|
|
Post by James E on May 5, 2023 8:57:36 GMT
Re the LibDems traditional overperformance in Local Elections compared to GE polling.
Here are their Net Equivalent Vote percentages for the past 6 years, compared to their average polling figure for the 6 polls before each years local elections.
2022 19% v 11% in GE polls(+8)
2021 17% v 8% in GE polls (+9)
2019 19% v 9% in GE polls(+10)
2018 16% v 8% in GE polls (+8)
2017 18% v 7% in GE polls (+11)
So if the LDs do achieve a NEV of 21% compared to 11% in recent polls, that would be entirely normal.
One rider to this: I am comparing to polls before the LEs in each year. It's also normal for the LibDems to get a polling boost afterwards when voters see that they have done 'surprisingly well'.
|
|
|
|
Post by James E on May 4, 2023 11:31:56 GMT
So given how in the last couple of years Labour’s polling fell while UKUP’s rose, and Tories and Lib Dem’s stayed relatively flat, where do you think UKIP votes came from in that period? Using my Dec 2012/Feb 2013 YGs linked above, the rise in the UKIP vote was only a couple of points from then onwards- from 10% to 12%. UKIP's rise was mostly before then, and clearly came at the expense of the Tories @ around 11-17% of their 2010 voters. But if you look at the Lab>UKIP switching as per the tables I linked, it was only 1% or 2% at that time, as opposed to 5-6% at GE2015. So this direct Lab>UKIP switching after Feb 2013 could account for most of the UKIP gains made from that point onwards (plus a bit of LD>UKIP which also rose from then onwards). |
|
|
|
Post by James E on May 4, 2023 11:20:24 GMT
RE General Election 2015. etc But a fair chunk of the 2010 Labour vote deserted the party in 2015 - and not just in Scotland. Where did those votes go? Using the last 4 pre-GE YouGovs from May 2015, the Lab 2010 voters divided as: Con 7% Lab 78% LD 3% UKIP 5% SNP 4% |
|
|
|
Post by James E on May 4, 2023 11:04:06 GMT
RE General Election 2015. Here's a couple of YouGov polls from Feb 2013 and Dec 2012 with the Labour lead at 14/15 point - so the peak of what Labour achieved under Ed Milliband. cdn.yougov.com/cumulus_uploads/document/amy2lnthes/YG-Archive-Pol-Sun-results-040213-Trackers-EU.pdfcdn.yougov.com/cumulus_uploads/document/f9ncl9h13g/YG-Archive-Pol-Sun-results-041212.pdfThe 2010Con VI (after DKs) divides: Con 74% / 77% UKIP 15% / 16% Lab 9% / 5% LD 2% / 2% So it must be clear that Milliband never achieved a great deal of Con>Lab movement. The high Lab VI rested on gains from the LDs, as well as Con losses to UKIP. And of course there are more ConDKs than Lab DKs. An ICM (or Opinium style) reversion of these would knock about 3 points off that 15% Lab lead. But in any case, the average labour lead was never 15% - but more like 11%. Once you allow for the reversion of the DKs and the large scale of the GE2015 polling error (a 3% Lab>Con swing), it's valid to ask how far ahead Labour really were at the time - as opposed to what the polling headline figures said. I would suggest that is was probably only really by 4-5%. |
|
|
|
Post by James E on May 4, 2023 10:09:45 GMT
There is a wealth of polling evidence from 2015 as to how voters moved - so rather than guessing or 'suspecting' it's best to refer to that. As I have mentioned before, Tory voters from 2010 were far more likely to switch to UKIP at GE2015 (15% of them) than the 6% of Lab 2010 voters who were found to have done so. It is impossible to tell how voters moved from their position in mid parliament… No James, if something is possible, but you don’t currently have enough evidence to gauge whether it’s happening or not, it’s not necessarily rational to dismiss the possibility just because you can’t currently prove it one way or another. You might not necessarily be able to prove up front that someone would definitely get run over the next time they walked across a road blindfolded, but that doesn’t mean it’s rational for them to ignore the possibility. comsidering possibilities you can’t currently prove is a survival trait. (But that’s enough about Covid…) I'm not 'dismissing' your theory, but I am saying that it is not evidenced. And as polling organisations do not have a record of their respondents' individual VI from the middle of the 2010-15 parliament, there is little-to-no likelyhood of that evidence ever becoming available to test your Con>Lab>UKIP theory. My approach is always to start by looking at the available polling evidence and to see what fits with the evidence, rather than speculating as to where the bounds of possibility lie. |
|
|
|
Post by James E on May 4, 2023 9:18:44 GMT
I suspect that in 2015 Labour picked up supprt from 2010 LDs whilst losing former supporters to UKIP. Cameron stemmed the likely flow of 2010 Tory votes to UKIP with his Brexit Referendum promise - and also picked up quite a few 2010 LD votes. There is a wealth of polling evidence from 2015 as to how voters moved - so rather than guessing or 'suspecting' it's best to refer to that. As I have mentioned before, Tory voters from 2010 were far more likely to switch to UKIP at GE2015 (15% of them) than the 6% of Lab 2010 voters who were found to have done so. It is impossible to tell how voters moved from their position in mid parliament, but generally the Con2010 cross beaks show high levels of switching to UKIP and a lot to 'Don't Know'. Lab made most of their gains from the LDs - and also from having lower levels of their 2010 supporters answering 'Don't Know'. I don't think that Cameron 'stemmed the likely flow' of Con>UKIP switching. The Con>UKIP crossbreaks were around 12% at Labour's polling peak in mid 2012, but were somewhat higher (15%) in May 2015. As for the 2010 LD voters, YouGov record these as splitting in all directions: 29% stuck with the LDs, 29% to Lab, 16% to Con and 10% to UKIP. |
|
|
|
Post by James E on May 3, 2023 20:38:52 GMT
That Savanta poll of 18-25 year olds is not as bad as it might be for the Tories.
YouGov's last 5 age 18-24 cross breaks average at:
Con 9%
Lab 60%
LD 8%
Green 10%
|
|
|
|
Post by James E on May 3, 2023 19:41:53 GMT
Who needs polling when you can just use guesswork? Considering different possibilities, forming hypotheses to consider is normal in science James. And so is evidence. |
|
|
|
Post by James E on May 3, 2023 19:33:58 GMT
That graph tells us nothing about where UKIP got their voters from. But an analysis of YouGov's polls immdiately before GE2015 showed that they took 15% of Con2010 voters, 10% of LD2010 and 6% of Lab 2010. Ah, but it’s not all about a comparison with 2010. For example, it might well be that initially UKIP took more from the Tories, but as time wore on, and immigration got hyped, UKIP took an increasing number from Labour. Including maybe some who had switched to Labour post-2010. Who needs polling when you can just use guesswork? |
|
|
|
Post by James E on May 3, 2023 19:25:07 GMT
As for Miliband, others argued in the past he got done in part by the rise in the UKIP vote. Here’s the graph anyways, see whst you think… [graph deleted] That graph tells us nothing about where UKIP got their voters from. But an analysis of YouGov's polls immdiately before GE2015 showed that they took 15% of Con2010 voters, 10% of LD2010 and 6% of Lab 2010. |
|
|
|
Post by James E on May 3, 2023 18:54:46 GMT
You are missing this point. They chose to vote Labour rather than Liberal Democrat, as they might have done and did in the 1980s. Why? Because the Tories were more loathed by then? I read somewhere that Tories stayed at home. 4.4 Million of them certainly stopped voting Tory, although yes Blair's vote went up 2 million. LD was not that much down. So I think not scaring the Tory horses could be said to be one of the Blair "achievements" to getting the landslide, just as it looks like being a win for Starmer on current polling. TBC but Starmer does have a higher lead that Miliband's single figure leads 18 months before the 2015 election. Looking back to 20 months before the 2015 Election (as we may have to wait that long for Jan 2025), Labour under Milliband had an average lead of 5-6 points, compared to 15 points now. Of course, it is doubtful that he even had that much of a lead, as the pollsters were all about 6 points adrift when the actual result emerged in May 2015. Comparing polls-with-polls, the Tories achieved a 3% swing-back in those 20 months to turn a 5 point Lab lead to a 1-point Con lead. |
|
|
|
Post by James E on May 3, 2023 18:46:17 GMT
I am rather busy atm but would be goof if someone unpick that Scottish R&W to see how many SNP2019-Lab net switchers there are. I expect much of the narrowing to be due to SNP 2019 DKs but facts would be nice to know. SNP 2019 voters now (Headline, excluding don't know): SNP 74% Labour 16% (3% of 2019 Labour voters have switched to SNP) Lib Dem 3% Con 2% Green 2% SNP 2019 voters including don't know but weighted by likelihood to vote SNP 70% Labour 16% (3% of 2019 Labour voters have switched to SNP) Lib Dem 3% Con 2% Green 2% Don't know 6% SNP 2019 voters including don't know unweighted by likelihood to vote SNP 68% Labour 16% (2% of 2019 Labour voters have switched to SNP) Lib Dem 3% Con 2% Green 2% Don't know 7% Won't vote 1% Edit: So more of a direct switch to Labour than to don't know by more than 2 to 1. Obviously, many of those voters will have voted Labour in the past (i.e. pre-2014) so it is presumably a less difficult move (in either direction) than switching between Labour and Tory is in England. Thanks to pjw for that. I would just add as a qualifier to this, that R&W do not use a panel, and their cross-breaks are far more variable than those of YouGov or Opinium. The last 2 full Scottish YouGov's showed 13% of 2019SNP>Lab and 13% of 2019SLab>SNP (after excl DKs). This was in the context of a 37/28 SNP lead. The previous YouGov in March had 12% SNP>Lab and 11% SLab>SNP. |
|
|
|
Post by James E on May 3, 2023 15:59:56 GMT
This does indeed appear to be new although the dates seem to be sluggish (so wouldn't cover what could be a stabalisation/improvement for Lab in the last 2 or 3 polls) and I've never held much belief in Santana to begin with: Stats for Lefties 🏳️⚧️ @leftiestats · 49m 🗳️ NEW: Labour lead drops to 11pts (-3) 🔴 LAB 42% (-3) 🔵 CON 31% (-) 🟠 LD 9% (+1) 🟣 REF 7% (+2) 🟡 SNP 3% (-) 🟢 GRN 3% (-) ---- 🔴 Labour majority of 28 seats Via @savanta_UK , 21-23 Apr (+/- vs 14-16 Apr) If the fieldwork really was 21-23 Apr, then we have had nine more recent polls since then, including the Survation released today. These are all in a tight range of Lab 13-18 points ahead. 10-poll average is Lab 15% ahead. |
|
|
|
Post by James E on May 3, 2023 13:49:10 GMT
( re "the Tories had made about half their gains, and on a steeply rising trend, before the war") The polls at the time don't support this idea of a 'steep gradient' at the time when the war started. The Conservative VI in polls around the start of the Fulklands War was: 28 Feb 1982 (Gallup) 34% 15 Mar 1982 (Mori) 31.5% 31 Mar 1982 (Gallup) 35% (+1) 2 Apr War begins 12 Apr 1982 (Mori) 31.5% (0) 14 Apr 1982 (Gallup) 33% (-2) It is from 21 Apr onwards, with a 36% with Mori, and then 44% on 30 Apr 1982 that there was a steep rise in the Tories' VI. Jeez James, yes you don’t see much effect over the course of a few weeks, esp. given polling variation. But if you take it from the trough the previous year, with polling in the twenties, then it’s a lot clearer. These polls cover 6 weeks and, as I say, look static.If you then look forward 6 weeks to May 1982 the Con VI really did rise steeply in a short time, to an average of 46%. |
|
|
|
Post by James E on May 3, 2023 13:35:20 GMT
c-a-r-f-r-e-w "There was indeed, though interestingly the Tories had made about half their gains, and on a steeply rising trend, before the war (which kinda matches with the fall in inflation)" The graph you have linked suggests that, but the polls from that time do not. Looking at the polling averages from the time, the Con VI rose by about 4 points in the 4 months before the Faulklands war and by about 14 points in the aftermath. ConVI Nov/Dec 1981 Av 28% ConVI Late Feb/Mar 1982 Av 32% ConVI July/Aug 1982 Av 46% en.wikipedia.org/wiki/Opinion_polling_for_the_1983_United_Kingdom_general_election Fair enough, I shall have another look James! EDIT: the only poll just before the war started was the Telegraph on 35%. You have to go back two weeks for one before that, which is a bit lower, and then another two weeks before that for a poll on 34%, which is perhaps why the graph is as it is. And the important thing is the gradient. There was a rising gradient in step with the inflation improvement, before Falklands kicked in. Though Falklands may well have contributed of course. The polls at the time don't support this idea of a 'steep gradient' at the time when the war started. The Conservative VI in polls around the start of the Fulklands War was: Five polls in Jan-mid Feb average Con 31% 28 Feb 1982 (Gallup) 34% 15 Mar 1982 (Mori) 31.5% 31 Mar 1982 (Gallup) 35% (+1) 2 Apr War begins 12 Apr 1982 (Mori) 31.5% (0) 14 Apr 1982 (Gallup) 33% (-2) It is from 21 Apr onwards, with a 36% with Mori, and then 44% on 30 Apr 1982 that there was a steep rise in the Tories' VI. |
|
|
|
Post by James E on May 3, 2023 12:00:25 GMT
Re Corbyn and the voters who deserted Labour in 2019. These were the YouGov findings at the time: yougov.co.uk/topics/politics/articles-reports/2019/12/23/their-own-words-why-voters-abandoned-labour The largest reason for not voting Labour again was 'Leadership/Corbyn' with 35%. 19% cited Brexit, two-thirds of them Leavers, and one-third Remainers. 16% cited 'economic competence'. To some extent, these were constant issues when Corbyn was Labour leader: in 5 years as leader, he never led either May or Johnson in a 'Best PM' poll; and Labour never had a lead on 'which party do you trust on the economy?' or similar questions. Yeah by 2019 there had been lots more time for the Blairite and media assault and for Tories, realising they were still behind despite the assault on Corbyn, to switch to BJ and position themselves more to the left. 2017 was the better chance since it caught some by surprise, to see left wing policies do better than expected, and they did wind up polling more than Tories for a while after the election All Leaders of the Opposition in living memory have leads in some polls - even the likes of William Hague, Ian Duncan Smith (link below) and Michael Foot. d25d2506sfb94s.cloudfront.net/today_uk_import/YG-Archives-pol-dTel-JulyTracker-030729.pdfMy point was that Corbyn never achieved any lead in the kind of polls that suggest an underlying lead - such as those on economic trust or 'Best PM'. The same was true for Ed Milliband. |
|
|
|
Post by James E on May 3, 2023 11:52:51 GMT
c-a-r-f-r-e-w "There was indeed, though interestingly the Tories had made about half their gains, and on a steeply rising trend, before the war (which kinda matches with the fall in inflation)" The graph you have linked suggests that, but the polls from that time do not. Looking at the polling averages from the time, the Con VI rose by about 4 points in the 4 months before the Faulklands war and by about 14 points in the aftermath. ConVI Nov/Dec 1981 Av 28% ConVI Late Feb/Mar 1982 Av 32% ConVI July/Aug 1982 Av 46% en.wikipedia.org/wiki/Opinion_polling_for_the_1983_United_Kingdom_general_election
|
|
|
|
Post by James E on May 3, 2023 9:23:19 GMT
Re Corbyn and the voters who deserted Labour in 2019. These were the YouGov findings at the time: yougov.co.uk/topics/politics/articles-reports/2019/12/23/their-own-words-why-voters-abandoned-labour The largest reason for not voting Labour again was 'Leadership/Corbyn' with 35%. 19% cited Brexit, two-thirds of them Leavers, and one-third Remainers. 16% cited 'economic competence'. To some extent, these were constant issues when Corbyn was Labour leader: in 5 years as leader, he never led either May or Johnson in a 'Best PM' poll; and Labour never had a lead on 'which party do you trust on the economy?' or similar questions. |
|
|
|
Post by James E on May 2, 2023 18:33:29 GMT
R&W's 30 April Red Wall compared to GE2019
Lab 48% (+10)
Con 30% (-17)
LD 8% (+3)
Ref 6% (-)
Green 5% (+4)
Swing Con>Lab 13.5%
This compares to a 14% swing per their latest GB poll with Labour 17 points ahead. So normal for this sample of constituencies.
Looking at the 40 constituencies polled, I think they would all go to Labour if this were replicated at the GE, although the most long-shot (Bassetlaw) requires a 13.8% swing.
|
|
|
|
Post by James E on May 1, 2023 21:08:53 GMT
davemR&W's regional cross-breaks are prone to huge fluctuations. If we are to believe the latest one, Labour lead by 57/24 in the South-West of England, but only by 34/30 in London. It's actually Lab 52%, SNP 26% in this one after excluding the DKs, but worth noting that last week's R&W had Lab in 3rd place in Scotland, one point behind the Tories.
|
|
|
|
Post by James E on May 1, 2023 16:54:25 GMT
Small but perceptable swing to Labour on the doorsteps as previous undecideds breaking better than typically. I would not be surprised if Labour's PNS was closer to it national poll %age than usual and if the Tories was a bit below. All about differential turnout with ABTs more determined (in aggregate) than Tories to vote. It's normal to dismiss any party's claims of 'encouraging responses on the doorstep' but as people here will know, your outlook is normally more pessimisitic for Labour's chances than many of us. A 'normal' Labour performance in English local elections is around 4-5 points lower than their average polling VI, and the Tories are generally a few points below theirs, too. So an 'expected' PNS for Thursday might be around Lab 39%, Con 26%, LD 17%. |
|
|
|
Post by James E on Apr 30, 2023 19:00:37 GMT
@rafwan I have edited my 15:04 post to show the YouGov churn figures postGE2019 combine with the 2017 vote shares to produce the 2019 result. If you are pondering my point about where the 2010 LDs went in GE2015, when the LDs were down by 15 points and UKIP up by 10 points - polling data suggested that only about 10% of the 2010LD vote (or 2.3% of those voting) switched to UKIP. Another one to consider: in 2017, the LibDems got 7.4% of the vote compared to 7.9% in 2015, so you might guess that they held on to 94% of their 2015 voters. But if you want to check the 2017 YouGov analysis linked here, it shows how many of their 2015 voters actually stayed with them (36%). yougov.co.uk/topics/politics/articles-reports/2017/06/22/how-did-2015-voters-cast-their-ballot-2017-generalThere is no way that you might work that out by looking at the net movements in percentage votes. EDIT I have removed my further responses, as it seems that Rafwan was simply trolling. Surely no-one really believes that a simple analysis of net gained/lost vote percentages can be used to determine churn? |
|
|
|
Post by James E on Apr 30, 2023 16:23:51 GMT
@rafwan YouGov use a panel of voters, so have access to people's previously-stated votes from earlier elections. This is part of the reason why their findings are considered by many to be more reliable than other pollsters. I don't understand why you claim that there is some kind of mis-match between YG's findings and the 'Actual' result from 2019 - of which we are all well aware. There isn't. I didn't know about the panel - yes that would make a difference. The actual results showed that: Between 2017 and 2019, Labour had a net loss of 7.8 points; Tories had a net gain of 1.3 points; LibDem was 4.2 up; SNP and Green were 2 points up between them. Anti-EU parties were net static. There is a net difference of 0.3 between those tallies (rounding? non-voters?). There is a hefty mismatch between these data and those you present. The net movements of support - of which we are all very well aware - cannot be used to calculate the 'churn' within any party's previous vote. Your approach reminds me of those people who claimed after the 2015 General Election that most of those who deserted the LibDems had switched to UKIP: after all, the LDs were down by 15 points, UKIP were up by 10, and Con, Lab were all up, too, so UKIP couldn't have gained support from them... There is no 'mismatch'. If you cannot understand the YouGov tables I have provided, or can't see that they do in fact match to the 2019 result, there is no point in any further discussion. EDIT - I have edited my post from over 4 hours ago, to show exactly how the YouGov churn-analysis data can be combined with the 2017 votes to produce the 2019 result. It is simply a matter of multiplying the various percentages, using the figures linked in my post at around 3pm. |
|
|
|
Post by James E on Apr 30, 2023 16:17:20 GMT
pjw1961 Indeed however the 33 was based on a vote share of around 9% It's currently around 11% and a good local elections performance might well focus some minds. LibDems got 11.6% in 2019 and 11 seats. Why would it be so different next time? The LibDems' performance in terms of the number seats depends on their performance relative to the Conservatives, rather than on their absolute vote share. Hence the LDs gained 28 seats in 1997 when they were down 1% in the popular vote, but lost a net 5 seats in 2010, when their vote share was up 1%. And of course in 2019, they gained 4% in the popular vote but landed up with one less seat than 2017. |
|
|
|
Post by James E on Apr 30, 2023 14:54:28 GMT
@rafwan
YouGov use a panel of voters, so have access to people's previously-stated votes from earlier elections. This is part of the reason why their findings are considered by many to be more reliable than other pollsters.
I don't understand why you claim that there is some kind of mis-match between YG's findings and the 'Actual' result from 2019 - of which we are all well aware. There isn't.
|
|
|
|
Post by James E on Apr 30, 2023 14:04:31 GMT
James E I simply cannot square those two sets of data. But which would you trust more? Actual data from actual voting? Or single (heavily weighted) opinion survey data relying on recall of events over two or three years before? The YouGov 2019 election analysis which I linked above was based on a poll of 40,000 respondents carried out in the week after the election. It was published on 17 December 2019.Data table are in the second link, below. Take a look at the diagram two-thirds of the way down the page of my first link, which maps the movements in all directions between GE2017 and GE2019. (Headed 'How 2017 voters voted at the 2019 general election). This sets out clearly all the movements, and, of course, matches to the actual result. yougov.co.uk/topics/politics/articles-reports/2019/12/17/how-britain-voted-2019-general-electiond25d2506sfb94s.cloudfront.net/cumulus_uploads/document/wl0r2q1sm4/Results_HowBritainVoted_2019_w.pdfFrom these tables, it looks like the Labour losses from GE2017 went equally to pro-Brexit and anti-Brexit parties. YG's figures showed that Labour held onto 72% of their 2017 vote - compared to 85% 2017Con voters who stuck with the Tories. A combined 13% of 2017Lab voters went to Con (11%) and BXP (2%), while another 13% went to the LDs (9%), SNP (2%) and Greens (2%). EDIT - 17:25hrs Rather than engaging in further back-and-forth posts, here are the YouGov Dec 2019 findings as detailed in the link above more than 4 hours ago, combined with 2017 results to produce 2019 levels of support: LibDem 59% x 8% LD2017 = 4.7 9% x 41% Lab 2017 = 3.7 7% x 43% Con 2017 = 3 Total 11.4% in 2019 Labour 72% x 41% Lab 2017 =29.5 20% x 8% LD2017 = 2 4% x 43% Con 2017 = 1.7 34% x 1.5% Green 2017 = 0.5 Total 33.7% in 2019 Conservative 85% x 43% Con2017 = 36.5 11% x 41% Lab 2017 = 4.5 14% x 8% LD2017 = 1.1 67% x 1.8% UKIP2017 = 1.2 Total 43.3% in 2019 |
|
|
|
Post by James E on Apr 30, 2023 12:37:02 GMT
Is part of that down to the unusually high number of 2017 Labour voters voting Conservatives in 2019? Between 2017 and 2019, Labour shed 7.8 percentage points of support. The Tories gained 1.3 points, and the rest went to other parties, overwhelmingly pro-EU. YouGov's GE2019 analysis presents a different picture to that. They found that Labour lost 11% of their 2017 support to the Tories, 9% to the LDs, and a slightly lower proportion to others (a mixture of Green and Brexit Party probably). yougov.co.uk/topics/politics/articles-reports/2019/12/17/how-britain-voted-2019-general-election |
|
|
|
Post by James E on Apr 30, 2023 10:47:01 GMT
PJW, as Opinium in effect reallocate those DKs and WVs to the proportions that did give a VI, that Lab 16% is from 76% - so over 21% of Tory 2019 voters switching to Lab on their methodology which as you imply accounts for the surprising headline numbers. Think Con 2019- Lab is typically above 10% I and was around 14/5% at its peak. NB) Pedant Alert, Graham Thrasher says 700 seats would be a good night for Labour, expectations would be a lower, perhaps closer to me 'decent' 600. Yes, their figures for Con 2019 without 'don't knows' become Con 59%, Lab 20% - which doesn't quite pass the smell test. Likely a bit of a rogue with a reversion in the next poll (which we also shouldn't get too excited about). For comparison to other polls: The previous two Opiniums showed 14% and 17% of Con2019 voters switching to Lab. YouGov, whose fieldwork findings are often similar to Opinium, show 14%, 12% and 17% Con2019>Lab in the 3 most recent polls (with leads of 18, 15 and 14), after excluding Don't Knows. Both YG and Opinium use panels, and because of this they tend to show higher churn than other pollsters who ask respondents to recall their 2019 vote (e.g R&W). I'd suggest that 15% Con>Lab is normal for the panel pollsters and 10% for non-panel. |
|
|
|
Post by James E on Apr 29, 2023 18:41:05 GMT
An 18-point labour lead with Opinium is a really surprising - perhaps just an outlier? Their re-weighting methodology has previously put them around 5-7 points lower than other pollsters. Now they show the highest Lab-lead of all recent polls.
I would have expected a lead of around 9 points, given that the latest youGov showed Laabour 14% ahead.
|
|
|
|
Post by James E on Apr 28, 2023 15:36:44 GMT
...and this is the result of that Omnisis Local Elections poll:
Local Elections Voting Intention:
LAB: 37% (+10)
CON: 26% (-5)
LDM: 17% (=)
GRN: 9% (=)
IND: 6% (-5)
RFM: 4% (-1)*
Via
@omnisis
, 27-28 Apr.
Changes w/ LE2019.
*Changes w/ UKIP share.
3:38 PM · Apr 28, 2023
These figures look credible enough to me. I recently did some analysis of Local Election 'Net equivalent vote' shares for the past decade, and the norm seems to be that Labour get 5% or so less than their General Election VI, the Tories a few points below theirs, too, and the LDs exceed their GE VI by around 7 points.
|
|
