|
|
Post by James E on Jun 17, 2024 13:44:07 GMT
Is it just me, but it does look like the Conservatives have given up on the national campaign and must be all focusing on defending their own seats?
I’ve been posting data about that, and the Telegraph graphics showing where the campaigning is happening, which supports your view. It also showed that Refuk are prioritising Labour seats for campaigning. If that is true, it is really bizarre. The last 4 YouGovs show RefUK taking an average of 32% of the Con2019 vote, and 3% of the Lab 2019 vote. And on the Leave/Remain axis, they are taking 35% of Leavers and 2% of Remainers. |
|
|
|
Post by James E on Jun 16, 2024 19:01:34 GMT
alecPolling Report's 'default model' is something described as 'incumbent weighted' which produces fewer gains per point of swing then even Uniform National Swing. There have been elections where such a model might have served well, such as 2001 or 2015 (when first time incumbents did especially well), but it is a useless model for 2024.
|
|
|
|
Post by James E on Jun 16, 2024 17:36:52 GMT
Ignore Batty. He’s a known troublemaker. Quite violent too. I need some advice. England defeats usually provide easy revenge targets after the games. If we lose to Germany, for example, parked BMWs near pubs can be vandalised. If we've lost to Greece, Greek owned fish and chips shops can be ransacked. Kebab outlets if the Turks have turned us over, parked Renaults torched if it's the French. Most countries have convenient recipients for our righteous rage, but what the hell do we do with the Serbs? Foreign made cars seem to be the favourite target, so it should be obvious what will to be targeted this time. Saabs. |
|
|
|
Post by James E on Jun 16, 2024 10:59:28 GMT
jib "The SNP are still on course for majority of MPs in Scotland. I failed to understand those who wrote off their chances, and their travails under Sturgeon and Yousaf look behind them now." Survation's MRP figures (SNP 37%, Lab 30%) are out of line with other polling in Scotland. Not one of the past 15 polls shows the SNP ahead and the overall average is for Labour to be 4 points in front. YouGov's cross breaks show Labour an average of around 6-7 points in front, too. It seems to me that where MRPs ask a 'constituency question', the SNP may be picking up some tactical support from the Scottish Greens, so perhaps this would explain some of the difference, but certainly not all of it. Or this may be a house effect of Survation's MRPs, as they have consistently shown higher SNP seats than others such as YouGov. en.wikipedia.org/wiki/Opinion_polling_for_the_2024_United_Kingdom_general_election#ScotlandOf course it is possible that Survation's MRP is entirely correct, and most of the polls listed above are wildly wrong. But I think that the more likely scenario is that this one is generous to the SNP, and the reality is somewhere near to a tie. This makes a accurate prediction of seat totals in Scotland almost impossible.
|
|
|
|
Post by James E on Jun 16, 2024 9:10:47 GMT
shevii "I'm sure other people can provide poll averages but the Wiki page shows a definite decline in Lab and Con since the election was called. There is a tick box below the graph where you can tick a box to get the graph since the election was called:" en.wikipedia.org/wiki/Opinion_polling_for_the_2024_United_Kingdom_general_election The problem with this is that the graph simply reflects the headline figures which are affected by changes in methodology. Some pollsters such as YouGov and MiC have moved to re-weighting the DKs, which has reduced Labour's lead by around 6 and 4 points respectively. Others such as Survation have changed to prompting according to declared candidates. This tends to boost Ref and Green, the latter at the expense of Labour. So if you look at the average Lab VI per the most recent 10 polls, it's now 41.2%, down by 3.9 points from 45.1% in the last 10 polls before the election was called. But by my preferred like-with-like comparison to the longer-term average of the first 20 weeks of 2024, the Labour VI has fallen by far less. Per the most recent polls we have from Opinium, Techne, Savanta, YG, R&W, WeThink, Survation, BMG and MIC, Labour are down by an average of just 1.2 points from where they stood in the pre-election period (from 1 Jan to 22 May). [Incidentally, I've found that Savanta are not the only pollster to show an increased Lab VI. MiC's 'unadjusted' figure is up by 0.5%, too, though their headline figure is down by 1.5%]
|
|
|
|
Post by James E on Jun 15, 2024 22:03:26 GMT
ping I understand that that's a new constituency however the 2019 result showed The Tories on 57% the Liberal democrats on around 19% narrowly ahead of Labour on 17%. What precisely did "putting the voters right on tactical voting" entail? The 2019 result showed the LDs 1% ahead of Labour, but the notional result for the new boundaries of Kenilworth and Southam puts Labour ahead by just over 2%. I would imagine that any Labour canvasser would want to point this out. www.electionpolling.co.uk/constituencies/uk-parliament/kenilworth-and-southamToday's Survation MRP shows the seat as Lab 33.7%, Con 32.5%, LD 14.6%, Ref 11.1%. Also, a few weeks ago I posted details of how seats like this (per the 2019 result) behaved in the 1997 election: that is with the Tories well ahead, but with a narrow gap from the LDs in 2nd to Lab in 3rd. The norm then was for Labour to make a net gain of around 15 points in relation to the LDs, so the outcome above (which would be a net gain of 17 points) looks plausible. |
|
|
|
Post by James E on Jun 15, 2024 20:36:44 GMT
Wiki has the percentages for that Survation MRP as: Lab 40 Con 24 Ref 12 LD 11 Green 6 SNP 4 The dates are 31st May to 12th June so a lot of those figures will be out of date anyway. That would make sense as Survation's last MRP had vote shares of Lab 43, Con 24, and Labour winning 487 seats. Compared to that, Labour now take 31 fewer seats on a swing 1.5% lower. Some people might be surprised to see these figures showing Labour taking so many seats on a lead of 'only' 16 points. For comparison to this, More In Common showed Labour taking 382 seats on a 15-point lead (43/28). Part of the reason is that the lower Tory VI with Survation makes them even more vulnerable to tactical voting. But also, More In Common show the lowest 'seats gained per 1% swing' of any MRP at just under 14. YouGov's MRPs show around 14-15, relatively low due in part to their 'unwinding' adjustment. Those who don't make such adjustments show higher rates of Labour gains. In Survation's case, 18-19 'seats per 1% swing' in this MRP. My view is that the actual result will fall between these two, but probably closer to Survation than to MIC or YouGov. One further detail on Survation's latest MRP: looking at their map, it would be possible to travel the entire length of Britain - from either Lands End or the White Cliffs of Dover to the North coast of Scotland without passing through a single Conservative-held constituency (though you would need to take a rather convoluted route). www.survation.com/mrp-update-first-mrp-since-farages-return/ |
|
|
|
Post by James E on Jun 15, 2024 19:52:08 GMT
JamesE Is your prediction of the polls converging as the election approaches proving right? It seems to me it is.... I don't recall predicting that the polls would converge, although it is normal to get somewhat fewer 'Don't Knows' close to an election. If and when that happens, the size of the adjustment made by pollsters such as Opinium becomes smaller. But as per the figures Neil has provided, this one is still a 6-point reduction to the 44/21 poll which the unadjusted figures would show. And YouGov's most recent 'adjusted' figures reduce the size of Labour's lead over the Tories by 7 points. There are other factors now coming into play, such as pollsters now using actual candidate lists as a prompt. This seems to provide some Lab to LD movement, so would have the effect of reducing Labour's lead if it were not for RefUK eating into the Tories' vote share. And we may also just see pollsters converging by making other methodological changes, but that's a separate issue to the re-allocation of Don't Knows. What I did predict (as did most people) was that the polls would narrow, with a normal-sized Government recovery turning opposition lead of 18-20 points into more like 12-14. That clearly hasn't happened. |
|
|
|
Post by James E on Jun 15, 2024 19:12:32 GMT
Opinium
Lab 40% (-2)
Con 23% (-1)
LD 12% (+2)
Ref 14% (+2)
Green 7% ( 0)
My prediction of 39/22/12/15 wasn't too bad !
[ By my preferred comparison to the pre-election averages for each party by pollster, this one is Lab -1.5, Con -3, LD +2, Ref +3. Very much the norm at the moment]
|
|
|
|
Post by James E on Jun 15, 2024 18:24:08 GMT
Savanta ends a run of 9 polls where Labour were below their pre-election average. In this one, they are up by 2.5 points.
Lab 46% (+2.5)
Con 21% (-5.5)
Ref 13% (+3.5)
LD 11% (+1)
Comparisons to averages for Savanta from the first 20 weeks of 2024.
Adding this one in with the previous nine from regular pollsters, Labour are down by 2 points as an overall average, with 7 of the last 10 showing falls of between 1 and 3.5 points.
|
|
|
|
Post by James E on Jun 15, 2024 14:08:56 GMT
pjw1961 I'd accept that with a best result since 2010 of 9.5% in Braintree in 2019 it wouldn't be particularly high in our target lists. If Labour can move from under 20% to displace not so Cleverly who received nearly 70% of the vote share in 2019 the very best of luck. Hopefully you'll be aided and abetted by the frog faced hate gimp targeting the hard of thinking , it's an area where tactical voting would also have to be very effective. I wouldn't call Braintree a Labour target in any normal circumstances, but it is not quite so hopeless for Labour as 2019 suggests. The area is very Brexity and this maximised the Tory vote in 2019 (in comparison Labour got 27.6% of the vote in 2017). .... I have mentioned this before, but where polling companies provide data by region, the East of England shows significantly higher swings than average. R&W's large poll of 10,000 dated 7-10 June showed a 26% Con to Lab swing in the East in the context of a Lab lead of 27 points. (19% GB swing) MIC's samples (in 5 polls of 2,000) from 23 May onwards show a 22% Con to Lab swing in the context of an average Lab lead of 19 points. (15% GB swing) Labour Together's April poll showed a 23% Con to Lab swing in the East in the context of a Lab lead of 22 points. (17% GB swing) Each of these has a sample of around 600-700 for the region, and in all three the East has the highest swing of any region. The swing needed in Braintree is 24% |
|
|
|
Post by James E on Jun 15, 2024 10:17:14 GMT
We should get an Opinium poll at 8pm this evening, so it will be interesting to see if the figures match to the same dynamics we've seen in other pollsters.
opinium's pre-election averages in 12 polls this year were:
Lab 41.5%
Con 26%
LD 10%
Ref 11%
If the movements are similar to others, we should expect something like this:
Lab 39%
Con 22%
LD 12%
Ref 15%
|
|
|
|
Post by James E on Jun 14, 2024 22:23:59 GMT
Very interesting article. Thanks. I'm interested in vote efficiency and my own figures show votes per seat since 1945: Party Best Worst Average
Con 32,777 1983 58,188 1997 41,182
Lab 25,968 2001 50,837 2019 39,615
Lib 92,583 2001 432,823 Feb 1974 254,596
I interpret this as meaning that Labour have a slightly more efficient vote on average than the Tories, contrary to the popular idea that they tend to pile up votes where they don't need them and also that the Tory vote has been slightly more volatile. And of course we can see why the poor old Libs/LibDems want PR.
A couple of other points - in every single post-war election including the seemingly anomalous 1951 the party with the most efficient votes per seat won, with the sole exception of 2010. Also the much-vaunted 2017 election was Labour's worst performance in votes per seat since the war at that time (until worsened further in 2019). It is logical therefore for the Labour campaigners in particular to be directed away from seats that they are likely to win anyway.
Historically, the 'efficiency' of each party's vote has varied wildly, as the graphs show. Neither has an inherently 'more efficient' vote, although from 2015 onwards it appears that the Conservatives have had the more 'efficient' vote. Looking at the graph about three-quarters of the way down entitled "the Conservatives vanishing FPTP bias" what James Kangasooriam now reckons is that lead required by Labour for an overall majority is about 4% (and about the same for the Tories, for what that's worth). So, the same as I wrote on this forum in May and August of last year.... ukpollingreport2.proboards.com/thread/67/labour-win-outright-majorityukpollingreport2.proboards.com/thread/63/april-2023-lab-con-ldem?page=82 |
|
|
|
Post by James E on Jun 14, 2024 14:44:19 GMT
That certainly seems the case in the south. With many thinking of a Labour victory as a nailed on certainty some are now switching to Libdems to unseat the tories This could be the final nail in the tory coffin for this election Have you encountered evidence of this? Most polls show them picing up a bit - though not universally so. I think we need another week to see whether this is just a manifesto boost effect. I noted yesterday that all five of the most recent polls showed the LDs up on their pre-election vote share (22:29, 13/6/24). The average rise is just over 2 points on a like-with-like comparison. As for the South of England - YouGov's figures after their change to their 'MRP methodology for all polls' from 1 June show a very significant rise for the LDs there. Part of this is driven by their 'constituency question', which appeared to benefit the LDs by 2 points overall when used in earlier MRPs. I don't have a good like-with-like comparison for this, but the LDs' share in the South has risen from an average of 13.5% in 6 polls immediately prior to the election being called to 21% now. Part of that is due to the changed methodology but at least half of that 7.5% rise is almost certainly 'real' movement. Incidentally, the post-change averages for the South in YouGov's 'adjusted' 4 polls since 1 June are: Lab 30% (+7) Con 23% (-32) LD 21% (+4) Ref 18% (new) [to add 15/6/24. I have done a like-for-like comparison of MIC's SE and SW England subsamples. Looks to me like the LD VI is up by 1-2 points in the past couple of polls, compared to their previous average.] |
|
|
|
Post by James E on Jun 14, 2024 13:45:01 GMT
They may not have the numbers, but I'd like to see pollsters split their sample into different flavours of constituency (Con/Lab marginals, Con/Lib marginals, safe Labour etc.) and compare the movement from Lab to Lib. If people are only moving from Lab to Lib in Con/Lib marginals then it would actually be a good thing for Labour. Same goes for movement to Reform. Is this movement geographically spread or is it lumpy? There is a noticeable effect in YouGov's MRPs (including those done in the early part of this year) that their constituency prompt produces a higher LibDem VI at the expense of Labour. This appears to be due to the LDs taking more votes where respondents are reminded that the LDs are in 2nd place in their constituency. This also explains why YouGov are showing the highest LD VI of any pollster. It is this which produces strong LD performances wherever they were in 2nd place in YouGov's seats predictions. I think this may be a bit overdone: a lot of voters simply aren't aware of the position in their own constituency, and previous elections show that there are different dynamics where the LDs are a distant second to the Conservatives, compared to where they are a close second. In 2019, they managed to squeeze Labour a lot where they were seen as running a close race with the Tories. As for Reform - their support levels are very closely aligned to those of the Conservatives. So in YouGov's latest poll, the regional figures show them a little ahead of the Tories across most regions, while for R&W (Con 19, Ref 17) most regions have the Tories slightly ahead. You could make a case for saying that Reform are stronger in the East Midlands and North East, based on R&W, and it also appears that they are relatively weaker in London, Scotland and the North West. Both YouGov and R&W show Reform as leading the Tories in the Midlands - YouGov by 5 points (24/19) and R&W by 2 points (20.5/18.5), so in each case, 4 points better for them than their headline GB figures. This reflects the Conservatives' losses being worst of all in the Midlands; all pollsters show them losing a greater-than-proportionate amount of their 2019 vote share there. |
|
|
|
Post by James E on Jun 13, 2024 22:09:26 GMT
The unadjusted figures show the Conservatives in 4th place in the South of England: Lab 31% LDs 22% Ref 19% Con 17.5% But the adjustment brings them up to equal 3rd: Lab 27% LD 23% Ref 20% Con 20% Even with the benefit of the adjustments, the Tories are still losing a larger proportion of their vote share in their strongest regions, the South and Midlands of England - per YouGov, or all other pollsters I have checked recently. |
|
|
|
Post by James E on Jun 13, 2024 21:29:07 GMT
I'm disliking this reduction in the Labour vote, regardless of what is going on elsewhere. Enough already. Nothing below 37% thanks all the same. I and others such as alec have been commenting on the noticeable fall in LAB VI in several polls over recent days, but it attracted little reaction, apart from some noting that some methodologies had changed. Still MOE, sort of, but there does seem to be a general, even increasing, decline in LAB support. And three weeks still to go... YouGov and Survation have changed methodologies, but most have not. The only really large decline in the Labour VI we have had was from People Polling, but we know little of their methodology. The 5 polls most recent polls we have had - all with fieldwork of 11-13 June show the Labour VI down on its Jan-May level but only by an average of 2 points: 12-13 June Techne 43% (-1) 12-13 June YouGov 37% (-3) - compared to their pre-election MRPs as a like-with-like. 12-13 June R&W 42% (-2) 12-13 June WeThink 43% (-2) 11-12 June BMG 41% (-1) Meanwhile, these 5 polls show an average 2% rise in the LD VI (+1%, 2.5%, 3%, 2%, & 2.5% across the same 5) when compared to the same companies' Jan-May figures. en.wikipedia.org/wiki/Opinion_polling_for_the_2024_United_Kingdom_general_election#National_poll_results |
|
|
|
Post by James E on Jun 13, 2024 21:14:32 GMT
Reform gains on the basis of the YouGov poll and UNS - all held by Con in 2019 Ashfield - Lee Anderson Clacton - Nigel Farage Wellingborough and Rushden - Ben Habib Boston and Skegness - Richard Tice Torridge and Tavistock - Andrew Jackson. So the big 4, and a guy with the same name as a former US President who is Donald Trump's political hero. I imagine that these RefUK gains are derived from applying Electoral Calculus rather than UNS. UNS with those YouGov figures would give these changes to vote shares from 2019: Con -27 Lab +4 Ref +17 So if you apply that to Clacton, we get Con 48%, Lab 20%, Ref 17% Or Ashfield, Lab 28%, Ref 22%, Con 13% |
|
|
|
Post by James E on Jun 13, 2024 20:47:39 GMT
Latest YG poll has two VI tables -
1. Westminster Voting Intention [Headline voting intention projected by YouGov MRP model]
2. Now, thinking specifically about your own constituency, and imagining that these were the political parties standing, which party do you intend to vote for in the July 4th 2024 UK general election?
I looked at the Scots column only (and recalculated the constituency table to exclude DK & WNV), but there are others for Wales and the YG English regions
1. 2. Diff Party
31% 35% (+4) Lab
34% 35% (+1) SNP
9% 7% (-2) Con
12% 11% (-1) LD
7% 6% (-1) REFUK
5% 4% (-1) SGP
3% 3% (=) Other
As has been mentioned upthread, YouGov are now using the same adjustment methodology as they use for their MRPs in their Voting Intention polls. Their actual responses in Scotland gave a 35/35 tie between Lab and SNP, but the re-weighting turns this into a 34/31 lead for the SNP. I would read all the pluses and minuses above the other way round - for example, their adjustment increases the Tory vote share in Scotland from 7% to 9%, but reduces Lab from 35% to 31% Looking at the GB headline figures, their adjustment/re-weighting has reduced Labour's lead over the Tories by 6 points overall. Without adjustment, these figures would be: Lab 40.3% Con 15.3% Ref 18% LD 13.8% |
|
|
|
Post by James E on Jun 13, 2024 20:18:44 GMT
Where are we at with the "Don't Knows"? Are they coming down % yet? Thank you, dodger ! This is what I want to know. Does anyone have access to the raw data? Labour's % seems to be coming down and others (RefUK, Lib Dems) seem to be going up, but is this because don't knows are splitting for the others/perhaps voting tactically or is it because people are switching from Labour? Must be there in the data.* *Hope I'm making sense. Just coming back from the pub. Thursday is the... old Thursday. Looking at YouGov's tables, the Con 2019 Don't Knows are a few points lower (from low 20s to 16%) than they were a few months ago, and Labour's Don't Knows are unchanged (8-9%). This has not made much difference to the polls, though it should mean that the 'adjusted' polling figures differ less from the rest than they did by a point or so. The largest change I can see for Labour is that they are taking less of the LD 2019 vote (from circa 40% to 20% per YouGov). Labour are also retaining a lower percentage of their own 2019 voters than they were. However, it's important to remember that the polling companies do change their methodology - in particular, by prompting according to the declared candidates in each constituency rather than a chosen list of parties. This has made a significant difference to some companies' figures. On a like-with-like basis, Labour have lost about 2 points on average from where they were in January to May . |
|
|
|
Post by James E on Jun 13, 2024 18:39:36 GMT
BMG poll Lab 41% (-1) Con 21% (-2) Ref 14% (-2) LD 12% (+3) Grn 6% (-) SNP 3% (-1) BMG's biggest Labour lead since October 2022. BMG only poll once a month, but comparing their 4 pre-election polls (averaged) with these figures shows a similar pattern to other pollsters: Lab 41% (-1) Con 21% (-5) Ref 14% (+2) LD 11% (+2.5) |
|
|
|
Post by James E on Jun 13, 2024 18:16:04 GMT
R&W - Labour leads by 24%.
Lowest EVER Conservative % (worse than Truss).
Highest Lib Dem % in 2024.
🇬🇧 Westminster Voting Intention (12-13 June):
Labour 42% (-3)
Conservative 18% (-1)
Reform 17% (–)
Lib Dem 13% (+3)
Green 5% (–)
SNP 3% (–)
Other 1% (–)
Changes +/- 7-10 June
....
These figures show almost exactly the same pattern as I have noted in other pollsters who have reported figures in the past 48 hours. Compared to R&W's pre-22ndMay averages, this one is;
Lab 42% (-2)
Con 18% (-4.5)
LD 13% (+3)
Ref 17% (+4)
|
|
|
|
Post by James E on Jun 13, 2024 16:48:12 GMT
WeThink @wethinkpolling 1/ Labour’s lead is cut by two but they still lead by 23 with just three weeks until the #GeneralElection. And independent candidates make our chart for the first time... 🔴 Lab 43% (-2) 🔵 Con 20% (NC) ⚪ Ref 14% (-1) 🟠 LD 11% (+1) 🟢 Green 6% (+1) 🟡 SNP 2% (-1) ⚫ Ind 2% I wonder if the Independents are or include Muslim Independent candidates which the recent poll of the VI of Muslim voters didn't seem to take account of. Savanta's recent poll of Muslim voters included the wording '...or would you vote for another party'. 5% of respondents chose this, up from 3% in a previous poll from Oct-Nov 2023. I suppose better wording might have been 'would you vote for another party or candidate', but there was an option for someone other than the named parties. I would be very surprised if Savanta did not include Muslim Independents under 'independents'. But I would also expect that any support for Galloways's WPGB would be considered as being for another 'party' (for which there is a total of 2% support, too) and not for an 'independent' candidate. |
|
|
|
Post by James E on Jun 13, 2024 15:58:38 GMT
We Think Westminster Voting Intention: LAB: 43% (-2) CON: 20% (=) RFM: 14% (-1) LDM: 11% (+1) GRN: 6% (+1) SNP: 2% (-1) Via @wethinkpolling , 12-13 Jun. Changes w/ 6-7 Jun. These figures reflect the same overall changes I have noted from the pre-election period ( i.e 1 Jan to 22 May, when Sunak called the election). Movements compared to WeThink's averages from that period are: Lab 43% (-2) Con 20% (-4) RFM 14% (+3) LD 11% (+2) GRN 6% ( 0) |
|
|
|
Post by James E on Jun 13, 2024 10:50:59 GMT
robbiealive Thanks for your kind words, although I don't think that my output of 9 posts in the past week can be described as "busy". Some regulars here have been known to post that many in an hour... As for Election Maps - I know no more about them than you can read from their twitter account. They seem to be the obvious source for newly released polls, and their nowcast looks decent to me.
|
|
|
|
Post by James E on Jun 13, 2024 10:15:57 GMT
Election Maps UK @electionmapsuk Westminster Voting Intention: LAB: 39% (-7) CON: 19% (-1) RFM: 17% (+3) LDM: 10% (+2) Via @peoplepolling , 11-12 Jun. Changes w/ 16 May. Do you (or anyone else) by any chance know if Election Maps have altered their methodology recently? If not, that's a chunky old reduction in LAB VI, along with several others recently. At least CON don't seem to be benefiting. This one seems to be an outlier, but there is a clear trend of a reduction in the Lab VI. I prefer, wherever possible to compare current polls to the pollsters' long term averages, and to factor in changes in methodology. So for me, YouGov's Lab 38% using their MRP approach needs to be compared to their two large pre-election MRPs which showed an average Lab VI of 40%. There is the additional factor of pollsters now prompting for parties according to the declared candidates for each seat, which does seem to have had some effect. Looking at the seven most recent polls by fieldwork, the Lab VI has done this: PP -6.5 (compared to 4 pre-election polls averaged) MiC -1.5 (compared to many pre-election polls, averaged) Norstat -4 (compared to their one other recent poll, as that's all we have from them) YouGov -2 (compared to 2 large pre-election MRPs) Focaldata -1.5 (compared to their 2 other election polls, as we have none from earlier in 2024) Survation -2 (compared to their most recent poll, as a change to their prompt prior to that has also clearly had an impact for smaller parties VI) Verian -0 (compared to their one previous poll a week earlier) That's an average loss of 2.5%, an a slightly reduced lead across these 7 polls of 19% ( Lab 40.4, Con 21.4, Ref 15, LD 11 ) |
|
|
|
Post by James E on Jun 11, 2024 16:25:23 GMT
Re YouGov' figures today:
Their 'unadjusted' figure for Labour has averaged 45% this year, while varying between 40 and 48.
Their 'adjusted' Labour figure in 5 previous MRPs and recent polls has averaged around 4 points lower, and varied between 39.5% and 43%. So this one is only about 2-3 points lower than what we should have expected, and may be random variation; it's a better figure for Labour than the 40% (unadjusted) that YG showed in one poll in late March.
|
|
|
|
Post by James E on Jun 11, 2024 9:10:49 GMT
Lots of talk about the Muslim vote deserting Labour, that may be a tad overblown... Savanta NEW Westminster Voting Intention among UK Muslims for 📈51pt Labour lead 🌹Lab 63 (-1) 🌳Con 12 (-7) 🔶LD 12 (+3) 🌍Green 7 (+2) ⬜️Other 5 (+2) 1,083 UK Muslims, 24 May - 3 June 2024 (all change from 27 Oct - 3 Nov 2023) These are very similar figures to those found by JL Partners in March, which were Lab 61%, Con 12%, LD 9%, Green 9%. Labour is still somewhat down from the 72% they were estimated to have achieved with Muslim voters in 2019, but not to the extent that some had predicted. But this probably should have been obvious, because with such a high share of Britain's 6.5% Muslim vote from 2019, a sudden really big loss (say, from 72% to 40%) would have knocked 2 points off the overall Lab VI last autumn. It should also have been very clear in VI figures for London, with a 15% Muslim population. But in fact, Labour's vote share in London appears to have held steady in London over the past year, or perhaps increased slightly, as the anti-Ulez campaign has faded. |
|
|
|
Post by James E on Jun 10, 2024 18:25:42 GMT
pingAll the MRPs I have looked at in detail assume that Con, Lab, LDs, Ref and Greens are standing, plus PC in Wales and SNP in Scotland. We didn't know until a couple of days ago whether this assumption would be correct, but it now seems that there will be a few seats where one of these will be absent. For the rare cases of one of these parties not standing (e.g Rotherham) the models won't work particularly well. I would be surprised if the pollsters now try to model this differently, for these odd cases, as it would require a different VI question to model for the absence of one party or another.
|
|
|
|
Post by James E on Jun 10, 2024 17:21:28 GMT
pjw1961 "...what's the current difference in RefUK VI between those that prompt for them and those that don't? Do Farage's estimates stand up?" There is no company which prompts for Con, Lab, LD and Green, but not Reform. The only pollster whose approach I am certain just about fits with Farage's description is Savanta, whose initial prompt does not include Reform UK, but they do prompt for Lab Con & LD . Those who choose 'another party' are then prompted for Green and RefUK. Savanta's VI for Reform has long been 2-3 points below the rest, but the same has been true of Survation with their 'no prompt for anyone' approach. I think that Deltapoll also don't prompt for Reform or Greens. The same trio - Survation, Savanta and Deltapoll - are the only pollsters who still sometimes register 1% (or sometimes 2%) for UKIP.
|
|
