United Kingdom General Election: July 4, 2024

[afleitch]

There's also some chatter that the 'nowcaster' and 'forecaster' polls that should be converging... aren't.

Some forecasting measures made assumptions about a Tory swingback that isn't happening with different voter movements now visible in polling.

With this in mind, it is worth noting that postal votes are going out now. Postal voting now makes up as much as 20% of the overall vote, so it will be interesting to see if this is incorporated into any of the latter models.

There's been an absolute glut of pollsters this year all measuring national voting intention and almost nothing else (at least publicly) It's quite archaic. Not much on constituency polling or more local polling (which didn't do too poorly in May) and MRPs seem like curios.

Someone doing a specific poll of postal voters and trying to see if there's any 'once in the polling station' effect could capture what's actually happening when people cast votes.

Labour's win looks big enough to be one that at the same time cannot be missed nor can it be measured. So no one can get it wrong, but they can broadly get it right.

[Alcibiades]

There's been an absolute glut of pollsters this year all measuring national voting intention and almost nothing else (at least publicly) It's quite archaic. Not much on constituency polling or more local polling (which didn't do too poorly in May) and MRPs seem like curios.

To be fair, constituency polling has a very poor record. The last time it was done on a widespread scale was 2015, when it failed even more miserably than the national polls (although admittedly that maybe had something to do with Lord Ashcroft doing most of them).

[Oryxslayer]

There's been an absolute glut of pollsters this year all measuring national voting intention and almost nothing else (at least publicly) It's quite archaic. Not much on constituency polling or more local polling (which didn't do too poorly in May) and MRPs seem like curios.

To be fair, constituency polling has a very poor record. The last time it was done on a widespread scale was 2015, when it failed even more miserably than the national polls (although admittedly that maybe had something to do with Lord Ashcroft doing most of them).

There was also a lot done by a pro-remain group in 2019 to raise awareness/judge viability of various Lib-Dem candidates. In almost every instance though they overshot the Lib-Dems, perhaps because of relative response rates having an oversized impact on a type of polling that is very vulnerable to those things.

[CumbrianLefty]

There's also some chatter that the 'nowcaster' and 'forecaster' polls that should be converging... aren't.

Some forecasting measures made assumptions about a Tory swingback that isn't happening with different voter movements now visible in polling.

shy don’t work with the shy tories voters aren’t voting for the tories shy reform shy lib dems voters perhaps?

Nobody is shy, is more likely.

Despite *still* being ceaselessly prayed in aid as something that will magically save them, actual "shy Tories" haven't been a significant factor in polling since the 1990s.

The most recent major poll error in their direction (2015) was actually for totally different reasons.

[Almost Anyone But Biden Or Trump (ABBoT but not Greg Abbott)]

https://www.twitter.com/focaldataHQ/status/1803059872390353022

NEW: Labour's lead climbs to campaign high.

Labour: 43% (+1)
Conservative: 21% (-3)
Reform UK: 16% (+1)
Liberal Democrats: 10% (+1)
Green: 5% (-)

Fieldwork conducted 14–17 June
2,604 respondents (GB)

Another poll with Conservatives declining and reform gaining, which has the effect of giving Labour a large lead.

At this point I think the biggest chance of Labour getting something other than a clear victory would be if the momentum for Reform continues and there is mass abandonment of the Conservatives, bringing them under 10% or so (with some of their voters going to reform, and some others to Lib Dems). If that happens, maybe Reform and the Lib Dems can win enough seats to prevent a huge Labour majority.

That's probably not the most likely scenario, but it seems more likely at this point than the Conservatives actually recovering.

[Almost Anyone But Biden Or Trump (ABBoT but not Greg Abbott)]

Count Binface has received his first Newspaper endorsement. No doubt the first of many, it seems he has got momentum on his side:

https://www.twitter.com/PolitlcsUK/status/1802824659559469159

🚨 NEW: The Daily Star has endorsed Count Binface

[CumbrianLefty]

Is that latest YouGov poll of Scotland using their new methodology?

[YL]

We have a new player in the MRP game, Ipsos:
https://www.ipsos.com/en-uk/uk-opinion-polls/ipsos-election-mrp

Lab 453, Con 115, LD 38, SNP 15, Plaid 4, Green 3, Reform 3

The seats where they show Reform ahead are Clacton, Ashfield (where their model seems to ignore Zadrozny) and NW Leicestershire. The seats where they have the Greens ahead are N Herefordshire, Bristol C and Waveney Valley (where Reform UK are shown as second), and not Brighton Pavilion.

[JimJamUK]

We have a new player in the MRP game, Ipsos:
https://www.ipsos.com/en-uk/uk-opinion-polls/ipsos-election-mrp

Lab 453, Con 115, LD 38, SNP 15, Plaid 4, Green 3, Reform 3

The seats where they show Reform ahead are Clacton, Ashfield (where their model seems to ignore Zadrozny) and NW Leicestershire.

61% apparently…

[Gass3268]

A constituency poll:

https://twitter.com/electionmapsuk/status/1803118581149761684?s=46

[JimJamUK]

LOL

https://twitter.com/electionmapsuk/status/1803118581149761684?s=61&t=I1DE8_c...

[Talleyrand]

Starting to look superficially like those massive Labour-SNP swings in 2015.

[Germany1994]

LOL

https://twitter.com/electionmapsuk/status/1803118581149761684?s=61&t=I1DE8_c...

If replicated in real life: How big of a swing would this be??

[oldtimer]

LOL

https://twitter.com/electionmapsuk/status/1803118581149761684?s=61&t=I1DE8_c...

If replicated in real life: How big of a swing would this be??

33% swing, about in line for the seat according to the national polls ,that give the Conservatives less than 20% nationally.

No wonder the Conservative frontline is West Devon.

[Joe Republic]

Well there goes Rehman Chishti's second attempt at the leadership.

[TheTide]

LOL

https://twitter.com/electionmapsuk/status/1803118581149761684?s=61&t=I1DE8_c...

If replicated in real life: How big of a swing would this be??

33% swing, about in line for the seat according to the national polls ,that give the Conservatives less than 20% nationally.

No wonder the Conservative frontline is West Devon.

A 39% decline nationally would put the Tories on less than 10%. What the poll suggests is something that many have long suspected - that the swing is likely to be fairly efficient (i.e. generally bigger in seats that the Tories hold).

[oldtimer]

https://www.twitter.com/focaldataHQ/status/1803059872390353022

NEW: Labour's lead climbs to campaign high.

Labour: 43% (+1)
Conservative: 21% (-3)
Reform UK: 16% (+1)
Liberal Democrats: 10% (+1)
Green: 5% (-)

Fieldwork conducted 14–17 June
2,604 respondents (GB)

Another poll with Conservatives declining and reform gaining, which has the effect of giving Labour a large lead.

At this point I think the biggest chance of Labour getting something other than a clear victory would be if the momentum for Reform continues and there is mass abandonment of the Conservatives, bringing them under 10% or so (with some of their voters going to reform, and some others to Lib Dems). If that happens, maybe Reform and the Lib Dems can win enough seats to prevent a huge Labour majority.

That's probably not the most likely scenario, but it seems more likely at this point than the Conservatives actually recovering.

The average change for the 5 pollsters with the same sample dates 14-16/17 June compared to their previous ones:

CON -1.4 (21.2)
LAB +0.2 (42.6)
LD 0 (10.4)
REF +2 (16.4)
GRN 0 (5)

The momentum is with Reform but not quite Corbyn 2017, if they keep up they should switch places in the polling average with the Conservatives by the end, something like this could be the final average :

LAB 42
REF 21
CON 16
LD 11
GRN 5

I guess that would result in seats something like this:

LAB 500
LD 75
REF 20
SNP 20
CON 10
PC 4
GRN 2

[oldtimer]

LOL

https://twitter.com/electionmapsuk/status/1803118581149761684?s=61&t=I1DE8_c...

If replicated in real life: How big of a swing would this be??

33% swing, about in line for the seat according to the national polls ,that give the Conservatives less than 20% nationally.

No wonder the Conservative frontline is West Devon.

A 39% decline nationally would put the Tories on less than 10%. What the poll suggests is something that many have long suspected - that the swing is likely to be fairly efficient (i.e. generally bigger in seats that the Tories hold).

When a party collapses don't look at Uniformed Swing, use proportional.

If the Conservatives are losing 2/3rds of their vote just divide their 61% in that seat by 3 = 20.3%

In the case of that poll it says they lost 5/8ths of their vote, replicated nationally would put the Conservatives at 17% , half of the pollsters already say 18-19% so not very far off.

[DL]

Have any polls explored 2nd choices of what's left of Tory voters? I wonder what proportion would be open to voting Reform or whether the hard core of Tory support when its down at 18% are people who would never vote for Farage under any circumstances - in which case Reform will soon hit a ceiling (if it hasn't already).

There has suddenly been all this interest among UK pundits in what happened in the 1993 Canadian election and all the obvious comparisons to another Reform Party supplanting the Tories in Canada - but those pundits all seem to ignore one key factor in Canada in 1993 that is missing this year in the UK. Canada's Reform Party was created due to feelings of alienation in the western provinces and they swept most of the seats in BC and Alberta that year. It was very much a regional protest movement. The Reform phenomenon in the UK has no regionalist element - I guess the equivalent would be if Reform UK had started out as a party protesting how the Midlands were being mistreated by Westminster and swept every single Tory seat in the midlands...

Another observation, I'm surprised at how according to surveys, so few 2019 Tory voters are switching to the Lib Dems. In the past if you were a "One Nation" Tory in the south and you were pissed off with the Tories, and could not bring yourself to vote Labour - you voted Lib Dem. Whatever happened to those people?   

[Philly D.]

The swing of 32,45 well exceeds the 22,7 BQ --> NDP swing in Quebec 2011 and is just a titch below the 35,65 point Lib --> PC in Quebec in 1984!

A swing of 26% in the South East would yield the Tories only 3 seats, two of which they would only hold for lack of a reform candidate (Weald of Kent being the only "genuine" hold).

Who had the South East Labour caucus being the largest regional one on their bingo card?

[TheTide]

Another observation, I'm surprised at how according to surveys, so few 2019 Tory voters are switching to the Lib Dems. In the past if you were a "One Nation" Tory in the south and you were pissed off with the Tories, and could not bring yourself to vote Labour - you voted Lib Dem. Whatever happened to those people?  

A lot of those were/are probably well to the right of 'One Nation' Toryism. The Lib Dems are good at picking up voters from all directions.

The swing of 32,45 well exceeds the 22,7 BQ --> NDP swing in Quebec 2011 and is just a titch below the 35,65 point Lib --> PC in Quebec in 1984! Who had the South East Labour caucus being the largest regional one on their bingo card?

Probably not many, as (AFAIK) there aren't regional caucuses in Labour or any other party.

[Logical]

This is uhhhh..... someone who has given up?
https://twitter.com/Mike_Fabricant/status/1802968304736194856

[IceAgeComing]

A lot of those people would have voted for someone other than the Tories in 2019; to be frank. The 2019 voting patterns were unique because Brexit was the defining issue; and this election is seeing a reversion to 'normal' in a lot of ways; with an incredibly unpopular incumbent government getting thrashed.

If replicated in real life: How big of a swing would this be??

Others have given the answer but I can give the method if helpful - to calculate the swing from party B to party A the formula is simply (A-B)/2. In this scenario; where A is Labour and B are the Conservatives; it would be (27+39)/2 (keep in mind that a negative negative is positive); which would be 66/2=33% swing from the Conservatives to Labour. You minus party B because if their vote share is positive then you have to subtract that from the gain of party A (see the many examples in 2017 where both Labour and the Tories increased their vote share); while if party A falls backwards you subtract that from the initial number. This is probably a slightly overly complicated way of explaining it but it's a good firm rule - but if one party is positive and the other is negative you add them together and divide by 2.

My view on that constituency poll is that it feels like nonsense but also the sort of nonsense that isn't good for the Tories - even a rogue poll that puts you 30% behind in Gillingham and Rainham is hardly good for the Conservatives. There's the by-election level swing (which we've never seen in a general election really - certainly not outside a handful of seats); there's Reform at 15% which feels low for a seat like this (although that might be an indication for them having a uniform vote share which is not great if you are Reform; you don't win seats with FPTP with a uniform share as the Alliance learned in 1983) and it just doesn't smell right. However I can't entirely rule it out - the area around the Medway Towns could be very interesting in this election (although where isn't in this election - probably Liverpool which will be 90% Labour again).

[oldtimer]

Have any polls explored 2nd choices of what's left of Tory voters? I wonder what proportion would be open to voting Reform or whether the hard core of Tory support when its down at 18% are people who would never vote for Farage under any circumstances - in which case Reform will soon hit a ceiling (if it hasn't already).

There has suddenly been all this interest among UK pundits in what happened in the 1993 Canadian election and all the obvious comparisons to another Reform Party supplanting the Tories in Canada - but those pundits all seem to ignore one key factor in Canada in 1993 that is missing this year in the UK. Canada's Reform Party was created due to feelings of alienation in the western provinces and they swept most of the seats in BC and Alberta that year. It was very much a regional protest movement. The Reform phenomenon in the UK has no regionalist element - I guess the equivalent would be if Reform UK had started out as a party protesting how the Midlands were being mistreated by Westminster and swept every single Tory seat in the midlands...

Another observation, I'm surprised at how according to surveys, so few 2019 Tory voters are switching to the Lib Dems. In the past if you were a "One Nation" Tory in the south and you were pissed off with the Tories, and could not bring yourself to vote Labour - you voted Lib Dem. Whatever happened to those people?   

That was basically tested in the end of May (which was also weirdly the end of May) 2019.

The Tories got 8.8% in the Euro Elections , most opinion polls overstated them but momentum clearly was downwards towards that level.

The General Election opinion polling unfortunately doesn't give a clearer view, because Theresa May resigned before they started polling, 17-20% as Boris turned it around, so it could have been lower than 17%.

[DL]

What sort of a place is Gillingham and Rainham? Is it a remain voting "blue wall" sort of place?