United Kingdom General Election: July 4, 2024

[Joe Republic]

Survation has the Tories on 0(!) seats in Wales. Is this a live possibility or another example of MRP oddities?

See: 1997

[adma]

Survation has the Tories on 0(!) seats in Wales. Is this a live possibility or another example of MRP oddities?

Realistic. I think they will hold Montgomeryshire but a wipeout is very plausible and Montgomeryshire is the only seat they are at decent odds to hold on current polling. Maybe Brecon at a push but Labour seems to have given the LDs a free run there so the LDs should win it with tactical votes.

Even with Montgomeryshire's own LD history?  Or did Lembit Opik stink up the joint *that* much?

[adma]

Nobody has yet mentioned Survation giving the SNP 37 seats.

Is this simply because they are too busy laughing at it?

It’s because I’m laughing at the name “Great Grimsby,” it sounds so close to Great Gatsby.

I think it sounds more like a comic-book exclamation.  "Great Grimsby!  Look at that fireball hurtling towards us!"

[Kalimantan]

That MRP is honestly weird and just looking through a few Northern seats it looks like Reform is taking sizable votes off of Labour. For instance, in Great Grimsby Labour is at 39%, the Tories at 29% and Reform at 23%, Labour received 33% of the vote there in 2019 and I am not inclined to believe that tons of Tory->Labour switchers jumped to Reform.

This is Redwall territory, where in 2019 huge numbers of Labour voters jumped ship and went to Boris on promises of Brexit and levelling-up and whatnot. Those voters will be heading across to Reform in their droves, and there will probably be more that join them. Expect Labour to sweep the board here with the Tory/Reform vote split, but there could be some quite low vote totals winning seats, and its not impossible some good local campaigns by Reform candidates can take the odd one. Grimsby might be prime territory for that.

[Philly D.]

Survation has the Tories on 0(!) seats in Wales. Is this a live possibility or another example of MRP oddities?

Realistic. I think they will hold Montgomeryshire but a wipeout is very plausible and Montgomeryshire is the only seat they are at decent odds to hold on current polling. Maybe Brecon at a push but Labour seems to have given the LDs a free run there so the LDs should win it with tactical votes.

Even with Montgomeryshire's own LD history?  Or did Lembit Opik stink up the joint *that* much?

It comes down to the LD by-election victory in Brecon in 2019. The LD vote in Wales is about 6%, which even with Reform, is too low to be competitive in more than one seat.

On another note, there is next to no polling below GB level (is this normal?). I don't think there's been a single Scotland-only poll! I'd also like a Greater London poll -- the giant R&W poll at the start had the Tories holding on the Chingford on a UNS!!!

[Kalimantan]

Survation has the Tories on 0(!) seats in Wales. Is this a live possibility or another example of MRP oddities?

Realistic. I think they will hold Montgomeryshire but a wipeout is very plausible and Montgomeryshire is the only seat they are at decent odds to hold on current polling. Maybe Brecon at a push but Labour seems to have given the LDs a free run there so the LDs should win it with tactical votes.

Even with Montgomeryshire's own LD history?  Or did Lembit Opik stink up the joint *that* much?

Don't forget Montgomeryshire is the seat where the incumbent Tory candidate, Sunak's PPS in the last parliament, placed a bet on the election date being in July and is now under investigation by the gambling commission. Sleaze can bring down any candidate anywhere.

[adma]

Survation has the Tories on 0(!) seats in Wales. Is this a live possibility or another example of MRP oddities?

Realistic. I think they will hold Montgomeryshire but a wipeout is very plausible and Montgomeryshire is the only seat they are at decent odds to hold on current polling. Maybe Brecon at a push but Labour seems to have given the LDs a free run there so the LDs should win it with tactical votes.

Even with Montgomeryshire's own LD history?  Or did Lembit Opik stink up the joint *that* much?

It comes down to the LD by-election victory in Brecon in 2019. The LD vote in Wales is about 6%, which even with Reform, is too low to be competitive in more than one seat.

On another note, there is next to no polling below GB level (is this normal?). I don't think there's been a single Scotland-only poll! I'd also like a Greater London poll -- the giant R&W poll at the start had the Tories holding on the Chingford on a UNS!!!

In which case, I have a measured skepticism about that generic Wales-wide LD share.

[adma]

That MRP is honestly weird and just looking through a few Northern seats it looks like Reform is taking sizable votes off of Labour. For instance, in Great Grimsby Labour is at 39%, the Tories at 29% and Reform at 23%, Labour received 33% of the vote there in 2019 and I am not inclined to believe that tons of Tory->Labour switchers jumped to Reform.

This is Redwall territory, where in 2019 huge numbers of Labour voters jumped ship and went to Boris on promises of Brexit and levelling-up and whatnot. Those voters will be heading across to Reform in their droves, and there will probably be more that join them. Expect Labour to sweep the board here with the Tory/Reform vote split, but there could be some quite low vote totals winning seats, and its not impossible some good local campaigns by Reform candidates can take the odd one. Grimsby might be prime territory for that.

And as far as low vote totals go--let's remember that the more enduring byproduct of the Red Wall paradigm shift might be in how even in the event of a Keirslide, the Blair-era astronomic Lab shares in many of these Red Wall seats aren't going to be seen again.  38% is the new 58%, IOW.

But a different twist to Great Grimsby: it's a place where the Tory incumbent's reoffering, and the former Labour MP's attempting a comeback, and who knows whether it could be a Tory-Labour vote split electing Reform instead--but actually, there's a *different* potentiality to consider, which once again echoes Canada '93: a record number of Tory incumbents finishing 3rd, and once again most likely among the '19 Red Wall scalp-claimers...

[Tintrlvr]

Survation has the Tories on 0(!) seats in Wales. Is this a live possibility or another example of MRP oddities?

Realistic. I think they will hold Montgomeryshire but a wipeout is very plausible and Montgomeryshire is the only seat they are at decent odds to hold on current polling. Maybe Brecon at a push but Labour seems to have given the LDs a free run there so the LDs should win it with tactical votes.

Even with Montgomeryshire's own LD history?  Or did Lembit Opik stink up the joint *that* much?

Don't forget Montgomeryshire is the seat where the incumbent Tory candidate, Sunak's PPS in the last parliament, placed a bet on the election date being in July and is now under investigation by the gambling commission. Sleaze can bring down any candidate anywhere.

I didn’t connect the dots on this.

Anyway the LDs were weaker in Montgomeryshire than Brecon pre-redistricting, and the new areas of the Montgomeryshire seat are more Tory than the new areas of the Brecon seat (where the new areas are solid Labour normally and could vote LD tactically). And as others noted the Welsh LDs don’t really have the resources to contest both. If Montgomeryshire goes to anyone it will be Labour.

[Sol]

No, just no. If you have to make an American comparison to Grimsby, then it would be some working class coastal town in like Maine or Washington, which coincidentally is the sort of place Biden generally bounced back a bit in after the 2016 apocalypse.

If it is analogous to Maine, I would say more so Maine's 2nd Congressional district rather than the 1st (which is more progressive). But Trump got a higher vote share in 2020 than 2016 there, so that doesn't really fit your argument at all. ME-02 is precisely a place where Democrats traditionally won and where Trump improved with WWC voters.

Washington (assuming you mean Washington State) is not really a WWC area, it is socially progressive Pacific coast, and Clinton did fine there in 2016.

It is true that in both Washington and Maine there were relatively high 3rd party vote shares in 2016, but those 3rd party voters are more analagous to UK Green and UK Lib Dem voters than to the sorts of disaffected voters who would be pro-Brexit or pro-Reform in the UK or pro-Trump in the USA.

There's already a thread for this kind of irrelevant speculation, and it isn't here.

[YL]

Survation has the Tories on 0(!) seats in Wales. Is this a live possibility or another example of MRP oddities?

Realistic. I think they will hold Montgomeryshire but a wipeout is very plausible and Montgomeryshire is the only seat they are at decent odds to hold on current polling. Maybe Brecon at a push but Labour seems to have given the LDs a free run there so the LDs should win it with tactical votes.

Even with Montgomeryshire's own LD history?  Or did Lembit Opik stink up the joint *that* much?

Don't forget Montgomeryshire is the seat where the incumbent Tory candidate, Sunak's PPS in the last parliament, placed a bet on the election date being in July and is now under investigation by the gambling commission. Sleaze can bring down any candidate anywhere.

Also, it's Montgomeryshire & Glyndŵr. The component rather absurdly called "Glyndŵr" is a substantial part of the constituency, and is a reasonably Labour area, unlike Montgomeryshire.

(That said, I'm not taking any of the details from the Survation MRP seriously.)

[YL]

I'm definitely not knowledgeable enough about previous MRPs to understand or offer an answer to this, but I wonder if the lack of a Leave/Remain dynamic (and the demographic factors aligned with it) compared to 2017, and 2019 in particular, makes these things quite a bit more unwieldy. Is it at all telling that the rise of MRPs benefitted from being at a time when we had another major and relevant electoral data point?

I think there may be something to this. An issue election - and the previous two both were - will always force binaries onto the electorate on an extent, and that ought to make outcomes easier to model as you have a pretty good reference point to start from. Take that away and things are much more chaotic.

MRPs in general weren't really that great in 2017 and 2019 though, other than the 2017 YouGov one which both got close on the headline figures and was fairly good on the pattern of the swings (including the famous Canterbury prediction). YouGov's 2019 MRP was also quite good on the patterns (though as in 2017 some local effects were missed and some were overcooked, e.g. East Devon) but it wasn't that close to the overall result.

[Blair]

On the subject of Great Grimsby & vote share it's worth remembering Labour actually won a fair few council seats in the area (I'm 99% sure- it's part of North East Lincolnshire council but my map reading suggests grimsby had gains)

I can't see them doing worse than 2019 on vote share- we had a BXP candidate in 2019 and they got 7% of the vote.

[Blair]

Worth remembering with MRP that YouGov's final one in 2019 was relatively poor; it got the vote shares roughly right but had Labour at around 235 seats.

It lead to a bit of fabled belief that Labour might overperform and keep the Tory majority low.

Of course the big problem as I said before in another post is that there's so many moving parts at this election; a falling tory vote, a relatively high amount of Lab-Con switchers, tactical voting, the reform vote and in a fair chunk of seats we could see at least 3 parties get north of 25%.

[TheTide]

Worth remembering with MRP that YouGov's final one in 2019 was relatively poor; it got the vote shares roughly right but had Labour at around 235 seats.

It lead to a bit of fabled belief that Labour might overperform and keep the Tory majority low.

Of course the big problem as I said before in another post is that there's so many moving parts at this election; a falling tory vote, a relatively high amount of Lab-Con switchers, tactical voting, the reform vote and in a fair chunk of seats we could see at least 3 parties get north of 25%.

I'm pretty sure no more than three parties can get more than 25% of the vote. 

[CumbrianLefty]

Is that really true re Brecon? As with Montgomery it expands to take in some more pro-Labour areas.

[President Punxsutawney Phil]

Survation has the Tories on 0(!) seats in Wales. Is this a live possibility or another example of MRP oddities?

See: 1997

Re: 1997, the Tories only won twelve seats north of Lincolnshire and none outside England.
They still won probably more land area than Labour though, somehow.
This time around history might be made there, in that department...

[TheTide]

The PA don't seem to have released their estimated declaration times yet. They appear to have released them for the local elections (2nd of May) on the 20th of April, so we should be getting them soon.

[adma]

Worth remembering with MRP that YouGov's final one in 2019 was relatively poor; it got the vote shares roughly right but had Labour at around 235 seats.

It lead to a bit of fabled belief that Labour might overperform and keep the Tory majority low.

Of course the big problem as I said before in another post is that there's so many moving parts at this election; a falling tory vote, a relatively high amount of Lab-Con switchers, tactical voting, the reform vote and in a fair chunk of seats we could see at least 3 parties get north of 25%.

I'm pretty sure no more than three parties can get more than 25% of the vote. 

In an election where the LD vote is *heavily* tactical, it's a possibility--particularly in those Great Grimsby-esque Leave seats w/a class-of-'19 Tory incumbent (indeed, UKIP *did* manage 25% 3rd in GG in '15, and that's w/Labour close to 40%).

Speaking of GG & vote share, keep in mind that even in her 22-point '19 loss, Melanie Odd got the exact same 32.7% as Austin Mitchell in his final 2-point *win* in '10.

[TheTide]

Worth remembering with MRP that YouGov's final one in 2019 was relatively poor; it got the vote shares roughly right but had Labour at around 235 seats.

It lead to a bit of fabled belief that Labour might overperform and keep the Tory majority low.

Of course the big problem as I said before in another post is that there's so many moving parts at this election; a falling tory vote, a relatively high amount of Lab-Con switchers, tactical voting, the reform vote and in a fair chunk of seats we could see at least 3 parties get north of 25%.

I'm pretty sure no more than three parties can get more than 25% of the vote. 

In an election where the LD vote is *heavily* tactical, it's a possibility--particularly in those Great Grimsby-esque Leave seats w/a class-of-'19 Tory incumbent (indeed, UKIP *did* manage 25% 3rd in GG in '15, and that's w/Labour close to 40%).

Speaking of GG & vote share, keep in mind that even in her 22-point '19 loss, Melanie Odd got the exact same 32.7% as Austin Mitchell in his final 2-point *win* in '10.

I might be having a moment and misunderstanding the quoted post, but I'm pretty sure it is not arithmetically possible. 25 + 25 + 25 + 25 = 100.

[adma]

Worth remembering with MRP that YouGov's final one in 2019 was relatively poor; it got the vote shares roughly right but had Labour at around 235 seats.

It lead to a bit of fabled belief that Labour might overperform and keep the Tory majority low.

Of course the big problem as I said before in another post is that there's so many moving parts at this election; a falling tory vote, a relatively high amount of Lab-Con switchers, tactical voting, the reform vote and in a fair chunk of seats we could see at least 3 parties get north of 25%.

I'm pretty sure no more than three parties can get more than 25% of the vote. 

In an election where the LD vote is *heavily* tactical, it's a possibility--particularly in those Great Grimsby-esque Leave seats w/a class-of-'19 Tory incumbent (indeed, UKIP *did* manage 25% 3rd in GG in '15, and that's w/Labour close to 40%).

Speaking of GG & vote share, keep in mind that even in her 22-point '19 loss, Melanie Odd got the exact same 32.7% as Austin Mitchell in his final 2-point *win* in '10.

I might be having a moment and misunderstanding the quoted post, but I'm pretty sure it is not arithmetically possible. 25 + 25 + 25 + 25 = 100.

Oh, I get it; the "at least" part ;-)

[Torrain]

Nobody has yet mentioned Survation giving the SNP 37 seats.

Is this simply because they are too busy laughing at it?

Very random seats at that: every glasgow seat is yellow while Labour gains Livingston and Falkirk.

The SNP’s Toni Giuliano has made a habit of losing winnable races in significant upsets for the party (2016, 2017, 2021), but even he’d struggle to lose Falkirk while the party holds 37 seats.

It’s just implausible - honestly not sure how you get those numbers. Some of the data looks like it’s trying to apply national swing to 2015, some 2017, and there’s not really much at the council level that explains it.

If they’d done something tenuous, like plug in the 2021 Holyrood list results, you’d see a bigger swing against the SNP in Central Scotland (Falkirk) than Glasgow, and poorer results for the SNP versus Labour in Lothian (Livingston). But it shouldn’t be anywhere near this stark.

[Blair]

Worth remembering with MRP that YouGov's final one in 2019 was relatively poor; it got the vote shares roughly right but had Labour at around 235 seats.

It lead to a bit of fabled belief that Labour might overperform and keep the Tory majority low.

Of course the big problem as I said before in another post is that there's so many moving parts at this election; a falling tory vote, a relatively high amount of Lab-Con switchers, tactical voting, the reform vote and in a fair chunk of seats we could see at least 3 parties get north of 25%.

I'm pretty sure no more than three parties can get more than 25% of the vote. 

Ha poor phrasing- I meant just three!

[afleitch]

Nobody has yet mentioned Survation giving the SNP 37 seats.

Is this simply because they are too busy laughing at it?

Very random seats at that: every glasgow seat is yellow while Labour gains Livingston and Falkirk.

The SNP’s Toni Giuliano has made a habit of losing winnable races in significant upsets for the party (2016, 2017, 2021), but even he’d struggle to lose Falkirk while the party holds 37 seats.

It’s just implausible - honestly not sure how you get those numbers. Some of the data looks like it’s trying to apply national swing to 2015, some 2017, and there’s not really much at the council level that explains it.

If they’d done something tenuous, like plug in the 2021 Holyrood list results, you’d see a bigger swing against the SNP in Central Scotland (Falkirk) than Glasgow, and poorer results for the SNP versus Labour in Lothian (Livingston). But it shouldn’t be anywhere near this stark.

There is a strong correlation between regional crosstabs in the national polling and constituency level results for regions in 2007. Glasgow is better for the SNP as is the Highlands and the North East (and there the Tories are doing much worse).

Polling and modelling might be struggling somewhat to deal with the demographic shift from a Surgeon/Yousaf to Swinney/Forbes SNP. I think the worse polling for the SNP was capturing that transition which is why, even with different methodologies, it's been marginally better of late.

Still, MRP generally is more bullish for the SNP. Though it stands to reason that if Labour are dipping or 'settling down' in the polls nationwide that this could be happening in Scotland. And the national aggregates might be showing a better SNP share because of tactical voting in a 'punish the Tories' election.

SNP voter retention also seems better in more recent polling. Labour support may be getting backfilled by the ongoing Tory collapse not going to Reform.

[Filuwaúrdjan]

Maybe Brecon at a push but Labour seems to have given the LDs a free run there so the LDs should win it with tactical votes.

No, the local Labour Party are fighting it quite hard and I understand that some activists have been heading there from parts of the Valleys (where things are quiet). That one is going to be a complete mess.