UK By-elections thread, 2021-

[ElectionObserver]

How credible is this?
https://cdn.survation.com/wp-content/uploads/2021/05/03181731/image-134.png

I think that it's a reasonable poll as constituency polls go, but it isn't hugely credible.

A few thoughts...
- Constituency polling in the UK has a terrible record. Some of them have been wildly inaccurate. That's due to many reasons, but small seat size and lack of experience by pollsters are two of the big ones.
- It's a poll of 301 people. That's a very small sample which is liable to a large margin of error even in a well conducted poll (+/- 5%). They actually polled 517 people. Then (unweighted) 464 people said they would vote, but 125 of them were 'undecided' and 32 'refused'. In total only 142 said they'd vote Conservative and 99 Labour.
- Survation only called landlines. Think who you know who answers landlines. Only my elderly grandmother would answer the phone and talk to a pollster. It's a very limiting, almost self-selecting, group.  It's now becoming more and more common for people to abandon landlines which I think would skew young and low income.
- This poll skewed pretty old in respondent age. Look at the tables. 57 people aged 18-34 responded which they weighted to 103. 145 responded who were 65+ which they weighted to 122. They weighted up the under 45s and down the 45+ age groups.
- It also skews wrong by party. They polled 124 2019 Labour voters, 155 2019 Conservative voters, 62 2019 Brexit Party voters and 32 others. Now they have weighted the numbers, but the raw unweighted sample is quite out-of-sync with the 2019 result when Labour got 39%, Conservatives 29%, Brexit 26% and others 8%.
- The data collected is a little dated at a week and a half old.

That said, I wouldn't like to be Keir Starmer or the Labour candidate right now.

[ElectionObserver]

Though the likeliest Tory gain in Wales isn't anywhere in the N, but the Vale of Glamorgan.

Missing out on that could be a sign they are failing to meet expectations overall.

How Labour have clung on so many times, I don't know. It's been a marginal in every election since 1999, with 2007 and 2016 being so incredibly close. It has been held by the Conservatives in Westminster since 2010. Though the Tories have never had a big majority in Westminster and it was actually pretty small both 2017 and 2019.

IMO the Conservatives best chances in order are
1 - Vale of Clwyd.
2 - Vale of Glamorgan.
3 - Brecon and Radnor.
4 - Wrexham.
5 - Clwyd South.
Three in the north, one the south and one the middle. The first three I think are likely gains, but I am not too sure after that.
Their most vulnerable seat, Aberconwy, is also in the north though.

[ElectionObserver]

We shall see. How many people will be caught out by the fact they need to bring their own pencil to the polling station?

You don't need to. It's just recommended.

For those who don't bring a pen, rather than having pencils in the voting booth, pencils will be given to voters and disinfected between each use.

People will also be asked to use alcohol gel on entry and exit, to wear a mask, and to maintain social distancing in polling stations.

[ElectionObserver]

It looks like Starmer's "insiders" have moved on from Hartlepool to start dooming about a byelection in Batley and Spen that hasn't even been made official yet

Are we expecting Brabin to go?

In Hartlepool, it will certainly be much closer than that poll predicted, but I still think the Tories will gain it by a couple of points.

[ElectionObserver]

So it appears Brabin has to go if elected mayor. Not because of the mayoral role, but because it includes the PCC powers.

https://twitter.com/SebastianEPayne/status/1390222580414656516

[ElectionObserver]

I was unsure whether Labour would hold this seat or not, but if I knew that Galloway would take over 20% of the vote, I would have had this down as a certain Conservative gain. They did incredibly well to hold on given that.

From what I have seen, Labour just ran a far superior ground game to the Conservatives who were far too complacent (as we saw the other week in Chesham and Amersham).

They also allowed the narrative that they would win to become the mainstream one. That meant a Labour hold, no matter how narrow, was incredibly good for Labour. Starmer woiuld have been in an exceptionally difficult position if they had lost, though I think he would have clung on.

[ElectionObserver]

Yes, Old Bexley and Sidcup is ancestrally Tory, but in a *very* different way to C&A.

The Tory protest vote is most likely to sit at home rather than consider another party. The poor weather in this part of the world today won't have helped.
I don't think ReformUK will get to double digits, but they should save their deposit reasonably comofrtably. It's good demographically with it being leave, very white, and well-off working class. They've also run a reasonably high profile campaign, but the party is a shell of its former Brexit self.
The Lib Dems have no history here and it isn't fertile ground at all (not as wealthy, educated or remain as C&A). They'll likely end up a point or two short of saving their deposit.
I'm interested to see Labour's numbers. There won't be many switchers, but a score above the 29.3% they got in 2017 would be very reasonable and it'd be an excellent night if they got to 1997's 35.1% no matter what the pundits say. I think the latter is quite unlikely, but it's quite dependent on how well RefomUK do.

[ElectionObserver]

Turnout 34.8% with 21,788 votes cast. It was 69.8% and 46,145 at the 2019 General Election.

People watching the count estimate a 3,000 to 5,000 Conservative majority which would be a huge fall from their 2019 majority of 18,952 in numerical terms. In terms of share, it would mean a drop from 41.1 points in 2019 to 14-23 points.