United Kingdom General Elections: December 12th, 2019

[TheDeadFlagBlues]

Even if the LibDems collapse over the course of the campaign, the distribution of their vote could be much less uniform than is usually the case, confounding our expectations and resulting in an outcome that none of us anticipate. In the event that the LibDems actually receive over 20% of the vote, I'd expect that result to be very distinct from past elections where LibDems received a similar vote share. Thus, this election is kind of a "black box" for long-time followers of British politics - you can punch the vote shares for each party into a seat calculator and the seat count could be wildly off!

I don't want to overempashize this because the present government is extremely unpopular, granting the possibility that the LibDems could very well appeal to their traditional electorate, which is far from being uniformly "eurosceptic" anyways, but even the possibility of a strategically-distributed LibDem electorate emerging strikes me as being very consequential

[TheDeadFlagBlues]

Nigel Farage throwing a lifeline to the remainers by announcing that they are going to stand in all constituencies.

Not 100% sure about that.  For sure BXP running everywhere will eat into CON Brexi votes but could also eat into LAB Brext voters that otherwise would never vote CON anyway.  The main problem with a de facto CON-BXP alliance is that it will trigger an even greater defection rate of CON Remain vote as well as trigger LAB-LDEM-Green tactical voting in response.

The most recent Yougov poll on a cross section of 2017 and 2016 vote is fairly instructive:

Voting intention among 2016 REMAIN voters
Lib Dem - 34%
Labour - 33%
Con - 16%
Green - 9%

Voting intention among 2016 LEAVE voters
Con - 58%
Brexit Party - 24%
Labour - 10%

LEAVE voters only
Voting intention among CON+LEAVE voters
Con - 77%
Brexit Party - 19%

Voting intention among LAB+LEAVE voters
Labour - 43%
Brexit Party - 25%
Con - 19%



REMAIN voters only
Voting intention among CON+REMAIN voters
Con - 61%
Lib Dem - 31%

Voting intention among LAB+REMAIN voters
Labour - 58%
Lib Dem - 27%
Green - 9%



There are almost no CON REMAIN vote that will defect to LAB and all such defection will go to LDEM.
The some LAB LEAVE vote will defect to CON but an even larger chunk will go to BXP.  That is the bloc of LAB LEAVE voters that will most likely never vote CON anyway and BXP running will keep those votes away from LAB.  Of course there is a bloc of CON LEAVE voters that will go to BXP but like in 2015 CON can hope for tactical voting.

Even if we are to take the Yougov estimates as given, it's hard to see how the Brexit Party focusing on "Labour heartland" seats benefits them. Let's do a little exercise. Assume that all Tory voters and UKIP voters in 2017 backed Leave (dubious and favorable to Labour but more close to being accurate than not in "Labour heartland" seats), that all LibDem/Green voters backed Remain, and use this to estimate the Remain/Leave share of the Labour vote. We can then applying these vote share estimates. I simply add the old LibDem and UKIP vote shares to the YouGov-based vote transfers for LAB/COn.

Doing this in Redcar (~68% Leave):
Labour - 55.5% -> Labour Leave 30%, Labour Remain 25.6%
Conservative - 33.2% -> Tory Leave 33.2%
LibDem - 6.7%
UKIP - 4.6%

Vote Estimates with Brexit Party (no tactical squeeze, just YouGov):
Conservative 31.2%
Labour 27.7%
Brexit 18.3%
LibDem 13.6%

However, let's assume a tactical "squeeze" among Labour remain voters, giving Labour ~90% of these voters but no tactical squeeze to benefit Tories because Brexit resources ensure voter confusion:
Labour 35.9%
Conservative 31.2%
Brexit 18.4%
LibDem 8%

Now, all of this is very "problematic", resting on a bunch of absurd assumptions but the idea is to illustrate the problem of the Brexit Party strategy. Even in Leave-heavy Labour heartland seats, something like ~40% of Labour 2017 voters backed remain - vast majority will back Labour again, even if they are annoyed, because they'll be squeezed. However, intervention of Brexit Party makes the same "squeeze" strategy for the Tories difficult - it'll be unclear which party is best positioned to win.

Let's assume that there is an accidental squeeze to Brexit that's more concentrated among Labour leave voters than Tory voters (we give 35% among Labour Leavers, reducing Tory share to 9% with this group and 25% among Tory leavers and 75% to Tories), holding all else constant:
Labour 35.9%
Conservative 27.6%
Brexit 23.3%
LibDem 8%

Such an accidental squeeze is possible (perhaps likely and seeing as it as a squeeze may be a mistake when one remembers how many "Tory Leave" types in Redcar or Rother Valley are just UKIPers who loaned their vote to the Tories last time!) in many of these seas and notice what it does - it produces a much larger Labour majority! Perversely, insofar as Brexit Party wants to coordinate with the Tories, its best strategy might be to abolish itself. Insofar as it wants to be independent and does not care about the Tories, it should probably target as many marginal Labour/Tory seats, as "Labour heartland" seats!

[TheDeadFlagBlues]

It's important for everyone to be aware that, per some unfathomable tradition, polling firms in the UK do not report toplines with undecided voters, as they do in the US. This creates the impression of the Tories having a substantial lead when, in fact, they do not.

As an example, on October 25th, Opinium released a poll showing: Con 40%, Labour 24%, LibDem 15%, Brexit 10%. With "Don't Knows" included, we have: Con 31%, Labour 21%, Don't Know 16% LibDem 12%, Brexit 9%.

Here are the share of "Don't Knows" by 2017/2016 voting history:
Conservative Remainers - Don't Know 22%
Conservative Leavers - Don't Know 9%
Labour Remainers - Don't Know 14%
Labour Leavers - Don't Know 18%

Backing this out, the Don't Knows in this opinion poll voted the following way in 2017 and 2016:
EU Party Referendum Groups - Labour Remain 32%, Conservative Remain 25%, Conservative Leave 24%, Labour Leave 19%
2017 GE - Labour 49%, Conservative 43%, LibDem 4%
2016 EU Referendum - Remain 53%, Leave 47% (Remain 42%, Leave 38%, Didn't Vote 19%)
Remain voters by party - Labour 53%, Conservative 41%
Leave voters by party - Conservative 53%, Labour 43%

Here's what the "Don't Knows" think about the leaders:
Preferred Prime Minister - None/Don't Know 65%, Boris Johnson 31%, Corbyn 3%,
Boris Johnson (excluding people without a view) - Disapprove 61%, Approve 39%
Jeremy Corbyn - Approve 25%, Disapprove 75%
Jo Swinson - Approve 42%, Disapprove 58%

Is performing this sort of analysis on the disgusting entrails of a polling sausage a bit misguided? Sure but I hope it's a reminder of certain facts: polls can whiff in dramatic ways when "undecided" voters are left unreported, there are many undecided voters and those undecided voters have a clear lean. If I had to guess, the undecided voters will gravitate towards Labour/LibDems throughout the campaign.

[TheDeadFlagBlues]

It's important for everyone to be aware that, per some unfathomable tradition, polling firms in the UK do not report toplines with undecided voters, as they do in the US.

Because "Don't know" is not an option on the ballot paper on election day.

Neither is "undecided" but it's typical for American pollsters to report this information because it gives a sense of the actual size of a lead. We all should have known that Hillary Clinton's lead was tenuous in 2016 because there are more undecideds and third party voters than usual. If you simply removed undecideds, she would have had an even larger lead in polls!

[TheDeadFlagBlues]

More polls...

Survation: Con 35, Lab 29, LDem 17, BP 10, Greens 1, Others ?*
ICM: Con 39, Lab 31, LDem 15, BP 8, Greens 3, SNP 3, Others 1

_{*Note that Survation polls include Northern Ireland, so mentally change those figures to 36, 30... for comparative purposes.}

It's certainly tightening up.

YouGov have just reported a 14 point lead for the Tories, so some of it is just about the differing house effects of the various pollsters.

At the national level, Survation knows what it is doing and YouGov does not.

[TheDeadFlagBlues]

Jermey Corbyn is so bad that he makes Ilhan Omar look good compared to him. Corbyn needs to lose in a landslide

Wow, thank you for the insight, Old School Republican, very cool!

Can we please deport people who know nothing about British politics from the thread?

[TheDeadFlagBlues]

Anyways, I think the debate went well for Jeremy Corbyn and serves as a reminder that Boris Johnson being an incredibly polarizing figure is hardly an asset when the aim of the Conservative Party is winning a majority - a repeat of 2017 won't do and a one-on-one debate with Corbyn was always going to push many people towards Labour.

Jeremy Corbyn could have wiped the floor with Boris Johnson if instead of evading his question, he had remarked that he wouldn't campaign in the referendum because he wants "the people to decide". This would still constitute a kind of evasion but would be more clever and would stand as a nice foil to Boris Johnson's polarizing rhetoric.

[TheDeadFlagBlues]

Jeremy Corbyn is British. Why would he know how Americans pronounce German surnames? The way we butcher these surnames is specific to the US, similar to how British people have a knack for mangling Spanish words and surnames.  

[TheDeadFlagBlues]

I have to say, it's really not a good look for the media to go after Corbyn for supposedly mispronouncing a name while he was pledging to defend the victims of Epstein's sordid pedophile ring. It's more than a bit anti-Semitic to assume that disdain for a literal pedophile is some sign of an anti-Semitic bent...

[TheDeadFlagBlues]

So there's this forecast site:

https://leantossup.ca/uk-constituency-map/

CON GAIN Torfaen anyone?

Even better: they have Swansea East (!!!!!!!!!) down for that as well. What the actual Christ.

Barnsley East: Tossup

uh okay?

[TheDeadFlagBlues]

Having fun messing around with the "Best for Britain" MRP forecast, noticed an absurd prediction for Rhondda: Labour 40%, PC 23%, Conservative 19%.

This would be the best Tory performance in Rhondda since *checks notes* 1923? Maybe such a performance is possible now, given the slow demise of local peculiarities, but it seems to me that we ought not expect to this to happen and that many of these MRP predictions don't pass the smell test. In order for Labour to plunge ~25 percentage points and for the Tories to increase their vote share by 7 percentage points, we'd expect many Labour voters to flip to the Tories in this constituency and there's little indication that this will happen. Also, it's worth emphasizing that this would be the worst Labour performance in Rhondda since, uh, 1910? Anything can happen, I could have sex with the Pope in theory, but anyone who expects this to happen is insane imo.

[TheDeadFlagBlues]

Remember that all this is is a demographic modelling tool that projects a large national survey down to constituency level. It relies heavily on a series of assumptions, and if any are even a little bit out then the projections will be wildly off. It's clever, but no more inherently accurate than doing a quick swing calculation and checking out the constituency pendulum.

Or, rather... everyone remembers that the model called Canterbury correctly last time. Everyone forgets its projected figures for Finchley & Golders Green...

Something to keep in mind also: this poll is a snapshot of time and things may move quickly over the next week. If there's a swing back to Labour (or away from Labour), it may not be uniform. Relying on MRP to think about the election is seductive for political geography enthusiasts, we all want to see detailed poll results, but you can use common sense and imagine similar predictions less the absurd outliers that no one should anticipate, even if they excite us on some level. Calm down, use your head folks!

[TheDeadFlagBlues]

Boris Johnson's ghoulish reaction to the image of the boy in the hospital is about the worst possible story for the Tories in the most critical period of the campaign because it puts the spotlight on the NHS' problems while suggesting that Boris Johnson will do nothing to solve them. Considering that 8-12% of voters remain Undecided and that the Tory lead is ~8 points when Undecideds are taken into account, this could be devastating for the Tories...

Seems like the sort of story that cuts through for a week tbh. I expect the Tories to try to slam a bunch of rats on the table to change the topic. The extent to which they do this will reveal how damaging the story is.