|
|
Post by James E on May 5, 2024 14:01:18 GMT
According to electoral calculus, who don't have an axe to grind the following are the " strong" seats for the Lib dems , where the Lib dems are regarded as significantly ahead of any of the other centre left parties according to their MRP. Bath, Bermondsey and Old Southwark, Bicester and Woodstock, Brecon Radnor and Cwm Tawe, Caithness Sutherland and Easter Ross, Cambridge, Cambridgeshire South, Carshalton and Wallington, Cheadle, Cheltenham, Chesham and Amersham, Chippenham, Devon North, Devon South, Didcot and Wantage, Dorking and Horley, Dorset West, Eastbourne, Eastleigh, Edinburgh West, Ely and East Cambridgeshire, Epsom and Ewell, Esher and Walton, Farnham and Bordon, Fife North East, Finchley and Golders Green, Frome and East Somerset, Glastonbury and Somerton, Godalming and Ash, Guildford, Hampstead and Highgate, Harpenden and Berkhamsted, Harrogate and Knaresborough, Hazel Grove, Henley and Thame, Honiton and Sidmouth, Kingston and Surbiton, Lewes, Mid Dunbartonshire, Newbury, Norfolk North, Orkney and Shetland, Oxford West and Abingdon, Richmond Park, Romsey and Southampton North, Sheffield Hallam, Shropshire North, St Albans, St Ives, St Neots and Mid Cambridgeshire, Sussex Mid, Sutton and Cheam, Taunton and Wellington, Thornbury and Yate, Tunbridge Wells, Twickenham, Wells and Mendip Hills, Westmorland and Lonsdale, Wimbledon, Winchester, Woking, Wokingham, Yeovil About 70 in total , including seats already held. There are an additional 30 odd seats with a "medium" possibility. The poll was conducted in February the polling for the Lib dems hasn't changed significantly since then. While of course there's no guarantee of picking up all 55 but they are all winnable, to only pick up 10 is highly unlikely. Their own middle ground estimate for lib dem mps is 50 with a minimum of 30 and a max of around 60 , which seems plausible, there's an outside chance of a few more and official opposition status of the election campaign goes exceptionally badly for the Tories. www.electoralcalculus.co.uk/blogs/ec_lib2seats_20240226.htmlElectoral Calculus's 'Top 60' seats are based on the " Lib Dems' closeness to winning", not on being "significantly ahead of other centre-left parties". This explains the inclusion of Cambridge, Bermondsey, and other Labour-held seats where the LDs were second. The list includes seats currently held by the LDs. EC's point is that while the LibDems topped the vote in these (after combining all 60) according to the FoN/EC MRP in Jan/Feb (as helpfully linked by oldnat yesterday) they are second to Labour in the 'medium' seats, which are the next-best 30 for them. |
|
|
|
Post by James E on May 4, 2024 17:23:10 GMT
FoN/Electoral Calculus MRP poll reported in the Mirror
www.mirror.co.uk/news/politics/bombshell-mega-poll-predicts-tories-32121361
Party VI Seats Diff (Scot seats)
Lab 42% 452 +249 (13)
Con 22% 80 -285 ( 0)
LD 11% 53 + 42 ( 4)
SNP 4% 40 - 8 (40)
REFUK 10% 0 0 ( 0)
Green 7% 2 + 1 ( 0)
As with all MRP polls I am very dubious about the reliability of allocating a "winner" in a marginal seat! This poll is 3 months old. The report is dated 14 Feb and the fieldwork was 24Jan-12 Feb. The polls have shifted a bit since then, so the Conservatives are probably not faring as well as that now. |
|
|
|
Post by James E on May 4, 2024 14:53:37 GMT
from the BBC
"Khan on course for comfortable win
Professor Sir John Curtice
With 11 of the 14 mayoral constituencies in London declared, we are now estimating that Sadiq Khan will win 44% of the vote and Susan Hall 33%."
If that is the outcome, it looks like there has not been a great amount of squeeze on the LD, Green and other parties votes. overall Lab + Lab share is up by 2 points, with Lab up 4 and Con down by 2, compared to the first round result in 2021.
|
|
|
|
Post by James E on May 4, 2024 13:38:35 GMT
Khan wins SW, which includes Richmond & Twickenham. That sounds unexpected? He was only just behind there in 2021 by 77,000 to 74,000 votes - and won by 5,000 votes after second preferences were added. The result here is a 3% swing from Con to Lab, so on the low side. The Bromley and Bexley result is the first we have had with a swing from Lab to Con - but its very small (less than 1%). However, the overall average swing must still be around 3-4% Con to Lab. John Curtice has made a forecast based on the first 6 results of Khan winning by a 10% margin (43 to 33). That would be a slightly lower overall swing of 2.5%. [TO ADD 15:40 - a Con to Lab swing of 1.2% in Brent & Harrow, so another case of a smaller swing in Outer London] |
|
|
|
Post by James E on May 4, 2024 12:50:50 GMT
Khan wins West Central, where he failed to win in 2021.
In 2024:
Sadiq Khan: 54,481
Susan Hall: 43,405
In 2021 (first preferences):
Sadiq Khan: 51,508
Shaun Bailey 53,713
All three areas to declare so far show around a 5% swing to Labour. If that was replicated all over London, he would win by around 15%.
.....TO ADD
North East (Hackney, Islington, Waltham Forest):
Sadiq Khan – 127,455
Susan Hall – 34,099
In 2021, Khan won the first preference vote here by 111K to 44K. Swing to Labour of about 7% here.
...TO ADD (2)
I make it around a 3% swing to Labour in the South-West Constituency, which is Outer London. The average Con to Lab swing is still 5% (and not varying much by area) from these early results.
...TO ADD (3)
Enfield & Haringey
Another comfortable win for Khan, and another 4-5% swing to him.
Very consistent results so far, and not a huge difference between the pro-Labour swings in Inner and Outer London (though they do seem smaller in the latter).
|
|
|
|
Post by James E on May 3, 2024 19:52:29 GMT
Surprised that Labour has won the Norfolk PCC - much bigger swing than Cambridgeshire. That must be because Labour delivered leaflets for the PCC in Norfolk. I didn't get one in Cambridgeshire. |
|
|
|
Post by James E on May 3, 2024 17:10:31 GMT
Sky projection based on these results Labour largest party around 20 short of an overall majority Lib dems on around 40: And Tories on around 230. My immediate thought was what a load of old codswallop. I would of course be delighted for Labour to have to have my parties support but I suspect the chances of that are about the same as Rob Edwards calling me to turn out to save the team from relegation. As I said yesterday: "PNV from the May 2023 Local Elections was: Lab 35% Con 26% LD 20% The norm is for the LDs to outperform their polling by 8-10 points, with a similar underperformance by Labour (when they doing well). The Tories might be expected to get around their current Westminster polling average of low-to-mid 20s. It should help them that there are so few RefUK candidates. So I'd expect similar to 2023 for Lab and LDs this time, and the Tories just a bit lower. [11:30am - If Labour achieve less than a 12% lead, we can expect the usual suspects in the Tories' client media claiming that this would not be enough for an overall majority, per UNS. I imagine we'll be hearing this by about this time tomorrow.] Last Edit: yesterday at 11:29am by James E |
|
|
|
Post by James E on May 3, 2024 9:41:00 GMT
@lululemonmustdobetter
"Possibly the most reliable gauge could be the estimated national % share. Will a Labour lead in that reflect what we are seeing in the polls?"
No, it won't.
Local Elections produce a different pattern of results to General Elections, even when the two are held on the same day.
Labour take a lower share of the overall vote than in GE polls, particularly when they are doing well. As I mentioned yesterday, the best comparison for NEV would be last year's figures of Lab 35%, Con 26%, LD 20%. Looking further back, in 1995 and 1996 when Labour were polling in the mid-to-high 50s, they Local Election NEV was still around 8-10 points lower than that.
|
|
|
|
Post by James E on May 3, 2024 9:33:22 GMT
Full result for Blackpool South By-Election 🌹 LAB: 58.9% (+20.6) 🌳 CON: 17.5% (-32.1) ➡️ RFM: 16.9% (+10.7) 🔶 LDM: 2.1% (-1.0) 🌍 GRN: 2.0% (+0.3) 🙋 IND: 0.9% (New) 👨👩👧👦 ADF: 0.8% (New) 🎩 MRLP: 0.7% (New) ....... I think Labour will be happy That is a monumental collapse of the Conservative vote. 17.5% is just 35% of their 2019 share, and is even lower than the 38-39% they recorded in Wellingborough and Rochdale. And the increase in the Labour vote will be very pleasing for Starmer. In rough terms, the Conservatives' lost vote share went two-thirds to Labour and one-third to Reform UK. That isn't what the polls are showing, as most have this the other way round. And even Survation, who I think are likely to be the most accurate for RUK, show Labour taking just a bit more of the Con 2019 vote than RUK (16% to 14%). Reform UK's vote share is marginally lower than the 17.3% which UKIP achieved in 2015, when UKIP took 12.5% of the GB vote. Given that the pre-2015 by-elections produced far better results for UKIP than the subsequent General Election, it should be clear that they are unlikely to be in contention for any seats, even if they retain 10-13% of the vote. |
|
|
|
Post by James E on May 2, 2024 14:48:48 GMT
|
|
|
|
Post by James E on May 2, 2024 14:11:53 GMT
leftieliberal "... Mayor: Khan 47%, Hall 25%, Garbett 7%, Blackie 6%, Cox 6%, Count Binface 3%
Regrettably it looks like Count Binface will not come ahead of the Reform UK candidate." Binface does very well with the 18-24 cohort (of 110), coming in joint 3rd place with the Greens and ahead of both of the Tory and RefUK candidates. Khan (Lab) 50% Blackie (LD) 14% Count Binface 11% Garbett (Green) 11% Hall (Con) 8% Cox (Ref) 7% Others 0% And from these figures, he looks very likely to finish ahead seven of the thirteen candidates.
|
|
|
|
Post by James E on May 2, 2024 8:29:19 GMT
The projected national vote share derived from today's local council elections results will be interesting. PNV from the May 2023 Local Elections was: Lab 35% Con 26% LD 20% The norm is for the LDs to outperform their polling by 8-10 points, with a similar underperformance by Labour (when they doing well). The Tories might be expected to get around their current Westminster polling average of low-to-mid 20s. It should help them that there are so few RefUK candidates. So I'd expect similar to 2023 for Lab and LDs this time, and the Tories just a bit lower. [11:30am - If Labour achieve less than a 12% lead, we can expect the usual suspects in the Tories' client media claiming that this would not be enough for an overall majority, per UNS. I imagine we'll be hearing this by about this time tomorrow.] |
|
|
|
Post by James E on May 1, 2024 10:08:52 GMT
I have always voted Labour in Norwich at local elections and had every intention of doing so this week. I find the failure to deliver a single leaflet thoroughly discourteous and for the first time ever will be spoiling my ballot paper. Didn't you previously tell us that you spoiled your ballot paper before over the issue of All Women Shortlists? You have also said that you voted for other parties during most of the time when Blair was Prime Minister, and at other elections, citing various issues which you have had with whoever was Labour leader at the time. For instance, you said last year that you would now vote for the Greens in protest against Jeremy Corbyn's expulsion from Labour. |
|
|
|
Post by James E on May 1, 2024 8:53:52 GMT
Labour leads by pollster 26 People Polling 25 Ipsos 25 YouGov 23 Redfield & Wilton 22 WeThink 21 Techne 20 Deltapoll 19 More in Common 18 Survation 18 Savanta 16 Opinium I won't mourn Opinium's impending relegation. It's noticeable from Neil's figures how those at the foot of the table have pulled up in recent months. Earlier in the year, we had an 11 from More In Common, a 12 from Savanta and even a 14 from the perennially inconsistent Deltapoll. Such results seem to be more rare now. But there is also BMG to consider - their most recent poll was a 16, which is normal for them. Looking at the lower half of the table for the first four months of the year, it's actually MIC who are right at the bottom, even though they seem to have hit a bit of form recently, as have Savanta. In the bottom 6, we have two who heavily 'adjust' (Opin & MiC), two who don't prompt for RefUK or Greens (Sav & Surv), and two who don't let us know much about their methods. Overall averages for the lower end of the table look like this: Deltapoll (12 polls in 2024) Av 18.0 Survation (6) Av 17.7 Savanta (16) Av 17.3 BMG (4) Av 16.3 Opinium (10) Av 15.7 More In Common (8) Av 15.3 One other observation on current polling: we have had figures from nine different pollsters since 22 April - BMG, YG, Survation, Techne, WeThink, Savanta, MiC, R&W and Delta. Every single one of them reports their Conservative VI at a figure below their previous 2024 average, some by a little, and some by several points. Overall the Tories are around a further 1.3% lower in these most recent polls than they have been over the first 4 months on 2024. |
|
|
|
Post by James E on Apr 30, 2024 10:09:05 GMT
Survation NEW: Westminster Voting Intention. LAB 44 (-) CON 26 (-) LD 9 (-2) GRN 4 (-) RFM 10 (+2) SNP 3 (-) OTH 4 (-) F/w 23rd -25th April. Changes vs. 19th April 2024 Survation's polls tend to get less attention than they should, perhaps because they don't look particularly unusual. However, they have a very good track record from previous elections, and to my mind mind produce figures which have far greater predictive value than many more frequent pollsters. These figures are the same as their last poll, and also the MRP they did last month. For those who are interested, their figures are here: www.survation.com/https-www-survation-com-archive-2024-2/While I do not have a lot of trust in the seats figures they have produced, the detailed tables are interesting. One anomaly is that while they have Labour 35/33 ahead in Scotland, the SNP still hold 41 seats. I do not think it is possible to make accurate Scottish seat prediction on a sample of a little over 1,000 - so they only had 400 SNP2019 voters. Their English regional figures show the same pattern as YouGov, Opinium and Deltapoll, with the largest swings in the South and Midlands of England - typically around 19% compared to 15% in the GB figures. Their highest swings are in the South-West and the East Midlands. And they show the Conservatives losing proportionately more voters in their strong regions, as per all comparable data. |
|
|
|
Post by James E on Apr 25, 2024 9:47:19 GMT
Redfield Wilton Wales Westminster Labour leads by 22% in Wales. Highest Reform and Plaid %'s in our polling. Wales Westminster VI (22-23 Apr): Labour 40% (-9) Conservatives 18% (+2) Reform UK 18% (+3) Plaid 14% (+4) Lib Dem 6% (+1) Green 4% (-1) Other 0% (-1) Changes +/- 23-24 Mar A couple of comparative figures for Welsh Westminster polling: 10 YouGovs (with collective sample of a little over 1,000) show: Lab 47% (+6) Con 17% (-19) And the last 6 Opiniums (about 500 respondents) show: Lab 40% (-1) Con 22% (-14) I think that the true position for Wales is likely to be somewhere between these two, with a Con to Lab swing of around 9-10%, and Labour leading by 20-25 points. The last full YG Welsh Westminster poll in December was Lab 42/Con 20 . It's clear from all polling that Labour have made only modest gains in vote-share in Wales compared to 2019, while the Tories are faring 'less badly' than in England : their losses of 19 and 14 points per YP & Opinium contrast to 25 and 19 points in those companies' GB poll headlines. |
|
|
|
Post by James E on Apr 24, 2024 19:02:56 GMT
Pollingreport's overall seats figures are:
Lab 350
Con 204
SNP 48
LibDem 27
But if you feed Pollingreport's vote shares of Lab 43.5%, Con 23.4%, Ref 12.1% etc into Electoral Calculus, and then compare the outcomes in seats, you get this:
Lab 468 Seats (+118)
Con 77 Seats (-127)
LibDem 57 Seats (+30)
SNP 25 seats (-23)
[To add - I believe that Pollingreport have not updated their figures for Scotland since some time last year or possibly earlier - hence while their headline data shows the SNP's GB vote share falling from 3.9% in 2019 to 2.7% now, they still have the holding the same overall number of seats. Also, I don't think Anthony Wells is involved now. The occasional editorials do not appear to be his work]
|
|
|
|
Post by James E on Apr 24, 2024 9:25:48 GMT
" I havnt looked in detail at UK polling lately, but I expect the normally non voters are way under represented as usual. Should be 1/3 to 1/2 of all polled, and it never is. This is a group which isnt really being asked what they will do." This comment intrigued me, as it's not an angle of polling which I have looked at. Having checked 3 pollsters, it looks to me like both YouGov and Techne do have the correct proportion of non-voters from 2019 in their samples. If you take the size of their Con, Lab and LD samples from 2019, upweight for the fact that these are 89.4% of those who voted in GB and then divide by the total sample size, both appear to be using samples which consist of 66% 2019 voters (and 34% non-voters) in line with the turnout 4 years ago. However, Deltapoll's samples appear to be 86% 2019 voters and 14% non-voters. d3nkl3psvxxpe9.cloudfront.net/documents/TheTimes_VI_240417_W.pdfwww.techneuk.com/wp-content/uploads/2024/04/R111-UK-2024-4-19-DATA.pdf |
|
|
|
Post by James E on Apr 22, 2024 11:13:04 GMT
Survation asked which party their voters (close to 1,000 of them) would support if Reform stood aside, and this produced the following: Con 29.3% (+3.1) Lab 46.5% (+1.5) LD 10.9% (+0.5) Green 4.6% (+0.4) SNP+PC 4.1% (+0.2) Others 4.5% (+2.6) (Comparisons to Survation's standard VI question) The benefit to the Conservatives is therefore only about 20% of the redistributed Reform UK vote (of which they get 37% to Labour's 18%). Note that the 'others' pick up over 30% of those who would have supported Reform, and that is in addition to those who had given UKIP, Reclaim, Heritage or whatever as their first choice. Because of this, the combined vote for the Tories + other small RoC parties is probably still around 33%. www.bestforbritain.org/spring-2024-polling(see link at bottom of page for results with Reform standing aside) I dont understand your analysis, the comments in the paragraph dont seem to have anything to do with the tabulated results. ie table says 29% of reform would switch to con, seems clear. Text says benefit to con is 20%? Then says benefit to con is 37%? Perhaps I didn't explain this well. The percentage figures above are the totals for each party in the absence of Reform UK; and the figures in brackets reflect the difference to Survation's standard VI question in the same poll, with RefUK as an option. So the Conservatives gain 3.1 points of the 8.5% Reform vote, which is 37% of it, and Labour get 1.5 of 8.5%, etc... The overall split of that 8.5% RefUK vote per Survation (with about 1,000 respondents) is: Con 37% Lab 18% LD 6% Green 5% SNP / PC 3% Others 31% This can be compared to YouGov's findings last December, when Reform were on around 8%. When asked what they would do if Reform did not have a candidate, their respondents divided as: Con 31% Lab 3% LD 4% Green 3% SNP/PC 0% Would not vote 27% Others 20% yougov.co.uk/politics/articles/47759-conservatives-unlikely-to-win-over-reform-uk-supporters-at-next-election |
|
|
|
Post by James E on Apr 21, 2024 17:05:06 GMT
James E - I acknowledge the numbers, but they are polls not actual votes. My point is that just as there has never been a UK GE where UNS has fully predicted the results correctly, nor has there ever been a UK GE where a proportional loss model would have done so. There was certainly some evidence of proportional losses in 1997 but it was not universal. This is not to say that it won't happen in 2024, and my guess is that proportional losses may well be more marked than ever before, but my instinct is that reality will still come up short of current polling due to political factors like the past voting history of constituencies, the degree of party capacity and campaigning, tactical voting and so on. Putting it another way, UNS and a proportional loss model when applied to individual constituencies will both be 'right' and 'wrong' to some extent. It will interesting to see which is more accurate and that can only be determined after the election. Kellner clearly thinks it will be UNS, I tend to agree results will be more proportional to the size of the previous Tory vote, but its a bit too soon to call the UNS advocates "silly" as NeilJ did as we don't have the hard evidence to back it up yet. I posted some comparative figures a few months ago, to contrast YouGov's regional cross-breaks to what UNS would predict. The two followed almost completely opposite patterns, and I think this is now so clearly established that it really is "silly" (or at least ill-informed) to expect the results to conform to UNS. And there are good recent precedents for this, as we have seen the UK's electoral geography shift from a point in 2001 and 2005 when the Tories appeared to need a 13% lead to achieve a majority to 2017 and 2019, when just 3-4% would be enough. Peter Keller himself has provided some useful detail on what UNS 'should' look like, as one of his blogs mentioned a rule-of-thumb of '5 seats per 1% lead', which is the same as 10-per-1%-swing. I believe that the pattern of Labour targets corresponds closely to this, so for example a 10% swing on UNS gives Labour about 100 seats. However, MRPs and other detailed models all seem to work out in the region of 14-17 seats per 1% swing. And I would be very confident that we will get at least the low end of that range. This is pretty much what happened in 1997, when the 'semi-proportionate' outcome gave Labour 147 seats rather than the 105-ish that UNS would have predicted, and overall 14 seats per 1% swing. |
|
|
|
Post by James E on Apr 21, 2024 12:32:49 GMT
pjw1961 The Conservatives' larger loss of support in their stronger areas is a consistent finding across all the pollsters I have looked at for the past year or two. So to add to the Opinium figures above and my regular YouGov analysis, the same is true with Deltapoll. Their last 4 GB polls show the Tories retaining 56% of their 2019 across GB as a whole, but this is just 50% in the South of England and 46% in the Midlands. Another example, but a slightly different approach, is to look at some of the localised polling Survation did earlier this year. Their Jan and Feb GB polls showed the Conservatives retaining 60% of their 2019 vote share, but this fell to 53% and 54% in their constituency polls of Clacton and Godalming & Ash, and to 57% in their larger poll of rural constituencies. And there is some good analysis on Electoral Calculus of the large scale polling they had from Find Out Now, which they have divided into 8 groups of 75 seats, ranging from the Conservatives strongest to weakest areas. They describe these findings as being a consistent minus 53% (or 47% vote share retention) across the board, but I would divide the sample in two, with the Tories losing 55% in their strong seats, and 51% in their weak seats. (See para 3.2, two-thirds of the way down this article). www.electoralcalculus.co.uk/blogs/ec_vipoll_20240215.html[This pattern is also evident in by-election results. Leaving aside the exceptional case of Uxbridge, the Conservatives have had a higher overall vote retention of 56% in the 4 Labour-held constituencies where we have had by-elections since Dec 2022. But this becomes 53% retention on average in the 6 seats where they have been defending.]
|
|
|
|
Post by James E on Apr 21, 2024 10:32:38 GMT
While Opinium don't poll very frequently, their figures are useful as a check on the regional patterns I have described before as found by YouGov and others. In particular, their methodology provides a good test to Peter Kellner's contention (his 'Grumblers and Defectors' article from last year) that the high swings in safe Tory areas may be the product of the larger numbers of 'Don't Knows' there. Opinium's 6 poll averages since 23 Feb are: Headline GBCon 25.3% (-19.4) Lab 41% (+8) Swing Con to Lab 13.7% LondonLab 51% (+3) Con 23% (-9) Swing Con to Lab 6% South EnglandLab 36% (+13) Con 29% (-26) LD 13% (-4) Swing Con to Lab 19.5% MidlandsLab 40% (+7) Con 26% (-29) Swing Con to Lab 18% North EnglandLab 49% (+6) Con 23% (-16) Swing Con to Lab 11% WalesLab 40% (-1) Con 22% (-14) Swing Con to Lab 7% ScotlandLab 29% (+10) Con 19% (-6) SNP 33% (-12) Swing SNP to Lab 11% So even after excluding the Don't Knows and counting only those who give a voting intention, the Tories are losing most support in their safest regions. Their overall figures for vote-retention, that is their current VI as found by Opinium's last 6 polls compared to GE2019 are: Scotland 76% London 72% Wales 61% North England 59% GB Average 57%South England 53% Midlands 47% kellnerpolitics.com/2023/05/24/how-grumblers-and-defectors-have-laid-a-trap-for-our-pollsters/ |
|
|
|
Post by James E on Apr 21, 2024 9:27:00 GMT
Opinium Labour lead remains at 16 while Reform climb to 13% • Labour 41% (n/c) • Conservatives 25% (n/c) • Lib Dems 10% (n/c) • SNP 2% (-1) • Greens 7% (-1) • Reform 13% (+2) I wonder if the Reform voting intention figure in these opinion polls is a "plague on all your houses" sentiment as opposed to a thought through expression of support for a specific party and its policy platform? A repository for the disaffected and politically homeless The clues seems to be in what these "voters" give as their second preference vote, and it's by no means all for the Tories, and also in the performance of Reform in actual elections. Their "voters" aren't turning up. Of course, this is often a problem with voting intention polls when it's not always clear whether respondents are being prompted or not. All that said, it seems pretty obvious that Reform are benefitting from disillusionment with the established parties, but I suspect they will be drastically squeezed when people are required to cast actual votes. I wouldn't be at all surprised if many of those indicating support for Reform now won't vote at all. We should remember too that there has always been a Far Right vote in British politics and Reform will be the beneficiaries of this often unspoken voting group who may feel ignored and uncatered for during general elections, certainly since the the electoral demise of NF, EDL etc. One of the effects of Opinium's methodology is to increase the Reform UK VI, as well as that of the Conservatives (as Ref are getting nearly all their support from Con2019 voters). For those who share the view that the Reform VI is overstated by most polls, the answer is to look to Survation's figures as they do not prompt by party names at all. Their most recent poll this week put Reform on 8%, while their much larger MRP (of 15,000 respondents) showed them on 8.5%. To me, the most interesting part of that was not their seats prediction, which looked too high for Labour and failed to account for the usual overperformance by LDs and Greens in their strongest seats, but the follow-up question for those who stated their VI as Reform. Survation asked which party their voters (close to 1,000 of them) would support if Reform stood aside, and this produced the following: Con 29.3% (+3.1) Lab 46.5% (+1.5) LD 10.9% (+0.5) Green 4.6% (+0.4) SNP+PC 4.1% (+0.2) Others 4.5% (+2.6) (Comparisons to Survation's standard VI question) The benefit to the Conservatives is therefore only about 20% of the redistributed Reform UK vote (of which they get 37% to Labour's 18%). Note that the 'others' pick up over 30% of those who would have supported Reform, and that is in addition to those who had given UKIP, Reclaim, Heritage or whatever as their first choice. Because of this, the combined vote for the Tories + other small RoC parties is probably still around 33%. www.bestforbritain.org/spring-2024-polling(see link at bottom of page for results with Reform standing aside) |
|
|
|
Post by James E on Apr 18, 2024 13:21:17 GMT
Ipsos 💥RECORD BREAKING @ipsosuk VOTING INTENTION 💥 Lab: 44% Con: 19% Reform: 13% Lib Dems: 9% Greens: 9% Other: 6% Lowest the Conservatives have ever polled with us across 45 years of surveys (breaking last month’s record) One for Dave. Con + RefUK = 32, which is unusually low, so may be an outlier. This seems to be a 'house effect' with Ipsos, as the 32% in this poll is actually the highest they have recorded recently for Con+Ref. Their 6 most recent polls have been: Oct '23 Con 24 + Ref 4 = 28 Nov '23 Con 25+Ref 4 =29 Dec '23 Con 24+ Ref 7 =31 Jan '24 Con 27 +Ref 4 =31 Feb '24 Con 20+ Ref 8 =28 Apr '24 Con 19+ Ref 13 =32 As of a few years ago, Ipsos were unique in not weighting by past vote, which eliminates the issue of false recall. Other than that, I don't know what would cause their figures to diverge so far from most other pollsters (though YouGov frequently show Con +Ref as 34%). |
|
|
|
Post by James E on Apr 16, 2024 13:25:43 GMT
If I have understood R&W's write-up correctly, they have Westminster VI figures for the West Midlands 'region' as well as the 'county' figures quoted by Neil (above). If this is so, their second 'West Midlands' figures can be compared to GE2019 thus:
Lab 48% (+14)
Con 22% (-32)
Ref 11% (+10)
LD 6% (-2)
Green 5% (+3)
Oth 1%
So a very high Con to Lab swing of 23%, but in line with YouGov whose cross-breaks show a 22.5% swing in the Midlands (6-poll av).
redfieldandwiltonstrategies.com/west-midlands-mayoral-election-voting-intention-10-14-april-2024/en.wikipedia.org/wiki/2019_United_Kingdom_general_election_in_England#Regional_resultsEDIT & CORRECTION 15:00- the two sets of figures are : 1) excl the 8% Don't Knows, and 2) with the Don't Knows left in. The ones I have quoted above are the latter, whereas Neil quoted the former. Confusingly, R&W refer to ' the Conservatives trailing Labour in the region by 28%') and if this were so the swing would be 24%, with Lab up by 18 and Con down by 30 points. However, on checking the weightings by past vote, it looks like the figures are for West Midlands County (as opposed to 'region'), where Con and Lab were tied at 44% each in 2019. So it's only a 14% swing. Lab 52% (+8) Con 24% (-20) Ref 12% (+9) LD 7% (+1) Green 5% (+3) Oth 1%
|
|
|
|
Post by James E on Apr 15, 2024 13:08:40 GMT
mercian " While the graph is interesting, I'm not sure I agree with Freedman. I've said before that I live in a very strong Labour area, and it was one of the highest Leave-voting constituencies in the country. Similarly the Red Wall seats were traditionally Labour and people lent their votes to the Tories to 'Get Brexit Done'. Now they'll revert back." It is certainly true that the swing among Leave voters is likely to be way higher than for Remainers - among whom Labour led by 49/19 at GE2019.However, this is largely due to the Conservatives' losses to Reform UK. Typically, YouGov show Labour gaining around 8 points with Remainers and 10 with Leavers. Looking at the figures quoted by Neil above, they currently show Labour on 27% with 2016 Leavers, which is well up on the 14% they achieved at GE2019. However, at GE2017 Labour was reckoned to have taken 24% of the Leave vote. I don't have any good analysis of what proportion of 2016 Leavers backed Labour in 2015, although we do know that the 2015 Labour vote split 35/65 for Remain. If the converse applies then Labour's proportion of those who went on to vote Leave in 2016 is still lower now than it was before the referendum.
|
|
|
|
Post by James E on Apr 13, 2024 20:02:50 GMT
graham Re South Cambridgeshire - it was a very close contest between Con (46%)and LDs (42%) in 2019, and the slight boundary changes have had little effect as it's now notionally 45/41. As I live in the constituency, I can also tell you that my paper recycling is being boosted by regular communications from the LibDems with 'two horse race' bar charts and the like. I think that the seat is an almost certain LD gain, and I expect the Tories to achieve only around 30% of the vote, in line with their national performance. But I think there is also some further relevant material from the article about GE1997 which I linked above. They provide details of the averaged movements in Con-held seats where the LDs were 2nd (and Labour third) from 1992, and it is clear that Labour did less well in these, particularly in the closest contests: that is the 'marginal' ones where the Tories were less than 10% ahead of the LDs. The figures were : Conservative held seats from 1992, LDs in second place: Marginal (15 seats) :Con -9.7%, Lab + 3.3%, LDs + 2.6%Possible (30 seats) : Con -12.1%, Lab +6.8%, LDs + 2.4% Safe (98 seats) : Con -12.6%, Lab + 8.8%, LDs -0.4%. So while Labour did slightly increase its vote share in these tight Con/LD marginals, they produced the lowest rises in the Labour vote of any type of seat at GE1997 (see page 14 of 18 in the article I linked above). Labour did somewhat better in seats where the LDs trailed the Tories by 10%+ or 20% +from 1992, but even here it is clear that starting in 3rd place was a disadvantage compared to being in 2nd. Also, I think it is worth considering the 2021 Cambridgeshire and Peterborough Mayoral Election, which was conducted using the supplementary vote system. First choices for the South Cambridgeshire Council area (which is larger but demographically very similar to the S Cambs constituency) were: LD 44% Con 34% Lab 22%
|
|
|
|
Post by James E on Apr 13, 2024 19:12:19 GMT
pjw1961 "... If you are criticising MRP models because they presently generate proportional swings, you need to be able to explain why we should expect uniform national swings right now, and why voters are able to internalize the tactical situation in their constituency several months before an election before parties have started actively campaigning."
In other words, when voters see that Labour is making little effort in Mid Devon and the Lib Dems are, they will switch their VI from Lab to LD. Same in South Cambridgeshire, another recent debate on here. I remain dubious about Labour capturing the Isle of Wight seats to give another example. " I agree with you on South Cambridgeshire because of the tactical position - the LDs were a close second last time. But on Devon Central, I think there are several reasons for believing a Labour win is possible. The swing required is 15%, and Labour were in 2nd place, 10-points ahead of the LDs at the 2019 election. en.wikipedia.org/wiki/Central_Devon_(UK_Parliament_constituency)One factor is that Labour have clearly fared well in this constituency in the past two elections, achieving a 6.1% swing in 2017 (v 2.0% for England as a whole), and suffering only a 1.7% swing to the Tories in 2019 (v 4.5% for England). But more importantly, there is a distinction to be drawn between the need for a strong local campaign in Local Elections, and trends at General Elections. Here, the evidence is of the wider swings sweeping up untargeted seats as well as officially targeted ones. Looking at the analysis of GE1997 (linked below), Labour fared 'better than average' across the board in Conservative-held seats where they were already second. www.dannydorling.org/wp-content/files/dannydorling_publication_id1318.pdfIf you look at the analysis of Con-held seats, listed as Marginal (under 10% Con lead), Possible (10-20% Con lead), and Safe (20%+ Con lead), Labour fared best in the Safe-Con ones. And this was despite the targeting of only the 90 most marginal at the time, all of which would have been 'Marginal' or 'Possible' Con-held seats, of which there were 106. Conservative held seats from 1992, vote change at GE1997 ( per pages 13 & 14 of linked article). Marginal (63 seats) : Con -12.3%, Lab +12.2% Possible (43 seats) : Con - 12.2%, Lab +13.7% Safe (69 seats) : Con -13.1%, Lab +13.8% These figures were in the context of national changes of Con -11.5%, Lab + 9.6%. So Labour outstripped UNS in 1997 by nearly a 3% further swing in untargeted safe Tory seats.
|
|
|
|
Post by James E on Apr 13, 2024 12:06:25 GMT
Another way of looking at those YouGov regional & Social Class figures I have listed above is a comparison to the outcome from GE1997, which is overall a 6% con to Lab swing (a 24 point swing compared to 12 points).
ABC1 Class
5% Con lead to 26% Lab lead
15.5% Con to Lab swing
South England
9% Con lead to 12% lab lead
10.5% Con to Lab swing
London
18% Lab lead to 28% Lab lead
5% Con to Lab swing
Midlands
14% Lab lead to 23% Lab lead
4.5% Con to Lab swing
North England
30% Lab lead to 33% Lab lead
1.5% Con to Lab swing
Wales
35% Lab lead to 29% Lab lead
3% Lab to Con swing
C2DE Class
30% Lab lead to 20% Lab lead
5% Lab to Con swing
Scotland
21% Lab lead to 2% Lab lead
9.5% Lab to SNP swing.
|
|
|
|
Post by James E on Apr 13, 2024 10:09:13 GMT
I've updated my usual analysis of the last 6 YouGov GB polls. This excludes the YouGov MRP, in which they use adjusted figures to counteract the 'Don't knows effect', so that these figures reflect an average of a 24-point Labour lead (44/20). As with the previous YouGov MRP in January, their adjusted figures came out with a lead around 7 points lower.
Overall GB
Lab 44% (+11)
Con 20% (-25)
Swing Con to Lab 18%
London
Lab 54% (+6)
Con16% (-16)
Swing Con to Lab 11%
South England
Lab 37% (+14)
Con 25% (-30)
LD 13% (-4)
Swing Con to Lab 22%
Midlands
Lab 44% (+11)
Con 21% (-34)
Swing Con to Lab 22%
North England
Lab 50% (+7)
Con 17% (-22)
Swing Con to Lab 15%
Scotland
Lab 34% (+15)
Con 14% (-11)
SNP 32% (-13)
Swing SNP to Lab 13%
Wales
Lab 46% (+5)
Con 17% (-19)
Swing Con to Lab 12%
Remain 2016
Lab 57% (+8)
Con 11% (-8)
Swing Con to Lab 8%
Leave 2016
Lab 23% (+9)
Con 33% (-41)
Swing Con to Lab 25%
ABC1 Class
Lab 46% (+13)
Con 20% (-23)
Swing Con to Lab 18%
C2DE Class
Lab 40% (+7)
Con 20% (-28)
Swing Con to Lab 17.5%
A few changes to note over the past few months: up to a few months ago, the Tories were faring relatively better with C2DE voters; this is no longer the case, and they are now losing a larger share of their support with this group than with ABC1s. The recent further drop in the Tories' VI has hit them particularly badly in the Midlands and North of England. And the same pattern of Conservative support falling by more in their stronger demographics still applies. Their vote-retention (averaging 45% of their 2019 vote) is as follows by region:
Scotland 56%
London 50%
Wales 47%
South England 45%
North England 44%
Midlands 38%
It's also worth noting that the drops in the Tories support don't all correspond to the largest rises in Labour's VI - 2016 Leave voters are the obvious example. Labour's strongest gains in support (but not their largest swings) are in the South of England, ABC1 voters, and Scotland.
|
|
