|
|
Post by James E on Nov 7, 2023 10:55:15 GMT
I don't think we've had this recent-ish youGov London poll (fieldwork mid-October)
Westminster Voting Intention [London]:
LAB: 55% (-3)
CON: 20% (+2)
LDM: 9% (=)
GRN: 9% (+2)
RFM: 4% (-2)
Via
@yougov
, 12-17 Oct.
Changes w/ 27-31 Mar.
These figures are nearly identical to the average of the past 6 YouGov London cross-breaks which I posted yesterday. This reduced lead of 35% is somewhat of a reversion to the norm after the record-breaking 40% in March of this year. As Labour led by 48/32 in London at GE2019, this is a Con to Lab 9.5% swing.
As I have mentioned before, the Conservative vote is actually holding up a little better in London than elsewhere: by these figures, they are retaining 63% of their vote, as opposed to 54% overall per YG (24% av/44.7% in 2019), while the Labour VI is 'only' up by 7 points, as opposed to around 13 points across GB. I think this pattern is also evident from LE results in other 'safe-labour' areas - places such as Bristol and Liverpool, for example?
|
|
|
|
Post by James E on Nov 6, 2023 10:56:48 GMT
Re YouGov's " Sadiq Khan leads Susan Hall by 50% to 25% for London Mayor"
A couple points on this:
YouGov's London figures normally seem to show larger Labour leads than other pollsters. I generally trust YouGov ahead of the likes of R&W as they have a good track record from previous elections, and to poll London effectively you need to sample very carefully.
However, this is still less then the comparable figures for London Westminster VI. An average of the last 6 YouGov London cross breaks is Lab 55%, Con 20%, LD 9%, Green 7% - a Con to Lab swing of about 9% on GE2019. This compares to around 16% for YG's GB figures.
The lower Con to Lab swing in London has been evident for perhaps a couple of years. So it predates the ULEZ row - in fact, I would go so far as to say that by YouGov's figures ULEZ does not appear to have had an noticable effect on overall London VI, although it may be doing so more locally in parts of outer London.
|
|
|
|
Post by James E on Nov 4, 2023 11:54:39 GMT
Re - Esher & Walton. I am not sure the Tory vote here will fall much further - because it has already happened! I would also expect Labour to be polling circa 10% next time rather than 4.5% - on the basis that 'a rising tide raises all boats.' That raises some doubt in my mind as to the extent the LD vote here is likely to rise further. A strange comment, given that the Tories continue to poll very poorly in the South of England . However, I would acknowledge that they certainly fared badly at GE2019 in the two Elmbridge constituencies, with their vote down by 9 points in Esher & Walton, and 6 points in Runnymede and Weybridge. In fact, R&W was one of just thirteen constituencies to show an overall Con to Lab swing in 2019; most of the others were either in London, or had been Lab gains in 2017. Looking at the recent YouGov cross-breaks (average of last 6), which should be relevant here, we have: South of EnglandLab 39% (+16) Con 29% (-26) LD 14% (-3) Con to Lab swing of 21%. Social Class ABC1Lab 48% (+13) Con 24% (-19) Con to Lab swing of 16% 2016 Remain votersLab 57% (+8) Con 12% (-7) Con to Lab swing 7.5% So the Tories are faring badly in all the demographics relevant to these two constituencies. Looking at their proportionate losses, YG's GB figures show them retaining a net 54% of their 2019 vote. This compares to 56% for just the ABC1 cross-breaks, and 53% for the South of England. It is only with their 2016Remain/2019Con voters that they are doing slightly better, retaining 63%. [ graham - Re your response below. I discussed the continued relevance of Brexit to Voting Intentions in a couple of long posts of 21 Oct (pages 16 & 17 of this thread), so won't repeat them in full. YouGov's figures continue to show huge differences between Remain and Leave voters, with Labour leading by 45 points with the former, and the Tories by 15 points with the latter. By my figures, current VI is aligned about 87% with 'brexit attitude' in that those who think it was 'Right' very largely support Con or Ref, and those who think it 'wrong' largely support Lab/LD/SNP/Grn/PC. ] |
|
|
|
Post by James E on Nov 3, 2023 19:46:14 GMT
|
|
|
|
Post by James E on Oct 26, 2023 21:43:23 GMT
..... Average Labour lead of over 20% Loving this. Neil or anyone else far more knowledgeable on this stuff than me - a quick question. Is this 20%+ average lead amongst all pollsters the biggest since Truss? Thanks in advance if anyone knows. Labour's lead was in the region of 20-22% in Jan and Feb of this year, so the current lead is just below that but still the highest for nearly 8 months. There's a nice graph in the link below which illustrates this. The comparison I like best is with the polling average on 1st Jan 2023. It was then Lab 47, Con 25, and now it's Lab 46, Con 26. www.politico.eu/europe-poll-of-polls/united-kingdom/ |
|
|
|
Post by James E on Oct 25, 2023 15:38:26 GMT
NEW from @ipsosuk : Labour lead unchanged at 20 points🚨 Labour 44% (no change from Sept) Conservative 24% (-)Lib Dem 13% (+1) Green 9% (+1) Reform 4% (-)Others 7% (nc) Labour lead was 21 pts last November (just after Sunak took over) Fieldwork Oct 11-18th It's a pity that Ipsos do not do more frequent polls, as their methods are (or at least were) rather different to other pollsters, and their results are more unusual than they may appear - I have amended Mark's figures above to show Reform UK separately. The 4% for Ref UK is the lowest in any recent poll, but that has been so for all eight Ipsos polls we've had this year: they have ranged from 2% at worst to 4% at best with Ipsos. But this is in a poll where the Conservatives are also on a low figure. The combined Con +Ref share is just 28%, and it was the same in their September poll. This compares to 32-24% from nearly all other recent polls (Techne, YG, Delta, R&W, Savanta, WeThink, Opinium), while BMG and MiC have them at a combined 37%. As to what is causing this - I believe that Ipsos don't prompt for Reform UK (but neither do Survation nor Savanta), and they also don't weight by recalled past vote on the basis that this is unreliable. I wonder if anyone knows any more about their methods.. |
|
|
|
Post by James E on Oct 23, 2023 10:28:11 GMT
birdseye (and others who have commented on my Brexit/Voting Intention posts) Many thanks, and I take the point that Brexit is not cited by many as a 'current issue'. But that's not really what my post was about. The question I am looking at it how past Brexit-vote and current attitudes to Brexit now interact with Voting Intention. I am basing this on recent YouGov polls, and in particular, the cross-breaks of their 'Hindsight' tracker with current voting intentions. With apologies for re-capping on some stats I have shared before: current VI by Remain/Leave (in 2016) compared to GE2019 is: 2016 RemainLab 57% (+8) Con 12.5% (-6.5) Swing 7% 2016 LeaveLab 25% (+11) Con 44% (-30) Swing 20.5% (These are the averages of the past 6 YouGovs from 14 Sept to 18 Oct) So it seems very likely that we will see higher swings in Leave seats than Remain seats at the next election. Somewhat of a reversal of the two most recent elections, but nowhere near to taking us back to the pre-2016 electoral landscape. And the Conservatives are actually retaining a higher proportion of their Remain voters than of Leavers (or others). But the YouGov 'Hindsight' tracker is also relevant here, and this provides a rather different narrative. From its first use in 2016 up to around October 2022, this always moved in tandem with Voting Intention, albeit in a more muted way, as attitudes to Brexit are now very entrenched. The divergenge of VI and 'Hindsight' in Oct 2022 was caused by Truss managing to make the Tories far more unpopular with the public than Brexit was. This has now returned to a point where Brexit-support matches the VI of Con + RefUK. The most interesting detail is the analysis of the 'Right' and 'Wrong' answers by current VI. In broad terms, around 70% of all those who voted Leave in 2016 still think their decision was 'Right' (with around 20% 'Wrong' and 10% DK). Within the current Lab VI as a whole, just 11% of their support thinks Brexit was 'Right', despite around 22% of it comprising 2016 Leave voters. This means that at most 50% of 2023Lab/2016Leave voters think their 2016 decision was 'Right' (and it could be a bit lower). It also means that among Leavers-who-don't-intend-to-vote-Labour, 75% still think their decision was right. So Labour's partial recovery among 2016 Leave voters involves taking a high proportion of those who have changed their mind on Brexit. And because of this, Labour's recovery with 2016 Leave voters does not mean that the political landscape is not still shaped by (current) attitudes to Brexit. @mark - "How much of that shift is from people changing their minds?" The demographic effect of younger voters supporting EU Membership, and older voters having voted aroun 65/35 for Leave is worth around a 0.6% swing per year, or 4% since 2016. Current polling shows around 58/42 for Rejoin or 63/37 for thinking Brexit was 'Wrong', so that's 4 points in an overall 10-15 point swing.
|
|
|
|
Post by James E on Oct 21, 2023 21:48:09 GMT
|
|
|
|
Post by James E on Oct 21, 2023 19:39:22 GMT
Very surprising to read one regular contributor writing here yesterday that he thought Brexit had 'ceased to be a salient issue'. This led me to wonder if people don't know where opinion currently stands, and how this is continuing to affect voting intention. You are answering the wrong question there James. Regardless of what view people take about Brexit now, it (and rejoining the EU) don't rank highly on the list of the most pressing issues. If people have switched VI it is for other reasons than Brexit. yougov.co.uk/topics/society/trackers/the-most-important-issues-facing-the-countryThe question I was addressing was the interaction of Brexit viewpoint and current voting intention. Perhaps I should have avoided the words 'salient issue' - although those were the words graham used when expressing surprise that the Tamworth constituency did not revert to its 2005 position in Thursday's by-election. The point is that it is still a dividing line. While Labour has recovered somewhat in Leave areas, and the Tory vote has declined steeply, the Brexit-divide is still a very strong feature of British politics. It's a strong predictor of how people vote, regardless of whether it is cited as an issue. Perhaps this is best illustrated by looking at how pre-referendum voting patterns have changed: Con2015 voters were reckoned to have divided 61/39 for Leave in 2016 . Lab 2015 voters were 35/65 for Remain. Now compare these to the most recent YouGov Brexit Hindsight tracker: Current Con VI divides 76/24 for believing that Leaving was 'Right'. Current Lab VI divides 12/88 for believing that Leaving was 'Wrong'. (excl DKs in both cases) Compared to 2015, the current Con VI has become more pro-Brexit by 15 points, and the current Lab VI is now more anti-Brexit by 23 points. And only 5% of the electorate thinks Brexit was 'wrong' but supports the Conservatives, while on the other side around 6% think it was 'Right' but supports Lab or LD. I think there's a little bit of SNP or PC or Green Leave support, but even allowing a point or two for that, voting intentions are now about 87% aligned with pro-Brexit /anti-Brexit views. I won't go into the full details of the YG Hindsight tracker, but suffice to say that this level of alignment is very slightly higher than as of 26 Jan 2020 - the closest YouGov on this issue to both GE2019 and the UK Leaving the UK. |
|
|
|
Post by James E on Oct 21, 2023 14:25:50 GMT
Very surprising to read one regular contributor writing here yesterday that he thought Brexit had 'ceased to be a salient issue'. This led me to wonder if people don't know where opinion currently stands, and how this is continuing to affect voting intention. At the last General Election, Brexit was a huge dividing line. Leavers supported the Conservatives by a 60-point majority (74/14), while Remainers backed Labour by a margin of 30 points (49/19). Since then, this divergence has narrowed somewhat: Labour is now faring considerable less badly with Leavers, closing the defecit from one of 60 points to around 20. Nevertheless, the net difference between the voting intentions of 2016 Leavers and Remainers is still well in excess of 60 points. This is still far larger than it was in 2015, let alone what it might have been in 2005. Another way of looking at this is that 2016 Leave/Remain vote is almost at strong a determinant of voting Intention as age: for comparison, per the past 5 YouGovs, the 18-24s and over-65s are typically around 80 points apart in their levels of support for Con/Lab. But the way in which the 'Brexit-divide' has narrowed recently is worth looking at in detail. At GE2017, Leavers were reckoned to have backed the Tories by 64/24 over Labour. This widened by a 10-point swing to 74/14 in 2019, but is now around 44/25 per the average of recent YouGov polls. So Labour is now doing only slightly better in absolute terms than in 2017. But the Tory Leave vote has collapsed by 30 points from their GE2019. This produces a 20% swing among Leave voters from then, and indeed a 10-point swing from GE2017. Looking at the current Con and Lab VI by 2016 vote, it is clear that these are now still more aligned with Remain/Leave than they were in 2015. So while the Conservatives' 2015 voters were reckoned to have backed Leave by 61% to 39% in 2016, their current VI divides around 23% Remain, 77% Leave (among those who voted). For Labour, their 2015 vote divided 65/35 for Remain. Per the most recent YouGov, their overall VI divides around 56/22 on 2016 Remain, but with a significant proportion (22%) who either could not or did not vote. This would include all of their support from those aged under 25. Re-basing the 2016 Referendum voters in the current Lab VI, they work out as around 72% Remainers, 28% Leavers. But there is still a distinction to be made between 2016 Referendum vote and current attitude/support for Brexit. I mentioned above that the most recent youGov showed 22% of the Labour VI coming from those who voted Leave in 2016. However, turning to the Brexit tracker* in the same poll, only 11% of the current Labour VI think we were 'right' to vote to Leave - and this is fairly typical of their recent polls. This means that among Labour Leavers, no more than 50% of those who backed Leave think their decision was the right one, compared to around 70% for Leavers as a whole. In other words, those Leavers who regret their decison are far more likely to now support Labour than those who still think Brexit was 'right'. And in the context of the Tamworth by-election, it's important to acknowledge how far opinion on Brexit has changed since we actually left. In 2016, Tamworth voted 66% for Leave. But given that YouGov's tracker has moved from 48/52 for 'Right to Leave' in 2016 to (typically) 63/37 for 'Wrong', it seems likely that the same question in the Tamworth constituency would now produce around a 50/50 result. * 'in hindsight, do you think Britain was right or wrong to vote to leave the EU?' d3nkl3psvxxpe9.cloudfront.net/documents/TheTimes_VI_231018_W.pdf |
|
|
|
Post by James E on Oct 20, 2023 10:33:01 GMT
|
|
|
|
Post by James E on Oct 20, 2023 7:40:07 GMT
What we don't know is how many people in Mid Beds voted tactically for Labour even though they would really have preferred to vote Lib Dem. The result probably underestimates the true Lib Dem support. Really? The Mid Bedfordshire constituency has until now been a very good predictor of the LibDems GB vote share - at least so far this century, and probably longer. And this has happened in a constituency where there was no case to be made for tactical voting between Lab and LD, as it has until now been very safe for the Conservatives. LDs' Mid Beds vote v GB shareGE 2019 12.6% v 11.7% (+0.9) GE 2017 6% v 7.4% (-1.4) GE 2015 7.2% v 7.9% (-0.7) GE 2010 24.9% v 23.0% (+1.9) GE2005 23.8% v 22.0% (+1.8) GE 2001 19.7% v 18.3% ( +1.4) So now with the LDs on around 11% in national polls, you are saying that 23% in a by-election "underestimates the true LibDem support". |
|
|
|
Post by James E on Oct 20, 2023 6:05:23 GMT
I'm sticking to my view there really isn't any such thing as 'inside knowledge' of elections, just speculation. The fact that so many of us on this site, all dedicated followers of politics, often for decades, are amusing ourselves by trying to guess the results tends to support that. After the results are known, those who guessed right can pat themselves on the back for their profound insight and those who were wrong can find reasons to explain it away as usual!  . P.s. - a light-hearted post, please don't take offence! In line with the above rule, as I predicted two Labour wins I am now declaring my genius!  ....and I'm doing the same. ukpollingreport2.proboards.com/user/48/recentInteresting that the Conservatives fared worse in Mid Beds than in Tamworth. With the polls currently averaging Con 28, Lab 45, the Tories are retaining 0.62 of their GE2019 share. The 40.7% share in Tamworth was 0.61 of their 2019 vote, while the Mids Beds share of 31.2% was just 0.52 of their 2019 vote there. The 'on par' result in Tamworth is very much consistent with what has happened to the Conservative vote in recent Con/Lab by elections, with them losing votes proportionately in line with what contemporaneous polls were showing. This was also true for the Chester, Stretford and Selby by-elections. The fared somewhat better in West Lancashire, a lot better in Uxbridge, but worse in Wakefield. The effect is different when the LibDems become the main challengers : the Conservative vote share falls by even more than the proportion implied by the polls. This was true in Somerton and Froom, Tiverton, North Shropshire and Chesham and Amersham. Because of this, I think it is at least arguable that they may have contributed to the very poor showing by the Tories in Mid Beds. |
|
|
|
Post by James E on Oct 14, 2023 18:20:41 GMT
Re the two by-elections next week, my prediction Tamworth a Labour gain Mid-Bedfordshire- Tories narrowly hold against a split ABT vote Come the General Election Labour doesn't need to win either one to get a huge majority The timing of these two by-elections looks favourable for Labour, and I'm allowing myself a couple of optimistic predictions. Mid Beds Lab 34 (+12) Con 32 (-28) LD 18 (+5) Ind 8 Grn3, Ref 3, oth 2 Tamworth Lab 43 (+19) Con 39 (-27) Ref 7 LD 4, Grn 4, oth 3. |
|
|
|
Post by James E on Oct 14, 2023 17:13:45 GMT
Lulu/Steve, Mids Beds doesn't really matter, other than a morale boost or deflation as it will be Tory at the next GE. As such important for both LDs and Lab to contest for narrative purposes. Failing to win Tamworth would be a more serious blow for Labour given that the party held the seat 1997 - 2010. Tamworth is currently the Conservatives' 55th safest seat, although there are good reasons for regarding it as an easier Labour target than Mid Bedfordshire at No 98. But there are a number of seats which Labour won in 1997 which are now even 'safer' for for Tories than Tamworth. Looking down the list of currently safe Conservative seats which were Labour's in the 1997 landslide, we have: Wyre Forest, No 62 Tamworth, No 55 Chatham and Aylesford, No 53 Cannock Chase, No 51 Cleethorpes, No 32 Sittingbourne and Sheppey, No 22 Braintree, No 17 Castle Point, No 3 Heading in the opposite direction, the Conservatives' safest seat in GE1997 was Huntingdon. As of 2019, it has fallen to No 149. www.electionpolling.co.uk/battleground/defence/conservative |
|
|
|
Post by James E on Oct 14, 2023 11:08:01 GMT
Summary of my above post Average seat projection: LAB: 420 CON: 149 LD: 23 11k+ poll, then modelled using MRP. Polling done 11-25 Sept Westminster VI LAB: 45.7% CON: 29.1% LD: 12.1% PC: 0.7% SNP: 3.5% Other: 9% There's a comparison to be made here between the 420 Labour seats produced by this MRP, and the result of the same swing (of 14.35%) using UNS. Unfortunately, I don't have a "notional majority-sorted" list of target seats on the new boundaries, but using the old boundaries, such a swing would only produce 356 Labour seats. That's 64 fewer seats - or a majority 128 lower - using UNS (and old boundaries which were probably a little bit better for Labour than the new ones). Incidentally, the seat total imply that they have the SNP on around 34 seats, so still a plurality in Scotland in the context of a 190-seat overall Labour majority. To add to the above - there figures from Savanta's MRP (and indeed any other MRP) provide a possible rationale for determining the percentage lead Labour might need for a majority, which we discussed a few days ago, in the wake of Peter Kellner's article 'Labour's mountain is not as high as it was'. Those 420 Labour seats would imply 218 gains on GE2019, in the context of a 14.1% Con to Lab swing. That's 15.46 seats for each 1% of swing, so noticably more than the "10 per 1% swing" implied by UNS*. With that rate of seats gained, the implied Labour lead needed for 326 seats is 4.2%. These figures are remarkably similar to an earlier, but larger MRP we had from Survation with fieldwork to 1 Sept. This put Labour on 426 seats, so 224 gains. The 14.5% swing for those gains yielded 15.45 "seats per 1%". And from this, the implied Labour lead needed for a majority would be 4.4%. The slight weakness of this is that with Labour leads of 16.5% and 17.4% respectively in these two MRPs, they may be gaining more 'seats per 1%' if their overall lead is in double figures. On the other hand, both of these MRPs still show the SNP dominant in Scotland (36 seats, per Survation), and significant gains for the LDs, so this is not over-generous to Labour on either of those criteria. *This is the same as the '5 seats per 1% lead' cited by Peter Keller a few days ago, and is consistent with a majority-sorted list of Labour targets on the 2019 boundaries. kellnerpolitics.com/2023/10/12/labours-mountain-is-not-as-high-as-it-was/ |
|
|
|
Post by James E on Oct 13, 2023 9:58:08 GMT
YouGov tables for their 11-12 Oct poll are here: d3nkl3psvxxpe9.cloudfront.net/documents/TheTimes_VI_AdHoc_231012_W.pdfA couple of things to note, beside the 47/24 Labour lead: As this is all post Rutherglen by election, the SNP leading Labour by 39/34 in Scotland may seem surprising (sample size 180). No real impact on the SNP's lead which was averaging around that (5-7%) across YouGov's data in the past 2-3 months. Likelyhood to vote rarely gets a mention but it may be significant that 71% of their Lab2019 voters say they are 'certain to vote', as opposed to 58% of Con 2019. And this is fairly typical of previous YouGovs. If this were to happen in an actual election, it would undermine some of the rationale for Opinium's upweighting to reflect the 2019 turnout of Con and Lab voters. My best guess is that the 'real' position is probably in the middle of Opinium's and YouGov's headline figures. |
|
|
|
Post by James E on Oct 12, 2023 21:48:11 GMT
JamesE/jimjam I think Peter Kellner needs to talk to the BBC, and Chris Mason in particular, about his revised calculations. Throughout Labour's conference, Mason continually trotted out the line, "Labour need a swing of 12% just to get a one seat overall majority". He said it with a sort of implied incredulity that anybody in Liverpool could possibly believe it was possible to "climb the mountain." But, of course, the psephological illiteracy of political commentators like Kuennsberg, Mason, Neil, Wark etc is legendary. Canards, once they are recited repeatedly, are difficult to rebut. With apologies for re-posting figures I have used before - the easiest rebuttal is to quote the equivalent "Percentage lead Labour need to win a majority", as it has stood after the past 5 General elections. These were: Post-2005 -2% Post -2010 2% Post-2015 9% Post-2017 7% Post-2019 12% It has always been a moving target - as evidenced by both this and the huge shifts in the equivalent target for the Conservatives (which jumped from 6.5% in 1992 to 13% post GE1997). |
|
|
|
Post by James E on Oct 12, 2023 20:37:08 GMT
James E, Re Kellner, he really is not the guru he once was and still doesn't get that UNS understates seat changes when there is a large swing. I recall at the time explaining to party member friends who follow polls casually that he was wrong. Pleased that most accepted my analysis which was helped by cutting and pasting some of your contributions on here. Many thanks for that, JimJam. I remember that when he wrote the 'grumblers and defectors' article, the 'grumblers' were meant to be the Con2019 voters currently answering 'Don't Know' - of which there are of course many. The problem with this is that there was a very simple way to test it - look at the detailed findings of Opinium, who exclude the DKs and readjust their weightings to reflect the positive answers only - and hence produce Labour leads around 6 points lower. I checked Opinium's figures at the time, and these did not in any way bear out his 'grumblers' narrative. And looking at their most recent polls, I've found that, if anything, they show the opposite effect. While YouGov seem to show a proportionately-even vote loss for the Tories across the North, Midlands and South of England (and London), Opinium's post-adjustment figures actually show the biggest proportionate losses in the Midlands and South. These are their averaged cross-breaks in England in 6 polls since the start of August. These show a 12.5% GB swing from Con to Lab, and an overall Con vote retention of 61% (27.3% now, divided by 44.7% in 2019) - typically Con 27%, Lab 41%. LondonCon 24.5% (-7.5) Lab 51% (+3) Swing 5%, Con Vote retention 76% South EnglandCon 32% (-23) Lab 33% (+10) Swing 16.5%, Con vote retention 58% MidlandsCon 29% (-25) Lab 44% (+11) Swing 18%, Con vote retention 53% North EnglandCon 22% (-17) Lab 54.5% (+11) Swing 14%, Con vote retention 67% These figures show very much the same pattern as YouGov, albeit that the swings are lower due to Opinium's re-weighting methodology. So to take the comparison of 'local' swing to GB-wide swing: in London, YG and Opinium have swings 6 & 7 points lower respectively than for GB as a whole . In the North they both have a swing 2 points lower. In the South, they both have a swing 4 points higher, and in the Midlands, they diverage a bit, with YG 2 points higher, and Opinium 5 points higher. |
|
|
|
Post by James E on Oct 12, 2023 18:03:02 GMT
Interesting analysis, Labour may get an overall majority with as little as a 5% lead {Tweets of Peter kellner's latest blog} As this is what I have been saying on UKPR2 for most of this year, I can't help feeing rather vindicated: Kellner has seemingly now recanted on what he wrote on the same blog only 4 months ago, when he claimed that Labour would need a 13% lead for a majority. That was his 'grumbers and defectors' article dated 24 May, if anyone remembers that. I stated here at the time that he was wrong, and it appears that he's now accepted that, although he still seems to want to stress Scotland and tactical voting, rather than the Conservatives' proportional vote losses. And it's not as if the polls have moved much in Labour's direction since then, although they have narrowed the SNP's lead by a few points in Scotland. ukpollingreport2.proboards.com/user/48/recent?page=6(various posts near the bottom of this page) and... ukpollingreport2.proboards.com/thread/67/labour-win-outright-majority |
|
|
|
Post by James E on Oct 12, 2023 17:08:36 GMT
My one concern about Labour's performance is that they have been consistently losing support in the "Blue Wall" seats over the last year. See the Blue Wall tab on ukpollingreport2.proboards.com/post/103418That means they are going to need to rely more on the Lib Dems to win seats off the Tories in these seats, and that means encouraging tactical voting as in 1997, not hammering away at each other as in mid Beds at present. My reading of R&W's Blue Wall may be a bit different to yours. It has certainly moved more than other polling figures, though. It showed a Labour lead of 6-7% in Feb& March, whereas the last 4 average a Tory lead of 0.5% - so a movement of 7 points. But taking R&W's GB polls from the same periods, their average Labour lead has fallen from 24% in Feb & March to just under 17% across the past 9 (which correspond to the same time period as their last 4 'Blue Walls'). So we may be looking at R&W's house effects, or perhaps a methodological change in the past 6 months, rather than something locally specific to their 42 'Blue Wall' constituencies. R&W's 'Blue Wall' is a somewhat odd combination of 10 London seats, and 32 other South of England ones, but with few in the East. They were all at least 42.5% Remain-voting, and by my estimate would have been 55-60% Remain overall in 2016. As I noted a few weeks ago in relation to YouGov's cross-breaks, the Con to Lab swings are a lot lower with Remain voters and London voters. In the context of this, it may not be so surprising that the Conservatives appear to be retaining somewhat more of their 2019 support (66% v 61%) in the Blue Wall than in GB as a whole. The overall swing per R&W in the 'Blue Wall' compared to GB as a whole looks much the same on average in the past couple of months, at around 14% in each. Meanwhile, other pollsters continue to show a larger than average Con to Lab swing in their South of England cross-breaks. For both YouGov and Opinium, this is running at around 4 points higher than their general Con to Lab swing. So for YG, this is averaging 20% (v 16%) and Opinium 16.5% (v 12.5%). |
|
|
|
Post by James E on Oct 9, 2023 13:03:38 GMT
On VAT on private schools, has anyone done the numbers on what it will generate from VAT, versus what it may cost for those who elect instead to use the state system instead? Not that I am arguing for public schools, I was just wondering how much it will actually generate overall in practice? A few figures that could help inform this: Average state spending per pupil is around £7,500, but we should be aware that this is an overall average and is always higher in areas of deprivation. So as we are (presumably) discussing non-deprived pupils, I'd suggest a lower figure of around £6K per pupil, or maybe less. (My local primary school receives basic funding of a little over £4K per pupil). Average private school fees are around 17K per year (according to the souce I've just googled). It would be a slight overestimate to simply take this figure and multiply by the VAT rate of 20%, as private providers would be able to recover VAT on costs, which they can't do so long as private school fees remain exempt. If you estimate that one-tenth of their fee income is VAT-able costs, the average VAT generated would be around £3,000. By these figures, you would get an overall break-even if 1 in 3 of all privately-educated pupils left to go to state schools. |
|
|
|
Post by James E on Oct 7, 2023 21:42:47 GMT
Today's Opinium of Lab 42%, Con 29% would be a lead of 18-19 points without their re-weighting ( Lab 45.3, Con 26.7). As ever, their actual fieldwork findings are similar to YouGov, and it is the re-weighting which produces a 'headline' Labour lead around 6 points lower with Opinium than YouGov.
Opinium's average from July-Sept was a 14.5% lead, so this one is consistent with the pattern of a very slight narrowing recently.
|
|
|
|
Post by James E on Oct 7, 2023 15:22:13 GMT
Summary of my above post Average seat projection: LAB: 420 CON: 149 LD: 23 11k+ poll, then modelled using MRP. Polling done 11-25 Sept Westminster VI LAB: 45.7% CON: 29.1% LD: 12.1% PC: 0.7% SNP: 3.5% Other: 9% There's a comparison to be made here between the 420 Labour seats produced by this MRP, and the result of the same swing (of 14.35%) using UNS. Unfortunately, I don't have a "notional majority-sorted" list of target seats on the new boundaries, but using the old boundaries, such a swing would only produce 356 Labour seats. That's 64 fewer seats - or a majority 128 lower - using UNS (and old boundaries which were probably a little bit better for Labour than the new ones). Incidentally, the seat total imply that they have the SNP on around 34 seats, so still a plurality in Scotland in the context of a 190-seat overall Labour majority. |
|
|
|
Post by James E on Oct 6, 2023 15:24:04 GMT
LAB: 44% (-3) CON: 28% (+1) LDM: 10% (=) RFM: 6% (-1) GRN: 5% (=) SNP: 2% (-1) Via @wethinkpolling , 5-6 Oct. Changes w/ 28-29 Sep. It won't get much publicity as it still shows Labour 16 points ahead, but if compared to the norm for 'We think', this is more of an outlier than Opinium's 39/29. It is fully 5 points below Wethink's July-Sept average Labour lead of 21%, whereas Opinium's was 4.7 points below their 14.7% average. On the other hand, there are a couple of pollsters who have reported in the past few days whose latest figures have moved in the opposite direction (YG and Savanta). The overall average, taken across the 7 most recent polls, starting with Opinium's 39/29 is a 2-point reduction from the same pollsters' July-Sept average Labour lead.. en.wikipedia.org/wiki/Opinion_polling_for_the_next_United_Kingdom_general_election |
|
|
|
Post by James E on Oct 6, 2023 11:14:37 GMT
Their Brexit Hindsight tracker is showing a new record : a 26-point lead for 'Wrong to Leave' by 32/58 with 10% Don't Knows. This becomes a 29-point lead after excluding the DKs. And the age-divide is as stark as ever: "In hindsight, do you think Britain was right or wrong to vote to leave the European Union?" Voters aged 65+ : 59/41 for 'Right to Leave' - a 5% swing on how they voted in 2016. Voters aged under 65: 28/72 for 'Wrong to Leave' - a 20% swing on how they voted in 2016. |
|
|
|
Post by James E on Oct 6, 2023 10:59:07 GMT
EmCat, Both ON conceptually and James E empirically have amply demonstrated that a proportionate swing approach between all parties in Scotland is way better than UNS. Across E&W, while UNS will be less accurate than a proportionate approach when looking at specific seats it wont be that far off in aggregate; although James has shown Labour probably win a few above UNS.In Scotland, it is more complex with more players (and Unionists partly coalescing) so UNS is even less applicable and the tipping point when Labour make big gains off the SNP higher than the simplistic Election Maps modelling suggests. UNS generally works OK for an approximation of overall seats when there is a swing of 5% or less. So I would be happy to use it to predict the point at which the Conservatives lose their majority, for example (around a 3-4% lead). But it did not work in 1997 for a 10% swing. Labour overachieved the UNS model by around 30 seats (or +60 to their majority), getting 12-13% average swings in Conservative-held targets, but only around 8% in their own safe seats. I would expect the next GE to diverge from UNS by more than this in England. |
|
|
|
Post by James E on Oct 6, 2023 10:45:19 GMT
@techneuk NEW POLL: Labour lead by 19 points: Lab: 45% (nc) Con: 26% (-1) Lib Dem: 11% (+1) Reform: 6% (nc) Green: 6% (nc) SNP: 3% (nc) Others: 3% 👥 1,635 questioned. 🗓️ +/- 26-27 Sep Data - technetracker.co.ukFieldwork dates are given as 4-5 Oct, so the same as for the 21-point lead per YouGov. 10-poll average now stands at a Labour lead of 18%. So much the same as it has been for 6-7 months. |
|
|
|
Post by James E on Oct 6, 2023 9:19:21 GMT
Rutherglen & Hamilton West By-Election Result: 🌹 LAB: 58.6% (+24.1) 🎗️ SNP: 27.6% (-16.6) 🌳 CON: 3.9% (-11.1) 🔶 LDM: 2.9% (-2.3) 🌍 GRN: 2.0% (New) ...etc... Changes w/ GE2019 Congratulations to Labour on their best ever by election result, it's a by election I know , normal caveats apply but if Labour did the same at a general election in Scotland the SNP would be reduced to around 10 mps The swing here is way beyond what I, or most people, expected. What we have here is the tendency of by-elections to exaggerate the general swings shown by polling - so the SNP share is down by 16 points rather than the 8 or so per Scottish Westminster polls. There is very likely to be a short term boost for Labour in Scotland due to the 'halo effect'. This invariably wears of after a couple of months, so it may be best to base any predictions for GE2024 in Scotland on where the polls are in December or January rather than the next one or two. And, to address steve 's 'Just for fun' projection for seats in Scotland using UNS. Such projections are particularly misleading for the SNP/Con battleground. This one shows Con gaining a seat, as they were only down by 11 points in Rutherglen, whereas the SNP were down by 16. But all polls (and this by election) show the SNP holding a larger proportion of their vote than the Tories, and this is likely to be a more accurate pointer for such contests. |
|
|
|
Post by James E on Oct 4, 2023 11:23:09 GMT
sheviiHS2 should free up all the capacity of the southern half of WCML, which in turn enables the provision of a very large number of other services - many of which may only need to use a small part of the current main line. The article below is an admittedly pro-HS2 explanation of how that capacity could be used. The important point is that a wide range of new train services could be provided between places which currently have no direct train service. www.globalrailwayreview.com/news/89694/hs2-capacity-britains-existing-rail/"HS2’s capacity-releasing effects on the conventional network mean that Coventry will benefit from; new direct connections to and from Derby, Sheffield, York and Newcastle; more frequent services to and from Shrewsbury, Telford, Leamington Spa and along the Coventry-Birmingham commuter corridor; as well as enjoying less crowded trains on existing services to and from London. Benefits will also spread to stations such as Lincoln, where additional capacity can be deployed to provide more frequent connections to and from Newark, Grantham, Stevenage and London."
|
|
.
