|
|
Post by James E on Mar 19, 2024 12:07:23 GMT
A couple of weeks ago there was some debate on this site about the possibility of 'false recall' in relation to the 2016 EU referendum. While we know that false recall is an issue in relation to past General Elections, the argument put forward by some was that, as a one-off, it was far less likely for the referendum.
One factor which is useful to know in all this is that we are close to the point where the numbers of living 2016 Leave and Remain voters are equal. My own figures, using the ratio of 65/35 for older 2016 voters, suggest that we currently still have a very narrow majority of (living) 2016 Leavers, but by this time next year there will be more 2016 Remainers.
With this in mind, I have noted an odd pattern in Deltapoll, who rely on recalled vote. In 11 of their last 12 polls, their raw sample has contained more 2016 Remainers than Leavers. If the numbers are now really 50/50 in the population, as I believe, this is as likely as tossing a coin 12 times, and getting 11 'heads'. The probability of this happening is approximately 1 in 340.
And there are more quirks to Deltapoll's 'Remain' sample figures. If we start with YouGov's recent figures, their last 5 polls show the Con VI with 2016 Remainers as 12,12,14,11 and 11, averaging at 12%. This compares to a figure of 19% Remainers having voted for the Tories per their own analysis of the 2019 General Election, and in fact shows that the Tories are doing significantly "less badly" with Remain voters than Leavers.
But compare this with the 6 Deltapolls we have had for the same period: their Con VI with 2016 Remainers is 19,16,15,20,20 and 16, averaging at 17.7%. This is an astonishing figure, as polling generally shows the Tories losing 40%-50% of their 2019 support across all demographics. Yet Deltapoll have them losing just 7% of their 2019 support among Remainers, compared to 36% per YouGov.
Looking at the 2016 Leave samples with both pollsters, their figures show similar Con leads, albeit by 36/32 with Delta and 32/27 with YouGov. So the substantial overall difference in the Labour leads shown by YouGov (av 24%)and Delta (av 19%) must be due to their Remain samples, plus their figures for those who did not vote in the referendum.
In the context of General Election polling this might not seem to make much difference. However, Deltapoll also poll regularly on the question of Stay Out/Rejoin, and on this the divergence between pollsters could lead to very different conclusions. Deltapoll's figures are noticeably lower for 'Rejoin' than others, varying around an average of 55/45. YouGov put the current state of opinion closer to 60/40 for Rejoin.
EDIT
I have looked further back into this, and the same pattern is apparent. While YouGov show ConVI as around 12% with 2016 Remain voters, Delta have the Tories' support with this group as averaging 18.7% across their past 12 published polls. This is almost at the level it was at the 2019 Election - measured as 19% per YG's analysis or 20% per Ipsos. But these two companies show similar levels of support for the Tories in their 'Leave' samples, 35% per YouGov, 38% per Delta.
|
|
|
|
Post by James E on Mar 18, 2024 19:53:38 GMT
We have not yet reached the point of '4 in a row' for polling movements, but the most recent 2 polls show Labour's lead increasing substantially. R&W's poll reported this evening shows Labour's lead at 5 points higher than their average so far this year, and with Deltapoll, it's 6 points higher. And there's another indicator that things have moved: of the most recent 8 polls from various pollsters, 4 show their highest Labour lead of any poll this year, and another two their equal-highest. Only YouGov and WeThink do not show unusually large leads. 11 March Lord Ashcroft : Lab lead 22% (highest of 3 polls in 2024) 11 March More in Common : 15% Lab lead (equal highest of 5 polls in 2024) 13 March YouGov : 24% Lab lead 14 March Survation : 19% Lab lead (highest of 3 polls in 2024) 14 March Techne :22% Lab lead (equal highest of 10 polls in 2024) 15 March WeThink : 18% Lab lead 17 March R&W : 26% Lab lead (highest of 12 polls in 2024) 18 March Deltapoll :23% lab lead (highest of 10 polls in 2024) en.wikipedia.org/wiki/Opinion_polling_for_the_next_United_Kingdom_general_election |
|
|
|
Post by James E on Mar 16, 2024 19:23:58 GMT
The two MRPs we have had in 2024 support your optimism: YouGov showed the LDs taking 48 seats on 12.5% of the vote in January, and FON gave the LDs 53 seats on 11% in Feb. Current polling averages around 10-11%, and Ashcroft's figures are an outlier, almost certainly a product of his non-standard Voting Intention question. Having said that, Electoral Calculus made some interesting findings last month on LD prospects in their more distant targets. While the LDs are well-placed in their 60-ish strongest seats , FON/EC find them being overtaken by Labour in most of the next 40 seats, in all of which the LDs were in second place in 2019. So alongside their 40+ LD gains, they also found 19 Labour gains in seats where the Lib Dems were in 2rd place (and Labour third) in 2019. www.electoralcalculus.co.uk/blogs/ec_lib2seats_20240226.htmlAnother way of looking at it is to compare to years when the Tories did very badly in the relatively recent past. Tory% Tory seats LibDem% LibDem seats 1997 30.7 165 16.8 46 2001 31.7 166 18.3 52 2005 32.4 168 22.0 62 Not strictly comparable of course because the Tories would have to recover quite a bit even to get to those levels, but so would the LibDems. Is there any real evidence of the likelihood of an increased level of tactical voting? The total of seats the LibDems get depends largely on their performance relative to the Conservatives. So for the three elections you quote above, the Tories led the LDs by 14% in 1997, 13.4% in 2001, and 10.4% in 2005. In the two MRPs this year, the Tories led by 13.5% with YouGov (and 48 LD seats), and 11% with FON (and 53 LD seats). This explains why the FON result is better for the LDs in terms of seats, even though their VI percentage is lower. As per the third para of my post immediately above, the historically low (11%-ish) current LibDem VI limits the number of seats that they can win beyond about 55. They can be expected to finish in second place in far fewer seats at the next GE. As for evidence of an increased level of tactical voting : parliamentary by-elections give some good evidence of this, as the Labour vote has been very heavily squeezed in recent Con/LD contests, often reduced to a fraction of what it was at GE2019. This has happened before, but has been even larger in the past few years than it was in (for example) 1992-97. Also, the polling by Lord Ashcroft which I linked yesterday shows 56% of the current LabVI who say that they would be 'willing to vote for the LibDem candidate if they stood the best chance of beating the conservative candidate' (and only 30% who would be 'reluctant' to do so). A smaller sample of 231 who intend to support the LDs would be willing to switch to a better placed labour candidate, but there are also more who say they are 'reluctant'. I don't have any comparable data from 1997 to 2005 for this, but in general, the lower the overall Tory vote is, the more they become vulnerable to the effects of tactical voting. |
|
|
|
Post by James E on Mar 16, 2024 10:59:43 GMT
There was a comment a couple of days ago that the Labour vote share is slightly down in recent polls. To check this, I've looked at the most 9 most recent with fieldwork ending from 10 March onwards. In each case, comparing their most recent Lab VI with the average of the same pollster's previous polls in 2024:
15 Mar We Think 43% (-2.4)
14 Mar Survation 45% (+0.7)
14 Mar Techne 44% (+2)
13 Mar YouGov 44% (-1.7)
11 Mar MiC 42% ( 0 )
11 Mar Deltapoll 42% ( 0 )
11 Mar Lord Ashcroft 45% (+1.5)
10 Mar R&W 42% (-2.6)
10 Mar Savanta 43% (-0.6)
The aggregate loss is -4.5, so the Lab VI is down by 0.5% on average compared to previous 2024 polls.
However, the same polls also show an aggregate rise of 6.3 points in Labour lead, or 0.7% on average. So overall, the Tories are down by 1.2% across the same 9 polls.
These polls also show the combined 'Con+Ref' VI averaging 35.5% - so a bit less than some figures discussed here. Overall averages are:
Lab 43.5%
Con 24.3%
Ref 11.2%
LD 10%
Grn 5.3%
|
|
|
|
Post by James E on Mar 16, 2024 10:30:31 GMT
James E While the Ashcroft poll on the face of it doesn't look good for the Lib dems and is highly likely to not reflect vote share, the tactical voting element brings dozens of otherwise unwinable seats for both the Lib dems and Labour from the Tories into play. I don't know of anyone bonkers enough to anticipate a lib dem majority , but there are plenty.of reasonably sane people, including me, who thinks it's quite probable that the Lib dems could come close to.or exceed their best parliamentary representation, despite no where near our best percentage of the vote. In order for this to happen it requires a disastrous defeat for the Tories a result I'm more than happy to see, I'm less sanguine about the 200+ seat majority it also entails for Labour. .... The two MRPs we have had in 2024 support your optimism: YouGov showed the LDs taking 48 seats on 12.5% of the vote in January, and FON gave the LDs 53 seats on 11% in Feb. Current polling averages around 10-11%, and Ashcroft's figures are an outlier, almost certainly a product of his non-standard Voting Intention question. Having said that, Electoral Calculus made some interesting findings last month on LD prospects in their more distant targets. While the LDs are well-placed in their 60-ish strongest seats , FON/EC find them being overtaken by Labour in most of the next 40 seats, in all of which the LDs were in second place in 2019. So alongside their 40+ LD gains, they also found 19 Labour gains in seats where the Lib Dems were in 2rd place (and Labour third) in 2019. www.electoralcalculus.co.uk/blogs/ec_lib2seats_20240226.html |
|
|
|
Post by James E on Mar 15, 2024 10:55:03 GMT
I don't think anyone has highlighted the Lord Ashcroft 'poll' of 5,300 respondents of 7-11 March. Con 23% Lab 45% LD 6% Ref 11% Grn 8% This is not a standard polling question, but is based on respondents' answers of how likely they are to vote for each party, on a scale of 0 to 100%. Their highest score (where it exceeds 50%) is counted for the VI figures, but note that an unusually high proportion of respondents (36%) don't meet the 50% threshold. Ashcroft has done the same in 2 previous polls in the past few months, which came out as 27/43 and 27/44. It seems that his unusual methodology always produces a low LibDem VI figure (6,7,& 6% ). There is some interesting detail in the question asked of (current) Lab and LD voters as to whether they would switch to the other party if better placed in their seat. The short answer is that a majority of LDs and a large majority of Lab voters would do so. view.officeapps.live.com/op/view.aspx?src=https%3A%2F%2Flordashcroftpolls.com%2Fwp-content%2Fuploads%2F2024%2F03%2FLord-Ashcroft-Polls-Political-survey-March-2024-Full-tables.xls |
|
|
|
Post by James E on Mar 12, 2024 12:36:25 GMT
I haven't seen these polls reported: Deltapoll 8-11 March 1502 GB adults Con 27% (=) Lab 44% (+3) LD 10% (+1) SNP 2% (-1) Grn 4% (=) Ref 11% (-1) Oth 2% (-1) Changes from 1-4 Mar More in Common 8-11 March 2027 GB adults Con 27% (-1) Lab 42% (-1) LD 10% (=) SNP TBC Grn 6% (+1) Ref 10% (+1) Oth TBC Changes from 23-27 Feb These might look like small Labour leads, but in both cases they are really par-for-the course with these pollsters. Deltapoll's previous 8 polls so far in 2024 average a 17-point lead, while a 15-point lead with MiC is in fact slightly higher than their norm of 14%. Indeed, MiC's average is currently the lowest of any pollster, as Opinium are typically now at around a 15-16 point Lab lead. |
|
|
|
Post by James E on Mar 11, 2024 12:40:56 GMT
For comparison to Opinium's figures, the current YouGov 6 poll 'regional' averages, from 31 Jan onwards are:
GB headline figures:
Lab 45.5%
Con 21.3%
Swing Con to Lab 18%
London
Lab 56% (+8)
Con 17% (-15)
Swing Con to Lab 11.5%
South England
Lab 38% (+15)
Con 27% (-28)
LD 12% (-5)
Swing Con to Lab 21.5%
Midlands
Lab 45% (+12)
Con 27% (-28)
Swing Con to Lab 20%
North England
Lab 54% (+11)
Con 17% (-22)
Swing Con to Lab 16.5%
Wales
Lab 47% (+6)
Con 18% (-18)
Swing Con to Lab 12%
Scotland
SNP 32% (-13)
Lab 36% (+17)
Con 13% (-12)
Swing SNP to Lab 15%
|
|
|
|
Post by James E on Mar 11, 2024 12:19:24 GMT
We have now had 6 polls from Opinium so far in 2024, all with Labour leads of 14-18 points, and averaging lab 42.0%, Con 26.3%, so an overall swing of 13.6% on the 2019 result.
The regional pattern is much the same as for YouGov and various MRPs, with the Tories losses being largest in the South and Midlands of England. These figures of course reflect Opinium's re-weighting, which still reduces the overall swing by about 3% (or 6 points).
London
Lab 51% (+3)
Con 23% (-9)
Con to Lab swing 6%
South England
Lab 36% (+13)
Con 31% (-24)
LD13% (-4)
Con to Lab swing 18.5%
Midlands
Lab 40% (+7)
Con 29% (-26)
Con to Lab swing 16.5%
North England
Lab 51% (+8)
Con 24% (-15)
Con to Lab swing 11%
Wales
Lab 42% (+1)
Con 24% (-12)
Con to Lab swing 6.5%
Scotland
SNP 33% (-12)
Lab 32% (+13)
Con 16% (-9)
Swing SNP to Lab 12.5%
The pattern of the Tories vote losses follows YouGov, and the MRPs, too, with them losing a larger proportion in their strongest areas. Their current averages, measured as a proportion of their 2019 vote, per region are:
London 72%
Wales 66%
Scotland 64%
North England 61%
GB overall 59%
South England 56%
Midlands 53%
|
|
|
|
Post by James E on Mar 5, 2024 17:57:24 GMT
Talking about the blue wall Redfield Wilton Labour leads by 9% in Blue Wall. Lowest % for Conservatives since October 2022. Blue Wall Westminster VI (3 Mar): Labour 37% (–) Conservative 28% (-2) Liberal Democrat 19% (-2) Reform UK 10% (+3) Green 5% (+1) Other 1% (-1) Changes +/- 11 Feb Compared to GE2019, that's: Lab 37% (+16) Con 28% (-21) LD 19 % (-8) It's the equal-lowest VI for the Conservatives since they started this series of polls in Oct 2022. The LDs generally fare a bit better than this, but are typically around 3 or 4 points down from the 27% they got in 2019, in a set of seats which includes a high proportion of their targets for this year. Other than that, I have rather lost interest in R&W's 'walls'. Their blue wall is an odd mixture of 32 seats across the South of England regions plus 10 in London, with all seats having a relatively high 2016 Remain vote. In the last 5 polls dating back to early November, it shows an average of 17% Con to Lab swing, compared to 16% for the 17 GB R&W polls over the same period. Their Red Wall figures are also rather 'par-for-the course' , showing an average 15% swing against the same GB average of 16%. This would still be enough for Labour to win back all of the seats in question, but suggests to me that the Tories may be getting a slight incumbency boost in their 2019 gains. |
|
|
|
Post by James E on Mar 5, 2024 11:16:55 GMT
Re Bristol Central The Electoral Calculus figures are almost certainly based on the FON/Electoral Calculus MRP of 18,000 voters in Jan-Feb. They showed the Greens gaining Bristol Central by 52/39 (compared to 51/39 per EC's current nowcast) www.electoralcalculus.co.uk/blogs/ec_vipoll_20240215.htmlBut there is an alternative in the YouGov MRP (of 14,000) from Dec-Jan. Their write up this states: "The Greens would hold on to their Brighton Pavilion seat, without making any further gains – although they come incredibly close in Bristol Central at 38% to Labour’s 40%. " So a 16% swing in Bristol Central compared to EC's 23%. This was in the context of YouGov's 'adjusted' overall vote shares of Lab 39.5%, Con 26% and Greens 7.5% - without re-weighting, the figures would probably have shown Labour several points higher as in YouGov's normal polls. yougov.co.uk/politics/articles/48371-yougov-mrp-shows-labour-would-win-1997-style-landslide-if-election-were-held-todayNot all MRPs are equal, and I would trust YouGov more than others. However, FON's finding are far more credible than some others such as Focaldata, and their sample is a decent size: MRPs and seat forecasts with fewer than 10,000 respondents should be taken with a very heavy pinch of salt. |
|
|
|
Post by James E on Mar 4, 2024 11:20:23 GMT
.... But the Rochdale result also illustrates where Reform UK is getting support from. It is very much a party for discontented Con2019 voters (who were also overwhelmingly Leavers in 2016) rather than for all Leavers. So they got 13% from a 62% Con2019 vote in Wellingborough and about half of that from a 31% Con2019 vote in Rochdale. The fact that Rochdale voted almost as strongly for Leave as Wellingborough did not help Reform UK there. Although Reform is a descendant of the Brexit party, that doesn't mean it will only attract Leave voters. ..... Per YouGov's figures, Reform UK is drawing support overwhelmingly from those who voted Leave in 2016, and even more so from Con2019 voters. Over the 6 YouGov polls since 23 Jan, Reform's 12.5% average VI with YouGov divides as 27% Leavers and 2% Remainers by respondents who give a voting Intention. By 2019 votes, RUK take 26% of Con2019, but only 1.5% from each of Lab2019 and LD2019. Of the 12.5% VI. As a rough approximation, Reform's 12.5% consists of 12% from Con2019 (or BXP) and 0.5% from Lab +LD. Incidentally, these figures also show that RUK are taking very few new voters or non-voters from either 2016 or GE2019, as the overall percentage of those who voted then supporting them is above 12.5%. The low Lab to Ref figures is remarkable in the context of Labour having taken 14% of the Leave vote in 2019 (compared to 74% for the Tories). So Labour took about 1/5th of the amount of Leave voters that the Tories took. A party which appealed equally to all 2016 Leave voters, as opposed to Conservative-supporting ones, might expect to take around 5% of the Lab2019 vote, along with the 26% they are taking from the Tories. But that clearly isn't happening - hence my comment about the different levels of support for RUK at Wellingborough and Rochdale. |
|
|
|
Post by James E on Mar 3, 2024 19:16:16 GMT
leftieliberal . Rochdale, Kingswood and Wellingborough are all constituencies where Reform UK should have been expected to perform above their GB average. These three constituencies voted collectively by more than 60% for Leave in 2016. And at the 2015 General Election UKIP out-performed their overall 12.5% vote share in each of them, getting 19.6% in Wellingborough, 14.8% in Kingswood, and 18.8% in Rochdale. If Reform are really in line for 10% of the vote they might have been expected to get around 14% across these three seats. Their actual average of 9.9% was therefore still an underperformance, and the Rochdale result especially so. But the Rochdale result also illustrates where Reform UK is getting support from. It is very much a party for discontented Con2019 voters (who were also overwhelmingly Leavers in 2016) rather than for all Leavers. So they got 13% from a 62% Con2019 vote in Wellingborough and about half of that from a 31% Con2019 vote in Rochdale. The fact that Rochdale voted almost as strongly for Leave as Wellingborough did not help Reform UK there.
|
|
|
|
Post by James E on Mar 3, 2024 17:28:24 GMT
That Godalming & Ash Constituency poll: Con 29% (-24) LD 35% (-4) Lab 23% (+15) Ref 8% Green 3% Oth 2% Sample of just 507, so a high margin of error. And on first sight, the figures look outlandish. However, the proportions are right in line with general polling: the Tories are reduced to 0.55 of their 2019 vote share - which is consistent with them getting 24% in GB polls. And Labour pick up 62% of the Conservatives' lost vote share, which would take them to around 45-46% for GB. Perhaps the biggest surprise should be the LDs being down a bit in a target seat, although that too is the same as in national polls, and of course even this does not prevent them from coming on top. Probably more likely that Labour has picked up LD votes whilst LDs get votes at Tory expense. A lot of churn. Yes, YouGov show Labour taking around 17% of Con2019 voters who give a voting intention, and around 37% of LD2019, so the overall picture is more complicated. However, in most of the analysis of local cross-breaks I have done, the pattern is that the more (numerically) that the Conservatives lose, the more Labour gain. What seems to be constant across all detailed polling is that the Conservatives' losses are proportionate rather than linear, and if anything are losing a higher proportion in their stronger areas. |
|
|
|
Post by James E on Mar 3, 2024 15:31:35 GMT
That Godalming & Ash Constituency poll:
Con 29% (-24)
LD 35% (-4)
Lab 23% (+15)
Ref 8%
Green 3%
Oth 2%
Sample of just 507, so a high margin of error. And on first sight, the figures look outlandish. However, the proportions are right in line with general polling: the Tories are reduced to 0.55 of their 2019 vote share - which is consistent with them getting 24% in GB polls. And Labour pick up 62% of the Conservatives' lost vote share, which would take them to around 45-46% for GB. Perhaps the biggest surprise should be the LDs being down a bit in a target seat, although that too is the same as in national polls, and of course even this does not prevent them from coming on top.
To Add:
The movements in this Survation poll are very similar indeed to the South of England cross-breaks from an average of 5 Opinium polls so far in 2024. These are:
Con 31% (-24)
LD 14% (-3)
Lab 38% (+15)
That's with the 'adjusted' headline figures Opinium shows. YouGov's 'unadjusted' figures show Con down 29 points , Lab up 16 and LDs down 4.
To Add 2 (5/3/24):
I've noticed that the Focaldata/Electoral Calculus MRP done in Jan/Feb produced figures for the new Godalming & Ash seat of:
Con 32% (-21)
LD 35% (-4)
Lab 20% (+12)
|
|
|
|
Post by James E on Mar 2, 2024 20:33:24 GMT
oldnat"you could agree with Michael Gove that ‘tens of thousands’ of Scots could switch from SNP to Con!" www.thenational.scot/news/24158204.michael-gove-claims-tens-thousands-scots-switch-snp/It looks to me like Mr Gove is technically correct. SNP to Con switching is 2% in the recent Survation Scottish Westminster poll, and also 2% in the two most recent YouGov Scottish polls. If we allow for 80-85% turnout of SNP2019 voters, 2% (x 0.85) of the 1,240,000 people who voted for the SNP in 2019 is roughly 20,000 voters - which is literally 'tens of thousands' . He might have added that polling suggests the Conservatives picking up a further ten thousand new votes from Scotland's quarter of a million first time voters, plus a further 25,000 or so from the LDs and Labour. On the other hand, polling also suggests that the Tories will lose the support at least 300,000 of the 693,000 who voted for them in Scotland in 2019, but Gove doesn't mention that.
|
|
|
|
Post by James E on Mar 1, 2024 11:34:58 GMT
Latest Yougov Labour lead at 26 points in latest YouGov poll for The Times CON 20 (=) LAB 46 (=) LIB DEM 7 (-2) REF UK 14 (+1) GRN 7 (=) SNP 3 (=) Fieldwork 28 - 29 February Personally I really don't think Reform are anywhere near 14% Tables are here: d3nkl3psvxxpe9.cloudfront.net/documents/TheTimes_VI_240229_W.pdfOne detail to highlight: their sample of 2016 Leave voters is: Con 22% Lab 21% Ref 20% Don't Know 19% Would not vote 8% The Conservatives took around 60% of the 2016 Leave vote at the 2019 election (or 74% of leavers who voted then, per YG). Even using the 'decided' voters figure of 31%, they are keeping barely half. This Leave sample is closer than normal, though. The 6-poll average for 'decided' Leave voters is Con 36%, Lab 27%. So a lead of 9%, compared to the 60% lead (74/14) YouGov found for leavers who voted in GE2019. |
|
|
|
Post by James E on Mar 1, 2024 10:48:19 GMT
John Curtice on Rochdale. Curtice said this was a “unique” byelection, making it harder than usual to draw wider conclusions from the result. He said: [This was] a unique contest contested by a candidate with a unique ability to appeal to the Muslim population in a town that also has a past record of voting on local issues. He said that the Tories suffered their biggest fall in vote share in a Labour seat in a byelection in this parliament. He said: Frankly, here is another poor Conservative performance. The Conservative vote down by 19 points – it’s the biggest drop in the Conservative vote in a Labour-held seat in a byelection in this parliament … There is no sign here of the electoral gloom that hangs over the Conservative party is in any way dissipated by Sir Keir Starmer’s difficulties. He said Labour suffered its biggest drop in support in a post-war byelection in Rochdale. The Tories' loss of support in Rochdale was (proportionately) the same as in Wellingborough, with them reduced to just under 40% of their 2019 share of 31.2%. And as in Wellingborough, Reform UK took almost exactly 20% of the Con2019 vote share, with 6.3%. This was a poor showing for them in a constituency where UKIP took three times that figure (18.8%) in the 2015 General Election, and they were even down 2 points on the Brexit Party's 2019 share. Looking back over by-elections since Sunak became PM, there has been a progressive worsening of the Conservatives vote retention in by-elections: Dec 22 Chester 58% (of the Con 2019 vote) Dec 22 Stretford &U 58% Feb 23 W Lancs 70% July 23 Somerton &F 47% July 23 Uxbridge 86% July 23 Selby 57% Oct 23 Tamworth 61% Oct 23 Mid Bedfordshire 51% Feb 24 Kingswood 62% Feb 24 Wellingborough 39% Feb 24 Rochdale 38% Leaving out the obvious exceptional case of Uxbridge, the Tories have averaged 53% retention in the 6 seats they had held, and 56% retention in the 4 Labour-held ones. |
|
|
|
Post by James E on Feb 29, 2024 12:20:20 GMT
You pointed out (correctly) that if Labour made as many gains as in 1997, this would give them a majority of 48. However, that's not the same as a 'landslide the same size as 1997'. A similar swing to 1997 (10%) might produce a similar result to that, but this would mean Labour finishing just 8 points ahead of the Tories compared to the current average of 19%. But if we got the same vote shares as in 1997 - Lab 43%, Con 31% - there would probably be a Labour majority just over 100. YouGov's MRP at the start of the year, showed a majority of 120 on a lead which was only slightly higher than that (39/26). It will be interesting to see how it turns out, but I'm still interested to know people's opinions on whether in the event of Tory wipeout there will be a centre-right party, and if so, who? My money would be on Labour. Current polls show a combined share for Con+Ref of about 35%, divided as Con 25%, Ref 10% on average. It's these voters who will form the basis of any RoC party of the future, and I'd be 90% confident that it will be the Conservatives who garner most of this potential vote, either at the next General Election or at the one after that. While Reform are still gaining support from 2019 Conservative voters, they must know that they stand little chance of gaining many seats, even if their current 10% were to rise to 15%. But a consequent fall in the Tories VI from 25% to 20% would be disastrous for them under FPTP. As pjw has mentioned, Farage & Co show little interest in establishing a party of government. His aim has always been to drag the Tories to the right, rather than to supplant them as the RoC party in the House of Commons. In that, he has been highly successful. So at some point, either before or after the GE, I would expect Reform UK's support to melt away back to the Tories very quickly just as UKIP's support did in the run up to GE2017, or the Brexit Party in the run up to GE2019. I suspect that the Conservative are currently is discussion with Farage with a view to gaining his endorsement prior to the Election. Starmer's surprising mention of Farage at PMQs yesterday suggests that he is aware of this. |
|
|
|
Post by James E on Feb 28, 2024 23:30:14 GMT
You could be right about all that though I suspect the likely defeat will not be as catastrophic as some are predicting/hoping. I showed a few days a go that if Labour get a landslide of the same size as 1997, they'll have a majority of about 40. Of course Tories are likely to lose some to LibDems too and possibly one or two others. Anyway, to get to my point - though it's imperfect the usual model for political positions is left v right. If the traditional centre-right party were to be virtually wiped out what happens then? There will still be a lot of centre-right potential voters. If there weren't there'd be no such thing as a centre ground, even if it does move about a bit. So who represents them? Do Labour under Starmer become the new centre-right party, or the LibDems, or Reform? I see that in Canada after the fabled wipeout, a reborn Conservative Party were in power as recently as 2011 and have been the official opposition since. Politics may well shift around for the next few years. Compared to the virtual two-party hegemony of the 1950s and 1960s things are already a lot more complex. I see that process continuing, and it may not be long (15-20 years?) before coalitions become more common and calls for PR become harder to resist. You pointed out (correctly) that if Labour made as many gains as in 1997, this would give them a majority of 48. However, that's not the same as a 'landslide the same size as 1997'. A similar swing to 1997 (10%) might produce a similar result to that, but this would mean Labour finishing just 8 points ahead of the Tories compared to the current average of 19%. But if we got the same vote shares as in 1997 - Lab 43%, Con 31% - there would probably be a Labour majority just over 100. YouGov's MRP at the start of the year, showed a majority of 120 on a lead which was only slightly higher than that (39/26). |
|
|
|
Post by James E on Feb 28, 2024 17:29:49 GMT
Redfield Wilton Red wall Labour leads by 24% in the Red Wall. Lowest Conservative vote % since Sunak became PM. Red Wall Voting Intention (25 February): Labour 49% (+1) Conservative 25% (-3) Reform UK 14% (–) Liberal Democrat 6% (+2) Other 7% (–) Changes +/- 30-31 January Loving your work Neil. I know this could all be moe, but can we at least for a brief, gleeful moment, hope that it isn't?  This is a 16.5% swing in the 'Red Wall', compared to a 15.9% swing per R&W's recent GB poll. As with previous polls of this sample of constituencies, it's not doing anything different to the country as a whole. And it may be that the Conservatives benefit from an incumbency bonus with these seats, all of which they gained in 2019. I don't know if the article below ('forget the Red wall') has been shared on UKPR2 before, but the authors state that based on the YouGov MRP around the turn of the year, the 'new marginals' will be a very different set of seats, more widely spread across England, but with a concentration in the South-East and West Midlands. www.politico.eu/article/red-wall-united-kingdom-general-election-poll-labour-conservatives-liberal-democrats/ |
|
|
|
Post by James E on Feb 28, 2024 15:59:08 GMT
James - thanks for the above. Not that an answer is available but I wonder re the below how those 20% will vote if only Lab or Tory available - in other words if tactically in a straight Tory/Lab seat. ''YG show that RefUK are typically taking 20% of the decided Con2019 voters'' I suspect many would be either RefUK to register support even though they know no chance of winning or WV. Question for James if he could obliged - 20% of decided 2019 voters is how much of the 2019 vote share? I could work out if I knew the DK/WV from Con 2019 of course but I don't have access. Thanks for this, @jimjam. You have caused me to re-check the figures which I had prepared rather hastily, and I now see that I had in fact mixed up the 'decided' and 'undecided' figures. So an important point of my post an hour ago was plain wrong - huge apologies for this. Labour's 'lost 2%' are not voters returning to the Tories. Reform UK are now taking 25% of the 'decided' Con2019 voters, per YouGov. And for ALL Con 2019 voters, the comparative losses are: Jan/Feb2023 To RefUK 9% To Labour 11.5% Don't Knows 24% Jan/Feb 2024 To Refuk 17% To Labour 12% Don't Knows 23% So the Don't Knows and switchers to Labour have been almost static, and the losses to Reform are up by 8% (of their 45% 2019 share). This means that the Tories' overall vote loss ( of 3 points per YG) is less than the VI which has gone from them to Reform UK in the past year (of near 4 points). Meanwhile, Labour are still a couple of points lower than 12 months ago : this looks like it is the result of lower retention of 2019 voters (down by 6% on average), largely to not voting . |
|
|
|
Post by James E on Feb 28, 2024 14:17:08 GMT
While Labour's polling lead is currently very much the same as it was 12 months ago, this disguises the fact that there have been some significant movements. Taking the polling averages of YouGov and Opinium from Jan & Feb 2024, and comparing these to the same months of 2023, we have:
Per YouGov
Lab 44.5% (-2.5)
Con 21.5% (-3)
Ref 11% (+4.5)
Opinium
Lab 42% (-2)
Con 26% (-2)
Ref 10% (+3.5)
It would be tempting to suggest that Reform UK have been drawing support equally from Con and Lab, based on a simple matching of the pluses and minuses. However, the cross-breaks of 2019 voters tell a different story: YG show that RefUK are now typically taking 20% of the decided Con2019 voters, and just 1% of 2019Lab (and LD). Their rise in the past 12 months has been entirely at the Tories expense, with Con2019 to Ref switching going up by 7 or 8 points. [Meanwhile, the Conservatives have been quietly taking back a part of the support they were losing to Labour 12 months ago. In rough terms, Con to Lab switching has fallen from around 18% in Jan/Feb 2023 per YouGov to 14% on average so far in 2024. And that 4% of the 45% Con2019 vote would explain Labour's loss of 2 points in the past 12 months. ( or not; see post below) ]
EDIT - My apologies: I've re-checked this, and found I was not making the correct like-for-like comparison. It is true that the Tories' losses in the past year have all been to Reform UK, but the level of Con to Lab switching is the same now as in Jan & Feb 2023. Labour's loss of 2 points must be due to something else.
|
|
|
|
Post by James E on Feb 27, 2024 11:19:35 GMT
Compared to GE2019, that's: Lab 52% (+4) Con 17% (-15) LD 10% (-5) Green 10% (+7) Ref 10% (+8) Very similar result to YG's London cross breaks which average 54/17 in 8 2024 polls. Swing of 9.5% is (as always) lower than the 17% which YouGov show for all of GB. And even on these record-low figures for the Tories, they are retaining more of their 2019 vote share in London than for GB as a whole : UNS would put them below 10% in London. |
|
|
|
Post by James E on Feb 26, 2024 12:03:47 GMT
As some might recall the liberal democrats won Chesham and Amersham from the Tories in 2021. The swing was massive over 30% of in a seat the Tories had previously held with a 30% margin and nearly 60% of the votes cast. Election maps now shows this as a lib dem" gain" as measured against 2019 in a general election, Chesham is 51st on the liberal democrats target seat for gain list. The Chesham & Amersham by-election was an extraordinary result, as it took place while GB polling averaged Con 43%, Lab 34%, LD 7%. So the Tories are now 18 points lower than they were when defending this seat in June 2021, and the LDs 2-3 points higher. Because of that, I would guess that the Conservatives may fare little better at a General Election than the 35% of the vote they managed at the by-election, with the LDs being the beneficiaries. Beyond the 9-or-so close Con/LD marginals from 2019, it looks one of the easiest LD gains for the next election. |
|
|
|
Post by James E on Feb 24, 2024 10:35:38 GMT
There are no more details of this published by either YouGov or the Fabians, but there was another poll of this 'sea wall' done in November 2022, with very different results. The sample then was only 631 people, and the result was a swing of 14%, at a time when YouGov's GB polls showed Con 25 Lab 47, as opposed to Con 22, Lab 46 now. So the small sample then showed a swing 3 points lower than their GB polls, whereas this one is 9 points above the current average. fabians.org.uk/wp-content/uploads/2022/12/V.08-Sea-Wall-ed-IW-3a-1.pdfThese Jan 2024 figures show a 27% swing from the Con 51%, Lab 29% in these constituencies in 2019, compared to 18% in YouGov's recent polls. Even if you apply a proportionate rather than uniform swing to the GE2109 figures, the expected result would be Con 25, Lab 45, as opposed to 21/51 in these figures. So this could be taken as yet more evidence of the Conservatives now doing especially badly where they fared well in 2019. The recent Survation poll of rural constituencies showed the same. However, if the sample is again only about 600, these figures need to be taken with a pinch of salt. This is about the same size as YouGov's usual South of England cross-break in their weekly GB polls of 2,000. One recent YouGov South of England sample showed a very similar 26% swing (Lab 43, Con 23 on 24 Jan), while another showed 'just' a 17% swing (Lab 34, Con 31 on 15 Feb). |
|
|
|
Post by James E on Feb 23, 2024 16:03:34 GMT
They've had a re-think
Westminster Voting Intention:
LAB: 44% (-2)
CON: 25% (-1)
RFM: 10% (+2)
LDM: 9% (=)
GRN: 6% (=)
SNP: 3% (+1)
Via
@wethinkpolling
, 22-23 Feb.
Changes w/ 15-16 Feb.
|
|
|
|
Post by James E on Feb 22, 2024 21:53:31 GMT
|
|
|
|
Post by James E on Feb 21, 2024 17:41:27 GMT
@mark "Some, such as the sadly departed "The Other Howard" on the original UKPR, recognised that there would be an economic hit, but, that Brexit was worth such a hit. Others thought we would be no worse off." The late ToH revised his views heavily on this during the course of the negotiations. At the time when Theresa May triggered Article 50 (so late March 2017) he stated that he expected an economic hit while we were leaving, followed by a swift rise in investment and economic growth once we had actually left. And he also expected at that time that we would leave with no deal. He later adjusted his views, stating (many times) that he did not expect the benefits to come until about 2030; it is those later comments that most here now recall. I am sorry that I do not have the actual quotes from the time, but noted his comments, as I thought then that they were likely to be proved wrong. If anyone can access the old UKPR1 comments, you would need to look at those from the last few days of March. And on the same subject - the proportion of Leavers expecting negative economic consequences from Leaving in 2016 was far lower than the "15%" mentioned a few times above. YouGov asked voters several times before the 2016 Referendum : "Do you think Britain would be economically better or worse off if we left the European Union, or would it make no difference?" The proportion of Leavers answering 'worse off' was just 3%/4% per the two polls linked below from 19th and 22nd June 2016 (and the same in 2 earlier YG polls of 6 June and 31 May). 'Better off' was the view of 45%/48% of leavers, with 38%/43% saying that they thought it would make no difference. d25d2506sfb94s.cloudfront.net/cumulus_uploads/document/atmwrgevvj/TimesResults_160622_EVEOFPOLL.pdfd25d2506sfb94s.cloudfront.net/cumulus_uploads/document/it82go26iz/TimesResults_160620_EUReferendum_W.pdf |
|
|
|
Post by James E on Feb 19, 2024 19:10:18 GMT
I think this question from the latest YouGov (14-15Feb) has gone largely unnoticed, but it's one poll where Rishi Sunak is ahead. "Which of the following do you think would make the best Prime Minister"? Rishi Sunak 32% Boris Johnson 17% Not sure 46% Refused 5% So Sunak leads by 15 points over Johnson, and this rises to 17 points among those with a current Con VI (50/33). However, Johnson does have a 2% lead with those who voted Conservative in 2019, taking 33% to Sunak's 31% and another 31% of Don't Knows. With the crucial group of those who intend to vote but 'Don't Know' who for, Sunak leads by 26% to 16%, with all the rest not sure or refusing the question. The group which most strongly prefer Sunak as PM over Johnson are 2019 LibDem voters, 54% of whom choose Sunak. However, only 6% of the same sample (of 173) intend to vote for the Conservatives. ygo-assets-websites-editorial-emea.yougov.net/documents/TheTimes_VI_AdHoc_Results_240215_W.pdf |
|
