarrassed, although perhaps ‘enjoyed’ isn’t quite the right word. Time and again over decades he’s been crushed, burned and fallen from a great height but always to return, unharmed, in pursuit of his great but unattainable aim. With such resilience and obsession, he should have been a politician.

    We should add one other politician’s anti-quality to resilience and obsession: an ability to suspend awareness – and even effects – of life as it appears to everyone else for as long as he concentrates on his own truth. Thus, he only becomes subject to gravity upon the dawning realisation that the ground is no longer beneath his feet (and even then, often with sufficient grace granted to be allowed a parting gesture).

    In truth, gravity is not really all that powerful. When you blu-tac a paper to a wall, for example, you’re overcoming the gravitational power of an entire planet with a small amount of fairly weak adhesive. That said, for us, if not for the Coyote, it is relentless.

    Politics, however, adheres to the principles of Coyotean rather than Newtonian physics: it is entirely possible for a party or a candidate to defy electoral gravity for periods of time, until the forces of cognisance align – which is exactly what Boris Johnson is doing now.

    The October Mori Satisfaction ratings gave Johnson a positive score of +2, a figure which is in line with other polling: YouGov this week found a net favourable rating of -6. These figures, taken against those for Jeremy Corbyn (-60 and -42, respectively), are no doubt a large part of what’s behind the consistent sizeable Tory leads in the polls – five of the last six companies to publish results have put the Con lead in double figures and the exception (ICM), had it at 8.

    If the Tories can deliver a lead of that scale on 12 December, Johnson will almost certainly be back with a comfortable majority. It’s true that the Brexit Party standing in only non-Con seats may make the Tory vote less efficient, piling up larger majorities in seats they already hold, as may the Remain Alliance, but given how super-efficient the Tory vote was in 2017 (very nearly a majority and a lead of 55 seats over Labour with an advantage in vote share of just 2.5%), that should only trim the Con majority, not eliminate it.

    But can the Tories deliver that lead? There are two warning lights flashing in the November gloom that should give us pause for thought, one of which is illuminated with a picture of our friend the Coyote, holding a sign saying “don’t look down”.

    The problem that the Conservatives have is that while Johnson is himself popular, the government is not. The PM’s unusually strong +2 rating has to be set against the net satisfaction score of the government he leads of a fairly awful -55. This is a disparity that cannot endure. Political gravity may be temporarily suspended but it will, sooner or later, come into play: either Johnson will scramble back to the cliff-edge by pulling the government’s rating up towards his own, or else he will plummet chasmwards as opinion turns against him. My firm expectation is the latter because the fundamentals driving that unpopularity are much stronger; the big uncertainty is when it happens.

    The second warning light is Johnson’s own behaviour during the campaign. Time was when Johnson was a considerable asset on the campaign trail: a slightly uncontrolled but exciting bundle of energy and enthusiasm. That time has passed. Johnson is more controlled but also more tetchy. His visit to the flood-affected parts of Yorkshire demonstrated a lack of empathy, compassion or practicality. His inability to make a cup of tea has also been noticed and while that kind of trivial story is usually irrelevant to the bigger picture, sometimes, when it plays to nagging doubts, it can be the sort of thing that helps to tip the scales.

    Other challenges await. His inability to do detail risks becoming a running theme in interviews or, later, in debates, should he turn up – though dodging them would open himself up the accusation of running scared. (This may be why Johnson hasn’t been pushing for Jo Swinson to be included in a 3-way debate, which might otherwise pose far more dangers for Corbyn than for the PM). Then there’s “events, dear boy”. We can predict some known unknowns, like how severe the winter stresses on the NHS will become; others will emerge or strike from out of the blue.

    For the moment, Johnson remains running in mid-air, held up by nothing more than a collective suspension of belief in the power of gravity, and in a disinclination to look down; indeed, a disinclination to think about anything other than making it to 13 December intact. What happens afterwards is for another day.

    In truth, of course, there will be an afterwards and in it he’ll have to keep chasing the road-runner – or, as we know it, Brexit. We know how that ends.

    David Herdson


    Throughout the whole of 2019 every time YouGov has asked its Brexit Tracker question voters those polled have said it’s wrong

    November 15th, 2019


    Election Battlegrounds: Guzzledown

    November 15th, 2019

    Political Betting doesn’t often discuss exhibitions of 18th century art, but I can highly recommend the Hogarth exhibition currently showing at the commendably eccentric Soane Museum in London. It is a very rare opportunity to see in one place all of Hogarth’s masterly sequences of ‘Modern Moral Subjects’, including A Rake’s Progress, A Harlot’s Progress, and Marriage A-La-Mode.

    Most topical, though, is The Humours of an Election (1754-1755), depicting a (literally) hard-fought election in the two-member constituency of Guzzledown, Oxfordshire. In the first painting we see a riotous dinner where the two Whig candidates (one is ‘Sir Commodity Taxem, Bart’) are soliciting votes in joyously dubious ways whilst the Tories parade outside with an anti-Semitic banner (some things were different in those days). The Whig election agent falls back having been hit by a brick thrown through the window. Look carefully and you can see his canvassing returns (‘Sure’ and ‘Doubtful’ – a lot of Doubtfuls but only one Sure). In the third painting we see polling day; the Tory GOTV operation seems to be very effective as the dying and infirm are being dragged up to cast their votes, while the Beadle, who is supposed to ensure that the election is properly conducted, is fast asleep. Meanwhile there is a dispute about whether a soldier veteran, who has lost both hands, can take the necessary oath by placing his hook on the Bible. No-one cares that the coach carrying Britannia is in trouble as the two coachmen are absorbed in playing cards, with one of them cheating. In the fourth painting… well, go and see for yourselves.

    Most relevant to today is the second painting, Canvassing for Votes, shown above. In the centre is a canny voter taking bribes from both sides. The landlady is counting her profits. The Tory candidate, beneath a campaign sign showing money pouring into Guzzledown, is buying trinkets from a pedlar to curry favour with the voter’s daughters on the balcony. At the back the Whig campaign HQ, the Crown Inn, is being attacked in an act of self-harm by a mob indignant at excessive taxes, with one stupidly sawing away at the sign which will crash down with him on it into the crowd below.

    Of course, nowadays our elections are much less crude. We no longer have candidates using their own wealth to bribe voters to support them. Instead we have politicians trying to bribe us with other people’s money, or more subtly, with our own. We have even progressed as far as having politicians successfully bribing us with an entirely fictional number painted on the side of a bus.

    Nonetheless the central problem remains: inasmuch as voters pay any attention, they pay attention to the wrong things, to pedlars’ trinkets rather than real issues. So let me cut to the chase and focus on a few key questions which voters and journalists should be asking of the three main party leaders:

    To Boris Johnson: “You know perfectly well that there is no chance of having a trade agreement agreed and ratified by the end of next year. So which is it: a crash out at the end of next year with the No Deal you’ve sought to avoid, or an extension which you’ve pledged not to ask for?”

    To Jo Swinson: “You know perfectly well that you are not going to be PM. If the Conservatives get a majority, we will have left the EU at the end of January; would you then help Boris implement a trade agreement against the inevitable opposition of the ERG? Or, if the Conservatives don’t get a majority, which of Corbyn and Johnson would you help into No 10, and on what terms?”

    To Jeremy Corbyn, or if you want a more coherent answer, to John McDonnell: “You know perfectly well that all the spending commitments you’ve made aren’t a programme for government, they are an unfiltered and unsorted wish-list of eye-wateringly expensive suggestions. Which are your priorities, and which are just fantasies?”

    Don’t expect answers, but if you are in London, do go to the Hogarth exhibition. It is on until the 5th January. Top tip: take a magnifying glass.

    Richard Nabavi


    Focus on Wokingham where two ex-CON MPs are slugging it out against each other

    November 15th, 2019

    The longstanding Brexiteer versus the defector to the LDs

    The Berkshire constituency of Wokingham is one of the most intriguing battle grounds at the General Election for it is where the veteran Brexiteer, John Redwood, is seeking to defend his seat against the remain backing former CON MP from the neighbouring constituency, Philip Lee. Redwood has been MP there since 1987.

    It will be recalled that when Parliament resumed in September it was while Johnson was speaking that Phillip Lee crossed the floor of the House of Commons from the Tory benches to sit next to Jo Swinson. His defection left the Tories with no working majority in the House of Commons

    Lee and Redwood, of course, have totally different views of the EU with the former being closer to voters in Wokingham than the latter. At the referendum Wokingham went 57% remain.

    This is a seat which is part of the Unite to Remain alliance and there is no Green standing. The key importance of this is that it is a signal to LD activists outside the area that it is a key battle ground and where they should help.

    There was a small sample Survation poll which had Redwood 4% ahead.

    In Euro2019 the LDs came top in Wokingham with 33.45% with the Greens getting 11.3%. The Tories were estimated to have got 13.4% with BP at 29.4%. LAB was at 4.9%

    The question here is whether Redwood has built up support over the past third of a century there to get re-elected. Everything depends on the importance of Brexit to those who voted for Redwood at GE2017.

    Ladbrokes currently make the Tories 1/4 favourites with the LDs at 11/4 which I regard as value. I’d assess this as closer to 50/50.

    Mike Smithson


    The voting polling’s bad for LAB but Corbyn’s ratings are even worse

    November 15th, 2019

    Above is the Wikipedia list of all the published polls since the general election campaign began. The overall picture is of not that much variation with the Tories in a range of 37-42%,  LAB 27-31% and the LDs 15-17%.

    The one party where there’s a lot of variance is Brexit which has a polling range of 4-10%. That’s largely explained by YouGov’s methodology change that factors in the fact that Farage’s party will be only be contesting non-CON seats.

    Generally I’m not a great fan of voting intention polling as a means of getting a feel for how things are going – leader ratings have historically been a better guide.

    What’s  pleasing about the current election is that we are getting a wider range of regular leader ratings than we’ve seen before.

    The weekly Deltapoll have joined Ipsos-MORI and Opinium in always including a ratings element. Ipsos MORI has also added a favourability question.

    Swinson has the the largest number, 33%, saying don’t know and that should get smaller during the campaign. Johnson tops on favourability while Corbyn has the most saying they have an unfavourable view. This is in line with other pollsters and other question formats.

    Mike Smithson


    The GE2019 podcast from Keiran Pedley – now with Ipsos MORI

    November 14th, 2019

    Keiran Pedley, familiar to PBers through his longstanding PB/PollingMatters podcast series is now doing a general election series for the pollster, Ipsos MORI, with home he works. Each week he’ll inviting a range of figures to add to our understanding of what’s happening. This is the first.

    I am sure we all wish Kieran the best of luck.


    Mike Smithson