The Bishop
Labour
Down With Factionalism!
Posts: 38,205
|
Post by The Bishop on Jun 21, 2023 11:16:14 GMT
Their latest "Red Wall" poll also shows the biggest Labour lead for a while.
|
|
|
Post by andrewp on Jun 26, 2023 17:56:34 GMT
Westminster VI (25 June):
Labour 44% (-2) Conservative 26% (–) Liberal Democrat 13% (+1) Reform UK 6% (-1) Green 5% (-1) Scottish National Party 3% (–) Other 3% (+2)
Changes +/- 18 June
|
|
|
Post by stodge on Jun 27, 2023 19:08:25 GMT
The Red Wall polling tonight shows an increase in the Labour lead which now represents an 18% swing since December 2019.
|
|
|
Post by andrewp on Jun 27, 2023 19:09:24 GMT
The Red Wall polling tonight shows an increase in the Labour lead which now represents an 18% swing since December 2019. Red Wall VI (25 June): Labour 53% (+3) Conservative 26% (-2) Reform UK 9% (+1) Liberal Democrat 6% (-1) Green 4% (–) Other 2% (-1) Changes +/- 11 June
|
|
|
Post by andrewp on Jul 3, 2023 20:54:26 GMT
Westminster VI (2 July):
Labour 46% (+2) Conservative 28% (+2) Liberal Democrat 11% (-2) Reform UK 5% (-1) Green 5% (–) Scottish National Party 3% (–) Other 2% (-1)
Changes +/- 25 June
|
|
|
Post by stodge on Jul 5, 2023 20:04:19 GMT
Dreadful Blue Wall polling for the Conservatives.
Just to explain what R&W call "the Blue Wall":
Various criteria have been used by different media, academic, and other sources to decide which constituencies constitute the ‘Blue Wall.’ For the purposes of our tracker polling, we have limited ourselves to studying constituencies which meet five criteria: 1) The constituency is in the South of England 2) The constituency elected a Conservative MP at the 2015, 2017, and 2019 General Elections 3) At least 25% of adults in the constituency have a degree 4) The Remain vote in the 2016 Brexit referendum in the constituency was greater than 42.5% 5) The Conservatives hold the constituency on a majority of less than 10,000 over Labour OR less than 15,000 over the Liberal Democrats.
The current polling has Labour on 36% (+15.5 on when these parties were contested in December 2019), Conservatives 29% (-21), Liberal Democrats 25% (-2.5). That's a swing from Conservative to Labour of in excess of 18% and a 9% from Conservative to Liberal Democrat.
Add in the likelihood of tactical voting and you look at a seat like Thornbury & Yate and the LDs would win that (Con maj 12,369) on the above numbers.
|
|
|
Post by robert1 on Jul 10, 2023 16:10:03 GMT
Con 27 (-1) Lab 48 (+2) LD 11 = Ref 5 = Grn 4 (-1) SNP 3 = Oths 1 (-1)
Fieldwork 02.07
|
|
|
Post by stodge on Jul 10, 2023 20:32:47 GMT
The England sub sample is a 16% swing from Conservative to Labour.
That in itself disguises key divergences in swing by region - the south and west are moving more strongly to Labour than the Midlands with a swing of over 20% while the swing away from the Conservatives in the Midlands is just over 10%.
|
|
Sibboleth
Labour
'Sit on my finger, sing in my ear, O littleblood.'
Posts: 15,745
|
Post by Sibboleth on Jul 10, 2023 21:03:32 GMT
The England sub sample is a 16% swing from Conservative to Labour. That in itself disguises key divergences in swing by region - the south and west are moving more strongly to Labour than the Midlands with a swing of over 20% while the swing away from the Conservatives in the Midlands is just over 10%. What do we learn from this? Not to bother with regional subsamples in polls from 'Redfield & Wilton'.
|
|
|
Post by batman on Jul 10, 2023 21:09:38 GMT
If I had a fiver for every time someone has set too much store on sub-samples in polls I could get seriously drunk. We should all know by now that subsamples may be quite fun to look at but have next to no validity in terms of accuracy.
|
|
|
Post by stodge on Jul 11, 2023 18:19:36 GMT
If I had a fiver for every time someone has set too much store on sub-samples in polls I could get seriously drunk. We should all know by now that subsamples may be quite fun to look at but have next to no validity in terms of accuracy. No, we don't. That's just your opinion. This is the second forum where I've been castigated by the psephological "purists" so I'll offer my take. Of course, sub samples of 100 aren't worth it but within polls sampling 2,000, some of the sub samples are much larger such as England and I quote the England numbers from as many polls as they are either published or are easily calculated since we know England and its 500+ seats are where elections are normally won or lost. The issue with sub samples is weighting and you can see from the numbers some of the pollsters (R&W) are struggling whereas I find YouGov closer to what I think the truth is (and I've been told their weighting is the most accurate - the curiousity is YouGov are the pollsters regularly showing the lowest Conservative shares while others (Opinium) are trying to model "swingback" or whatever. The regional English sub samples (indeed, I'd argue the same for the Scottish and Welsh) aren't wholly reliable but across pollsters, across many weeks of polling, the inescapable conclusion is the Conservatives are performing worse in the south and west and slightly less badly in the midlands and north. UNS is the bigger deception - IF we are looking at a 1997-style outcome, the variations in swing may be considerable (back then, the north and Scotland saw 7% swings from Conservative to Labour while in the south and London it was more like 13%). The regional variations in swing will be crucial. We may see some evidence of this in the forthcoming by elections.
|
|
jamie
Top Poster
Posts: 7,019
|
Post by jamie on Jul 11, 2023 18:41:55 GMT
Just to explain what R&W call "the Blue Wall": Various criteria have been used by different media, academic, and other sources to decide which constituencies constitute the ‘Blue Wall.’ For the purposes of our tracker polling, we have limited ourselves to studying constituencies which meet five criteria: 1) The constituency is in the South of England 2) The constituency elected a Conservative MP at the 2015, 2017, and 2019 General Elections 3) At least 25% of adults in the constituency have a degree 4) The Remain vote in the 2016 Brexit referendum in the constituency was greater than 42.5% 5) The Conservatives hold the constituency on a majority of less than 10,000 over Labour OR less than 15,000 over the Liberal Democrats. That’s seems… overly broad.
|
|
The Bishop
Labour
Down With Factionalism!
Posts: 38,205
|
Post by The Bishop on Jul 15, 2023 9:30:45 GMT
Yes, count me too as a subsample sceptic.
Maybe somebody should look back to the pollsters who were closest to getting the last GE right, and then look at how accurate their subsamples were.
My guess is, not massively so - but I am happy to be corrected.
|
|
msc
Non-Aligned
Posts: 908
|
Post by msc on Jul 15, 2023 14:34:19 GMT
Yes, count me too as a subsample sceptic. Maybe somebody should look back to the pollsters who were closest to getting the last GE right, and then look at how accurate their subsamples were. My guess is, not massively so - but I am happy to be corrected. Taking two pollsters who had fairly accurate campaigns, the results are mixed. (A few of the other on the ball polls are now longer available online.) (All figures rounded to the nearest...) Ipsos-Mori last election poll of 2019: 44-33-12 sub-samples of regions given North England – 37 Tory 43 Lab 7 Lib Midlands 49 Tory 35 Lab 8 Lib South England 54 Tory 19 Lab 21 Lib London 32 Tory 49 Lab 15 Lib England as a whole 45 Tory 35 Lab 12 Lab Scotland 22 Tory 19 Lab 7 Lib 48 SNP Survation last election poll of 2019: 45-34-9 subsamples London – Tory 25 Lab 44 Lib 6 Midlands Tory 42 Lab 25 Lib 9 North Tory 30 Lab 37 Lib 4 (with 5% Brexit Party) South Tory 44 Lab 19 Lib 11 (with 5% Brexit Party) England as a whole Tory 37 Lab 28 Lib 8 Scotland Tory 21 Lab 12 Lib 8 SNP 36 Wales Tory 26 Lab 28 Lib 1 Plaid 13 Actual regional results (sources - Electoral Calculus, Principal Fish, Wiki - they argue on the odd point here or there but let's just talk ballpark figures) London – Tory 32 Lab 48 Lib 14 Scotland – Tory 25 Lab 19 Lib 10 SNP 45 Wales – Tory 36 Lab 41 Lib 6 Plaid 10 Brexit 5 E Midlands – Tory 55 Lab 32 Lib 8 W Midlands – Tory 53 Lab 34 Lib 8 N E England – Tory 38 Lab 43 Lib 7 Brexit 8 N W England – Tory 37 Lab 46 Lib 8 Yorkshire Tory 43 Lab 39 Lib 8 S E England – Tory 54 Lab 22 Lib 18 S W England – Tory 53 Lab 23 Lib 18 England as a whole Tory 47 Lab 34 Lib 12 Survation, who I tend to think highly of, have all the wrong numbers in the subsamples, but get the overall direction of traffic correct. Ipsos are almost spot on in a number of regions. However, I'd agree that, at best, this can only give you warning whispers of danger on election night (ie the tightening of the Northern polling there, and Wales) and not a master plan. And we're talking days before the election, not a year out. Hope that helps!
|
|
|
Post by robert1 on Jul 17, 2023 16:41:32 GMT
Con 27 = Lab 44 (-4) LD 13 (+2) Ref 8 (+3) Grn 4 = SNP 3 = Oths 1 =
Fieldwork 16.7.23 (I believe)
Not sure that Reform figure looks credible
|
|
cibwr
Plaid Cymru
Posts: 3,587
|
Post by cibwr on Jul 20, 2023 18:23:34 GMT
|
|
|
Post by andrewp on Jul 24, 2023 17:06:46 GMT
Westminster voting intention:
LAB: 45% (+1) CON: 28% (+1) LDEM: 14% (+1) REF: 6% (-2) GRN: 4% (-)
via @redfieldwilton, 23 Jul
|
|
|
Post by andrewp on Jul 26, 2023 17:13:03 GMT
Some wall news
Labour leads by 18% in the Red Wall.
Red Wall VI (23 July):
Labour 48% (-4) Conservative 30% (+3) Reform UK 10% (+1) Liberal Democrat 6% (–) Green 4% (–) Plaid Cymru 2% (+1) Other 1% (-1)
Changes +/- 9 July
|
|
|
Post by Clark on Jul 27, 2023 9:36:34 GMT
The Tories polling higher in 'Red Wall' areas than the country as a whole is interesting.
Labour gains in places like Bishop Auckland for example might not be so straight forward as they seem
|
|
|
Post by borisminor on Jul 27, 2023 9:44:35 GMT
The Tories polling higher in 'Red Wall' areas than the country as a whole is interesting. Labour gains in places like Bishop Auckland for example might not be so straight forward as they seem I think the 'Red Wall' will be more stubborn to return to Labour and will have a slightly lower swing - but in the grand scheme of things I don't think it will make all that much difference looking at the national polling.
|
|