|
|
Post by James E on Apr 27, 2023 16:20:19 GMT
Re UKPR's Uniform National Swing Model.
I agree, it is very amateurish.
If we do see a large Con>Lab swing, I would expect similar effects to GE1997, with smaller swings in the Labour-held seats and larger ones in the marginals and longer range targets. In 1997, UNS from Con>Lab was exactly 10%, but in the target seats it was in the range of 9% to 18%. A swing of around 13% (give or take a few points) was the norm where Labour were attacking.
|
|
|
|
Post by James E on Apr 25, 2023 18:41:30 GMT
Re the 'Blue Wall':
Today's poll is still a 15.5% Con>Lab swing there, so as usual it's still somewhat higher than other GE polling (a lab 14-15% lead is a 13% swing).
Changes from GE2019
Lab 34% (+13)
Con 32% (-17)
LD 24% (-3)
Taking the very typically 'Blue Wall' constituency of Woking, these movements would turn it into almost a 3-way tie. (Con 30.9, Lab 29.4, LD 27.8 ). Dominic Raab's Esher and Walton seat, which is another of the 42 polled, would be LD 42%, Con31%, Lab 17%.
|
|
|
|
Post by James E on Apr 21, 2023 19:57:12 GMT
|
|
|
|
Post by James E on Apr 21, 2023 18:47:11 GMT
Not sure if we'll get a People's Polling but summary of today's polling updates: YG: LAB 43% (-1), CON 28% (+2). LAB lead 15% (-3)Omnisis: LAB 47% (-1), CON 27% (+2). LAB lead 20% (-3)Tecne: LAB 44% (-1), CON 31% (+1). LAB lead 13% (-2)Savanta: LAB 45% (uc), CON 31% (+2). LAB lead 14% (-2)
Average being a further narrowing of 2.5%. Obviously still a long way to go but as well as CON feeling more upbeat then the "Red on Red" attacks might increase as the shine comes off Starmer-LAB and the LW-LAB seek revenge. 'Popcorn' at the ready  An average Labour lead of 15.5%, equating to a Labour majority of 158 on the new boundaries. Labour leads continue to be well in excess of those pre-Truss. Another way of looking at these polls is to compare to the average from the first 3 months of 2023, when the Lab lead was around 20 points. the Tory recovery is a net 4 points - or a 2% swing-back. YG: LAB 43% (-1), CON 28% (+2). LAB lead 15% - compares to average of 22% in 14 Jan-Mar YG polls : so -7% Omnisis: LAB 47% (-1), CON 27% (+2). LAB lead 20% - compares to average 22% in 13 Jan-Mar Omnisis polls: so -2% Techne: LAB 44% (-1), CON 31% (+1). LAB lead 13% - compares to average 18% in 13 Jan-Mar Techne polls : so - 5% Savanta: LAB 45% (uc), CON 31% (+2). LAB lead 14% - compares to average 16% in 10 Jan-Mar Savanta polls : so -2% |
|
|
|
Post by James E on Apr 21, 2023 14:27:45 GMT
Article by Rob Ford on potentials for tactical voting: swingometer.substack.com/p/tactical-voting-how-much-can-the?r=8jnjk&utm_campaign=post&utm_medium=webHe's working off UNS and a minority (just) Labour government (pre tactical voting and using a 10% swing UNS) and I was a little surprised how little, OK 40 seats if you took the extreme end of cleaning up the 3rd party vote, that there were. He also makes the point that the higher Lab lead to begin with (ie lower Con percentage) the better the tactical voting scenario becomes. There's always a more or less fixed floor for a 3rd party vote, the Labour floor higher than the LD floor IMHO so some seats may come up against that and it might take a couple of elections for voters to appreciate how close things are but I found it interesting that he has put a range on the impact of possible tactical voting. It's an interesting enough article, but I'd query his use of UNS when there are very good reasons to expect a proportionate rather than uniform swing. He also states that "But this also reflects a real feature of the current polling landscape - right now, the big change since 2019 is a large swing from Conservative to Labour, with the Lib Dems treading water in national polls." This really isn't so, as the LDs actually have a lower 2019 vote retention than the Tories, but are picking up a few votes from other parties so as to be on 10% in the polls, as opposed to 11.8% in 2019. Looking at the cross-breaks, they appear to be losing betwenn 25-40% of their 2019 voters to Labour, with 3% of 2019Lab going the other way. This needs to be taken into account, particularly for seats where both parties may be in contnention. |
|
|
|
Post by James E on Apr 20, 2023 20:15:28 GMT
shevii re R&W Wales polling It's a lower lead for Labour in Wales than the last couple of Welsh YouGovs with 29 and 34 point leads. However, it is much in line with the the swing in R&W's most recent GB polls. In fact, most polls (and cross-breaks) of Wales show a lower swing than for England. On this result, the Tories would hold just one seat in Wales (Montgomeryshire). These figures imply a swing to Labour from the Tories of 7.5% in Wales. On the basis of UNS the Tories would retain five seats - Montgomery - Monmouth - Clywd West - Carmarthen West & Pembrokeshire South - and Brecon & Radnorshire.. Yes, you're right that the swing is only 7.5%, and as such would leave those seats as Tory on UNS. This is actually a lot lower than the 13% swing that current GB polls are showing (ie Lab 14-15 points ahead). However, it might be worth considering how 'ratio swing' would apply in Wales, with the Conservatives losing one-third of their vote rather than the simple 12 points that a UNS model shows. And as a precedent for this, it's worth remembering just how many seats Labour won in 1997 on swings far larger than the 10% UNS. |
|
|
|
Post by James E on Apr 20, 2023 19:49:05 GMT
Red Wall / Blue Wall shevii asked a couple of days ago about the differing patterns in R&W's so-called Red Wall and Blue Wall polling, which has led me to take a look at the most recent polls for each of these. This caused some debate as to why Labour appear to be gaining more in the 'Blue Wall' and what is driving this. But first, here are the averaged headline figures taking the most recent 4 polls for each (since late Feb) Red Wall Lab 49% (+11) Con31% (-16) LD 6% (+1) Con>Lab swing 13.5% Blue Wall Lab 38% (+17) Con 33% (-17) LD 20% (-7) Con>Lab swing 17% I should reiterate that the two 'Walls' are not really Northern and Southern equivalents. The 'Red Wall' conists of 40 Tory Northern, Midland and Welsh seats which were Labour seats in 2015 and 2017. 5 are in Wales, 12 in the Midlands of England and 23 in the North. Meanwhile, the 'Blue Wall' is made up of 42 seats, 10 in London, and 32 in the Southern regions. All Blue wall seats have a relatively high number of graduates, at least a 42.5% Remain vote in 2016, and a Tory majority of less than 10/15K over Lab or the LDs. In all but 15, the LDs came second to the Tories in 2019. While I don't have precise figures, the 'Red Wall' seats must have been around 40/60 Leave in 2016, and the Blue Wall maybe around 55/45 Remain. The really remarkable thing is the similarity of the 2019-vote cross breaks for these two types of seat. These are as follows. Of 2019Con voters 68% stick with the Tories in the Red Wall and 66% in the Blue Wall. Meanwhie 19% switch to Labour in the Red Wall and 20% in the Blue Wall. They are more likely to switch to RefUK in the Red wall than the Blue by 8% to 6%, while it's the other way round with Con>LD switchers by 5% to 3%. 2019Lab voters sticking with Labour are 90% in the Red Wall and 89% in the Blue Wall, while those going to the Tories are 2% (Red) and 3% (Blue). Lab lose 3% to the LDs and 1% to Refuk in each. 2019LibDems stick with the LDs at 64% in the Red Wall and 66% in the Blue Wall. Switchers to Labour are 27% in the Red Wall and 26% in the Blue Wall. They lose 5% to the Tories in the Red wall and 3% in the Blue Wall. So the simple answer is that the pattern per past vote appears to be almost exactly the same in the two samples. It is the pattern of 2019 votes which provides the higher swing in the Southern sample, as the LDs took 27% of the vote in the Blue Wall compared to just 5-6% in the Red Wall. So in a sense, the differing headline results are a demonstration of how proportionte (or 'ratio') swing works, as opposed to 'Uniform National Swing'. The only caveat I'd put on this is that R&W do not have a panel, and all of their polls show a much higher voter retention than panel-based pollsters such as Opinium and YouGov. |
|
|
|
Post by James E on Apr 20, 2023 16:39:15 GMT
shevii re R&W Wales polling It's a lower lead for Labour in Wales than the last couple of Welsh YouGovs with 29 and 34 point leads. However, it is much in line with the the swing in R&W's most recent GB polls. In fact, most polls (and cross-breaks) of Wales show a lower swing than for England. On this result, the Tories would hold just one seat in Wales (Montgomeryshire).
|
|
|
|
Post by James E on Apr 19, 2023 19:39:13 GMT
Brexit is a wholly Tory misadventure. At the 2016 referendum, the Labour vote went roughly 65/35 pro-EU. That's only because a lot of the Labour vote had already defected to UKIP. While UKIP's 13% vote share in GE2015 was predominantly working class, very little of it came from those who had voted Labour in 2010. Looking at the final 5 YouGov polls before the election, these show an average of 6% of Lab2010* voters voted for UKIP. That's less than 2% of those who voted. There were far higher proportions of Con2010 voters (15%) and LibDem2010 voters (10%) who switched to UKIP. en.wikipedia.org/wiki/Opinion_polling_for_the_2015_United_Kingdom_general_election*(6%, 7%, 7%, 3%, 6%) |
|
|
|
Post by James E on Apr 18, 2023 21:57:03 GMT
Meanwhile, their 'Blue Wall' polls show a higher Con>Lab swing than the GB average - normally by around 2-3 points. This is remarkable, because 25 of the 42 Southern seats where they do these polls have the LibDems in 2nd place and Labour 3rd. I do wonder whether since the pandemic London has just cleared out huge numbers of under 40s, who overwhelmingly vote Labour (or anti-Tory). I know plenty of people my age who have moved down here or to surrounding towns and villages, most of which are Tory seats. This might also explain the lower swing towards Labour in London.
Addendum to above: after reading shevii 's post, it could also be the fact that 4 more year's worth of younger people in Blue Wall seats have faced not being able to buy a property and all the other struggles which their forbearers wouldn't have had so bad, so are going off the Tories whereas their forbearers would have stayed put. @sotonsaint The large differential swings beteen London and the South of England regions are far too big to be explained by people moving out of London since 2019. To take YouGov's cross-break average for the past 2 months - it's 11% Con>Lab in London and 20% in the rest of the South. When I last checked R&W's regional figures a few weeks ago, it was 6% in London and around 23% in the rest of the South. And if we're talking about R&W's so-called 'Blue Wall' sample of 42 Southern constituencies - it must be remembered that 10 of these are actually in London. This may be a reason for R&W's 'Blue' Wall' Con>Lab swing being lower than their regional cross-breaks for the South. However, I think it's more likely that the lower London swing is due almost entirely to the large number of safe Labour seats in the capital. |
|
|
|
Post by James E on Apr 18, 2023 17:30:13 GMT
Red Wall swing 12.5% since GE - broadly in line with latest R&W national poll. Do you know if this is also in line with the "Blue Wall" polls they've been doing? I'm guessing if similar to national polling it might be but there are places which are not a wall at all. I would expect differences because I've have expected the seats they chose to have a lot of first time Tory voters, whereas Blue Wall would have demographic shifts. The pattern I have noted is that the 'Red Wall' polls are typically showing a swing 1-2 points lower than R&W's full GB polling. This latest one is an exception to that, but R&W's most recent GB poll with the Tories on 32% looks untypical. Meanwhile, their 'Blue Wall' polls show a higher Con>Lab swing than the GB average - normally by around 2-3 points. This is remarkable, because 25 of the 42 Southern seats where they do these polls have the LibDems in 2nd place and Labour 3rd. One factor is the LD vote. They seem to be holding steady on a small share (6%) in the 'Red Wall' polls, but are down from 27% to around 21% in the 'Blue Wall'. This also explains why Labour are picking up more votes there. I think the Con2019 cross-breaks are fairly similar, although in the 'Red Wall' they are losing more to RefUK. |
|
|
|
Post by James E on Apr 18, 2023 9:39:18 GMT
|
|
|
|
Post by James E on Apr 6, 2023 15:53:03 GMT
Electoral Calculus are basing their figures on current poling (with Labour nearly 20 points ahead), proportionate swings, and varying regional figures. Because of this, they have Lab up 16 points and the Tories down by 19 in Reigate, while for GB as a whole, they have Lab +13 and Con -17. Electoral Calculus's regional figures for the South-East of England are Lab 37.9% Con 33.5% - an 18.1% Con>Lab swing (link below). This does not look unreasonable, based on current polling. Redfield and Wilton's recent SE England cross-breaks average Lab 43%, Con 33% - so a 21% Con>Lab swing. No other company gives figures for individual regions, but I've noted a couple of pollsters' combined South of England ( SE+E+SW) figures. The last 9 YouGovs South England cross-breaks average Lab 41% Con 31% - so a 20.5% swing on GE2019 (they also show the LDs down by 3 points) And the last 6 Opiniums (with their DKs adjustment) average Lab 39%, Con 32% for the combined South - a 19.5% swing. www.electoralcalculus.co.uk/area_seast.html |
|
|
|
Post by James E on Apr 6, 2023 14:34:01 GMT
Worth noting that 42% of households in London do not own a car (or van). This is 31% for outer London, so even there it is roughly double the 16% with no car in the surrounding regions of the South of England. www.ons.gov.uk/peoplepopulationandcommunity/housing/bulletins/housingenglandandwales/census2021#car-or-van-availabilityOne other bit of data, relevant to the London Mayor: the May 2021 mayoral election took place at a time when the Tories were 10-12% ahead in GB polls, and London Westminster polls were averaging Lab 49%, Con 33%. Both of these were broadly the same as the position as of GE2019. The May 2021 first round result of Khan (Lab) 40%, Bailey (Con) 35% needs to be seen in that context. Khan's vote was significantly below the London Labour VI, and the timing could hardly have been better for the Tories, yet he still won fairly comfortably (55-45 in the run-off). |
|
|
|
Post by James E on Apr 6, 2023 9:34:51 GMT
Good to finally get a London poll. It's been a year since YouGov's last one.
Changes from GE2019 are:
Lab 58% (+10)
Con 18% (-14)
LD 9% (-6)
Green 7% (+4)
Ref 6% (+4)
These figures are very similar to the recent YouGov cross-break averages for London. For their 9 polls in Feb and Mar, these are Lab 56%, Con 18%, LD 10%. Other pollsters have shown much lower swings in London (e.g 5%) than the rest of GB, but this may be as a result of locally unrepresentative sampling. It is very noticable that YG's samples are far more consistent than R&W or Deltapoll, for example.
London's Con>Lab swing of 12% compares to 16-17% for GB as a whole, as YG's most recent poll was a 20% lead, and their recent average is around 22%.
|
|
|
|
Post by James E on Apr 5, 2023 17:04:50 GMT
The latest R&W "Red Wall" poll is a 14-point swing on GE2019, with Labour up 11 points (from 38%)and the Tories down by 17 (from 47%).
This is one point lower than GB polls are typically showing at the moment as an 18% Lab lead is a 15% swing. However, the 14% is right in line with R&W's own regional cross-breaks for the three English Northern 'regions', and with their most recent full GB poll of 2 April.
|
|
|
|
Post by James E on Apr 2, 2023 10:30:07 GMT
YouGov's long-running Brexit tracker question* has been used 9 times so far this year, with the latest poll of 29 March showing a 32/55 (or 37% to 63%) majority thinking the decision to leave is Wrong. Two interesting developments so far in 2023 - both involving the cohorts which until recent recently showed pro-Brexit majorities: Firstly, C2DE voters are now consistently saying that the decision was 'wrong' - all of of 8 most recent YG polls since early January show this, typically by about a 6% majority. In 2016 C2DE voters were 65/35 for Leave. Until around 6 months ago, there was still a pro-brexit majority in this sample. And even the 65+ cohort now seems to be having doubts. Last year, the figures for this group were working out as around 64/36 for 'Right to leave', whereas now it's a little under 60/40. meanwhile, all younger age groups have shifted against Brexit by around 15-20 points. * "In hindsight, do you think Britain was right or wrong to vote to leave the EU? " - YouGov have used this in 230 polls since 2016. whatukthinks.org/eu/questions/in-highsight-do-you-think-britain-was-right-or-wrong-to-vote-to-leave-the-eu/?pollster%5B%5D=yougov |
|
|
|
Post by James E on Apr 1, 2023 22:14:40 GMT
9 -mile queues at the port of Dover at the start of the Easter season, and the local Tory MP blames the French. Phil Moorhouse at his best... www.youtube.com/watch?v=fHZIpeQ58I8 |
|
|
|
Post by James E on Apr 1, 2023 20:22:22 GMT
LATEST @opiniumresearch Public opinion stays steady as Labour’s lead holds at 15 points. Con 29% (nc) Lab 44% (nc) ..etc James - you’re a ....[etc] Does that mean that as we get closer to a GE this ‘prediction’ becomes more closely aligned to what people are saying they would vote when the company polled them? In other words, the closer we get to the GE, the less awry these figures should be? Thanks, Dave. All I really do is literally read the detailed figures of some of the more reliable pollsters...and do a bit of basic analysis. Yes, it is likely that as we approach a GE, Opinium will converge with other pollsters. In fact, they already have converged quite a lot over the past 4 months, from around a 10-point gap when Truss was PM, as the Tories have become a bit less unpopular. It all depends on the relative levels of "Don't Know" responses from 2019 Con & Lab voters. And it's significant that Opinium's figures have not shifted in the past couple of weeks when others have. This reflects the fact that the Tories have been recovering support from some of their 2019 voters who were answering 'Don't know' a month or two ago. I wouldn't call Opinium's figures a 'prediction'. What they do is to exclude the 'Don't knows' (and WNV) and calculate their headline figures on the basis of those who make a positive choice. This means that those 2019Con voters who state a voting intention are upweighted. The resulting figures are surely likely to be closer to the real result of a GE than for those who treat all those who say 'Don't know' as non-voters. However, if you look at historic precedents of Tory governments recovering before a General Election, this needs to be taken into account. So while we might expect around a 10-point net recovery from YouGov's figures, it would only be around 4-5 points if you start from the way Opinium calculate their headline figures. As I've mentioned before, there isn't really much difference between YouGov and Opinium's polling findings if you look at their tables for those voters who state a positive VI (so excluding the DKs and WNVs). Opinium have 2019Con voters dividing as 66/17/3 (%) between Con/Lab/LD, compared to 61/16/5 with YouGov. For 2019 Lab voters, it's 1/89/3 with Opinium and 0/88/4 with YouGov. And for 2019LibDems, it's 6/34/51 with Opinium, and 8/41/44 with YouGov. There are other factors affecting the various pollsters: the largest of these is 'false recall' for those who don't have panels and recorded past votes, as YouGov and Opinium both do. The other significant ones are prompting (or not) with party names (which boosts Green and RefUK for those who name them) and the ordering of polling questions - it seems to make a difference whether you ask about VI first or likelyhood to vote first. |
|
|
|
Post by James E on Apr 1, 2023 19:21:27 GMT
LATEST
@opiniumresearch
Public opinion stays steady as Labour’s lead holds at 15 points.
Con 29% (nc)
Lab 44% (nc)
Lib Dems 9% (+1)
Green 5% (-1)
Reform UK 7% (nc)
@opiniumresearch
Rishi Sunak's net approval takes a fall compared to our last poll.
28% approve and 43% disapprove, giving him a net rating of -15%. This is down 7 points from Opinium's previous poll and essentially reverses the gains the PM saw in that poll.
|
|
|
|
Post by James E on Mar 31, 2023 16:15:24 GMT
Many thanks for that, shevii YouGov normally do their fieldwork on a Tuesday and Wednesday, so this is most likely 28-29 March. Their last poll with a 26 point lead was a particularly high one, whereas this 20 point lead is just a couple of points below YG's average for 2023 so far of a 22 point Lab lead. |
|
|
|
Post by James E on Mar 31, 2023 14:01:33 GMT
James, ''The last 8 YouGovs show 2,2,3,2,2,2,2,2 for 2019Con>Lab, while Opinium's last 4 polls are 2,1,2,2. - Average 2% The last 8 YouGovs show 4,4,11,4,6,5,7,6 for for 2019Lab>Green, while Opinium's last 4polls are 4,4,6,5. - Average 5.5%'' I think the 2019Con>Lab should be Green? 2% of 44% is .88% and 5.5% of 33% is 1.8% approx. Perhaps not 3 to 1 Lab-Green v Tory-Green and just over 2:1; maybe the rogue 11% skewed to appear 3:1 using YG crossbreaks? Indeed - and thanks for pointing out my error. As you say, it is more like 2 2019Lab>Green for 1 2019Con>Green. (Although the proportion of 2019Lab>Green is 3 times their 2019Con equivalents) |
|
|
|
Post by James E on Mar 31, 2023 10:54:34 GMT
@jimjam, @trevor For what it's worth, the last 4 YouGovs show Con2019>Green, as 2,2,3 and 2% The comparative figures for Lab2019>Green are 4,4,11 and 4%. So it looks like the small Con>Green movement that JJ has mentioned in Local Election VI does apply to GE voting intentions, too - albeit that the Greens are getting maybe 3 Lab voters for 1 Tory. and rounding but CON'19->Green VI was 1% versus LAB'19->Green VI of 4% on last YG. I did already mention that Green VI had dropped and I expect turnout to be low but 'stay at home' (ie a DK that abstains) is clearly not a vote for someone else - just a rounding up exercise for headline VI. I have checked all of YouGov and Opinium's recent figures. I tend to trust these two more than other pollsters, as they have panel-based samples, and a strong track record. And I am using the figures after excluding DKs in all cases. (It should not matter whether one chooses the full data or the figures after excluding DKs - so long as a consistent approach is taken). The last 8 YouGovs show 2,2,3,2,2,2,2,2 for 2019Con>Green, while Opinium's last 4 polls are 2,1,2,2. - Average 2% The last 8 YouGovs show 4,4,11,4,6,5,7,6 for for 2019Lab>Green, while Opinium's last 4polls are 4,4,6,5. - Average 5.5% So the 4 most recent YouGovs I quoted earlier are right in line with this - and certainly not "selective". |
|
|
|
Post by James E on Mar 31, 2023 9:10:26 GMT
@jimjam, @trevor
For what it's worth, the last 4 YouGovs show Con2019>Green, as 2,2,3 and 2%
The comparative figures for Lab2019>Green are 4,4,11 and 4%.
So it looks like the small Con>Green movement that JJ has mentioned in Local Election VI does apply to GE voting intentions, too - albeit that the Greens are getting maybe 3 Lab voters for 1 Tory.
|
|
|
|
Post by James E on Mar 30, 2023 18:12:02 GMT
How well would Alba do here? They will surely stand - if only to make life more difficult for SNP. A bit of an exaggeration there, surely? The LDs are polling a steady (if underwhelming) 6% in Scottish Westminster polls, while Alba's VI is also consistent, ranging between 1% to 0% |
|
|
|
Post by James E on Mar 30, 2023 17:21:04 GMT
What a polling based approach can't allow for is tactical voting in a one-off by-election context. For example, it is easy to imagine the Conservative vote collapsing completely in favour of Labour if Tory voters see this a free opportunity to give the SNP a kicking, knowing there is no effect on the Westminster position. I am sure both the SNP and Labour would throw everything they have at this and I am happy to confess I have no idea what the outcome would be. Labour really ought to win a seat which it gained in 2017 under Corbyn. Dire for Starmer if they fail. Current Scottish Westminster polling (SNP around 39%, Lab around 31%) is very much in line with the overall Scottish 2017 result, when SNP finished about 8 points ahead of SLab - another indication that if there is a Rutherglen by Election, it will be very close. Labour has of course made much greater progress in England than in Scotland under Starmer - that much must be obvious from even a brief perusal of polls from the past year or two. |
|
|
|
Post by James E on Mar 30, 2023 17:10:59 GMT
oldnat "SCon vote going to SLab should be enough for SLab to win, even if the SNP vote remained stable." Surely the Scon vote is unlikely to move over 'en masse'? YouGov's recent Scottish cross-breaks for Con to Lab (17%) and LD to Lab (40%) are effectively identical to their figures for GB as a whole - the last 4 GB YouGovs show 2019Con>Lab as 16% and 2019LD>Lab as 40%. [to Add - 19:22hrs] The most recent Scottish Westminster by election, Airdrie and Shotts, May 2021, saw no collapse in the Conservative share, despite Lab being the clear challengers to the SNP. They were down 4.7 points from 17.6% to 12.9%. At the time, the Tories were polling around 22% in Scottish Westminster VI - so 3 points down on GE2019. So maybe a couple of points of by-election squeeze on the 3rd-placed Tory vote, but no more than that. en.wikipedia.org/wiki/2021_Airdrie_and_Shotts_by-election
|
|
|
|
Post by James E on Mar 30, 2023 15:51:04 GMT
Savanta MRP prediction from December for Rutherglen together with lots of caveats: I have had a go at modelling Rutherglen and Hamilton West based on averaged cross-breaks from the two most recent YouGovs. I have ignored the very small movements to and from the SNP with the Cons and LDs. SNP lose 14% to Lab and 7% to SGP, but gain 11% of 2019Lab. Lab lose 11% to SNP, gain 14% of SNP2019. Lab gain 17% of Con2019. Lab gain 40% of LD2019. Lab lose combined 5% of their 2019 vote to LD or Con. This works out as: Lab 40.0% (+5.5) SNP 38.8% (-5.2) However, this assumes that SGP stand. The 3% which YG's figures suggest they would take from the SNP could be crucial. As for Savanta's MRP in December: The SNP's lead has fallen from around 13 points then to 9 now. If you adjust for that, it too would suggest a knife-edge result. |
|
|
|
Post by James E on Mar 29, 2023 17:44:36 GMT
Tory MPs revolting, not a surprise Surely any village with an RAF base would be a "noisy village"?  |
|
|
|
Post by James E on Mar 28, 2023 17:16:47 GMT
Blue Wall poll is back to normal - the last one was somewhat of an outlier in the Tories' direction. In the 6 'Blue Wall' polls they've done so far in 2023, the average is Con 32%, Lab 40%, LD 20%, Grn 4%, Ref 4%.
Comparing the figures to GE2019, this latest one is:
Lab 39% (+18)
Con 31% (-19)
LD 21% (-6)
If reflected in a General Election, these figures would (I think) see Labour gaining nearly all the target seats where they are 2nd, and a number where they are third. For example - Woking, which is one of the seats polled and is very close to the overall 2019 result forR&W's 'Blue Wall' sample, becomes Lab 35% (+18), Con 31% (-19), LD 25% (-6).
|
|
