United Kingdom General Election: July 4, 2024

[Torrain]

I know, I know. But balance, or something.
https://twitter.com/JLPartnersPolls/status/1795500304571490715

[oldtimer]

I know, I know. But balance, or something.
https://twitter.com/JLPartnersPolls/status/1795500304571490715

A usefull guide:
https://twitter.com/patrickjfl/status/1795416143521255429?

[Trans Rights Are Human Rights]

The mental image of Rishi on the Big Balls is one I desperately needed. Thank you, Jake ToryWipeout Twitter.

(Sorry for the meaningless spam to all the international posters; continue on)

[Torrain]

Per the Times - Abbott won't be allowed to stand as a Labour candidate.

[JimJamUK]

Is there anything interesting here? The relatively even distribution of support for all non-Lib-Dem parties makes me wonder if these have just become larger versions of the same skeptical crosstabs as regular polls. Same with the large Lib-Dem number in the east midlands (maybe the excel file had east and east midlands flipped?) but lacking any targets whatsoever.

I struggle to think of a constituency where the Lib Dem’s would get 13% in the East Midlands (okay, I’ve thought of 2 by the time I’ve typed out this post), nevermind average it across the region. FWIW, Redfield and Wilton tend to have quite flat distribution of votes and have for a while, to the extent that I’m a bit sceptical of their polling.

[JimJamUK]

Also LOL at London being the tied best region for the Conservatives, followed by Wales and then the North East.

[afleitch]

'Trend is your friend.'

Even if not a single person in the country changed their mind since the election announcement you're still going to get these sorts of variances up or down 2 points. JL is also a few days old; fieldwork ended Saturday. And on balance, the newer polls were polling over the Bank Holiday weekend which had there not been an election, would not have happened (most took a pause in the Bank Holiday earlier this month)

What's worth noting is the variance in party share; a 19% to 28% share for the Tories is effectively one that's worth a third of their potential vote share. Labour's variance of 40% to 47% is closer to 1 in 10.

Reform seem to be quite sticky; given the official starting block for the 2019 election was less traditional, Week 2 polls in 2019 started to see Brexit (starting at a similar position then as now) start to ebb. That's something to look out for in the next few days.

[afleitch]

A few more thoughts, with 'pushing undecideds', that's going to start happening either through methodology or naturally as the campaign progresses. That doesn't necessarily mean pollsters converging around those who 'push' either because those pollsters will also find that people need less prompting.

Panel polls in particular attract the politically engaged; it's why they tend to be bouncy and (potentially) protest heavy. For example, 67% voted last time, so polls finding 80% to 90% as 'likely voters' are already screwed towards an existing sample. If that sample isn't 'refreshed' it can be less polling and more 'focus group'. This isn't a bad thing of course, unless there's a turnout increase at which point you're not sampling enough 'did not vote.'

[DL]

When polling shows a huge landslide for any party - the final results almost always end up being a bit closer. I think that's because when (in this case) Tory support is that low you have a lot of traditional lifelong Tories who are parked in DK or with Reform and when push comes to shove some of them will go back to their traditional home when their "political DNA" reasserts itself. This won't change the outcome much but even in 1997 you may recall that the polls had the Tories doing even worse than their final vote share ended up being.  

I was actually looking back at the polling in 1997 - in the end Labour won the national popular vote by only 12 points but that was enough to get them 418 seats and a huge majority. The Labour vote did erode in the polls over the course of the campaign - but it was mostly lost to the Lib Dems not to the Tories. I suspect that many Labour voters in "blue wall" type seats voted tactically for the Lib Dems in the end

[MABA 2020]

https://twitter.com/JAHeale/status/1795455911139430537

Excellent news

[Philly D.]

When polling shows a huge landslide for any party - the final results almost always end up being a bit closer. I think that's because when (in this case) Tory support is that low you have a lot of traditional lifelong Tories who are parked in DK or with Reform and when push comes to shove some of them will go back to their traditional home when their "political DNA" reasserts itself. This won't change the outcome but even in 1997 you may recall that the polls had the Tories doing even worse than their final vote share ended up being.  

This doesn't always happen (see Canada in 1993 and 2011, and BC in 2001!!!) -- the assumption that it will seems hard baked in the United Kingdom of Because Tradition and Supermarket Tabloids.

Looking good Mr. Prime Minister

https://twitter.com/RedfieldWilton/status/1795485183073775620

London (1199 weighted):

45% Labour
26% Conservative
11% Lib-Dem
10% RefUK
7% Green
2% Other

East Midlands (526 weighted):

44% Labour
24% Conservative
14% RefUK
13% Lib-Dem
4% Green

East (826 weighted):

46% Labour
22% Conservative
17% RefUK
9% Lib-Dem
6% Green

It makes sense that the swing in London would be a bit lower than elsewhere -- it's been swinging towards Labour vis-a-vis the Tories for the past 3 elections, it has relatively more areas where the Greens would be doing relatively well and it has more visible minorities who are already more Labour. (Although I don't think the Tories will limit their losses to 5 seats).

But those East of England numbers... on a UNS, Maldon, Castle Point and just barely Clacton for the Tories. Extinction-level numbers! The East Midlands aren's as awful (because Lincolnshire), but again, only 3 seats for the Tories after that (Bosworth, Rutland and Melton and Daventry).

[DL]

This doesn't always happen (see Canada in 1993 and 2011, and BC in 2001!!!) -- the assumption that it will seems hard baked in the United Kingdom of Because Tradition and Supermarket Tabloids.

Funny you should mention BC in 2001. As I recall the NDP was polling in the mid teens during the campaign and the BC Liberals were over 60% - but when the votes were counted the BC Liberals beat the NDP 58% to 22% - a crushing defeat to be sure, but 22% is higher than the 17% the polls were predicting.

I don't want to overstate my theory. I suspect the polls will be largely accurate but if the "poll of polls" has the Tories getting 23% - look for them to get a "dead cat bounce" to 25%

[Philly D.]

You are right -- my bad, I thought for some reason the NDP did even worse than they did. Quebec 2022 would be a better example.

[oldtimer]

'Trend is your friend.'

Even if not a single person in the country changed their mind since the election announcement you're still going to get these sorts of variances up or down 2 points. JL is also a few days old; fieldwork ended Saturday. And on balance, the newer polls were polling over the Bank Holiday weekend which had there not been an election, would not have happened (most took a pause in the Bank Holiday earlier this month)

What's worth noting is the variance in party share; a 19% to 28% share for the Tories is effectively one that's worth a third of their potential vote share. Labour's variance of 40% to 47% is closer to 1 in 10.

Reform seem to be quite sticky; given the official starting block for the 2019 election was less traditional, Week 2 polls in 2019 started to see Brexit (starting at a similar position then as now) start to ebb. That's something to look out for in the next few days.

There are 3 reasons why I think Reform is sticky:

1. Since 2005, parties to the right of the Conservatives have gotten a larger share than Goldsmith's Referendum party in 1997, for many voters it won't be the first time to vote that way, UKIP was also sticky in 2015.

2. Reform rose by itself in the opinion polls, it has practically no leader and no organisation, it's a shell. Yet that empty shell still rose without any effort, meaning there's genuine demand.

3. The main issue about Brexit was Brexit, Boris ate that issue and it's votes in 2019. Reform now polling that high without Brexit means it's more complex than just 1 thing.

[Filuwaúrdjan]

An issue to be aware of with historical GB polling figures is that the polling industry only polls GB, whereas the usual summary figures for percentages at elections are for the entire United Kingdom. This is not always accounted for by the people who have put together poll lists for the various Wikipedia articles, as useful as they are. Unfortunately that kind of sloppiness in an otherwise useful resource is quite typical.

So for the 1997 election, the correct figures to compare polling numbers to are: Labour 44.3, Con 31.4, LDem 17.2, not 43.2, 30.7 and 16.8.

[DL]

Another reason the Tories may do a bit better than the polls say is that they still will have hundreds of MPs running for reelection each with some semblance of a local organization to identify votes. Reform UK is non-existent on the ground

[Conservatopia]

Conservative Party emails begging for money are HILARIOUS. Pathetic and sad but hilarious.

[beesley]

For Rishi Sunak, his visit to Caithness etc. a few days ago, and his visits to Stoke North etc. and Hinckley and Bosworth today.

For Keir Starmer, his visit to Glasgow sometime ago, and his visits to Derby South (for what that's worth, but as I said, no specific criteria and no assumptions) and Stevenage today.

For Ed Davey, his campaign launch in Cheltenham and the very beginning, and his visit to the bottom of Lake Windermere today.

[Filuwaúrdjan]

Rolls Royce employ over ten thousand people in Derby and the workforce is spread out across the region, as is usually the way these days.

[Filuwaúrdjan]

It would appear that some sort of deal has been made at Hackney North & Stoke Newington.

[beesley]

Rolls Royce employ over ten thousand people in Derby and the workforce is spread out across the region, as is usually the way these days.

Yes - however I couldn't think of a more objective way of doing it than to put the specific seat itself.

[Torrain]

Wee bit of candidate news:

LBC’s Iain Dale has quit radio to run for the Tories in Tunbridge Wells. 15k majority in 2019, and held by the party since its 1974 creation.

And in Birmingham Northfield (one of Labour’s easiest target seats), the Labour candidate Alex Aiken is standing down over “personal challenges which have arisen recently”.

[Torrain]

https://twitter.com/BallotBoxScot/status/1795568833069244431

[Chickpeas]

No Conservative MPs have announced they're standing down since the Michael Gove and Andrea Leadsom double whammy on Friday evening. Surely that's not the last of them?

[Blair]

Wee bit of candidate news:

LBC’s Iain Dale has quit radio to run for the Tories in Tunbridge Wells. 15k majority in 2019, and held by the party since its 1974 creation.

And in Birmingham Northfield (one of Labour’s easiest target seats), the Labour candidate Alex Aiken is standing down over “personal challenges which have arisen recently”.

Dale is fascinating; was chief of staff to David Davis, and a firm Brexiteer but over the recent years has been more of a Tory Wet- for those who listen to his podcast it’s quite funny as he has been a strong critic of the government over immigration (he’s very pro) and various other issues.

The nature of his radio show means he can claim to have a relatively wide ranging interest in policy that isn’t exactly typical for a Tory; he did a lot on British airways poor treatment of staff during covid.

He would be an interesting MP; but I am surprised as it’s signing up for 5 years in opposition.