|
|
Post by James E on Apr 23, 2022 20:23:46 GMT
Opinium feels a bit like an outlier to me. I know it's "new money" but mostly their new money was taking them closer to the averages of other pollsters. Since their 'new methodology' was introduced, Opinium's figures have been very steady, with Labour leads of 3,2,4.2,4and 2. So to me this looks like polldrums. And presumably it would be about an 8-9 point lead under their old methodology, which produced results around 6-7 points better for Labour. |
|
|
|
Post by James E on Apr 23, 2022 10:48:33 GMT
jibThe Telegraph article refers to Tory support 'collapsing' to 38% in 'Red Wall' seats, so it would seem likely that this refers to seats that they won in 2019.
|
|
|
|
Post by James E on Apr 23, 2022 10:36:50 GMT
These need to be taken with a pinch of salt, as they are based on small and unweighted sub-samples. But for what they're worth, these are the average of the last 4 R&W 'regional' sub-samples, compared to GE2019.
SE. Con 40& (-10), Lab 32% (+10)
Lon Con 28% (-4), Lab 45% (-1)
SW: Con 34% (-19), Lab 32% (-9)
WMids Con 43% (-10), Lab 37% (+3)
East Con 46% (-11), Lab 39% (+14)
NE Con 27% (-11), Lab 50% (+7)
NW Con 28% (-10), Lab 52% (+6)
Yorks Con 30% (-13), Lab 50% (+11)
Scot Con 22% (-3), Lab 21% (+2)
Wales Con 26% (-10), Lab 40% ( 0)
Taken with other evidence, such as the YouGov sub-samples, and polling of Scotland and London, it seems clear that Labour are making progress in 'England-outside-London', particuarly in the South. Not much progress in Scotland, Wales is a bit uncertain, and probably only a little further progress in London, where Labour had a 16-point lead and most of the seats at the last GE.
|
|
|
|
Post by James E on Apr 23, 2022 10:17:12 GMT
re the 'More In Common' poll for the Telegraph showing Tory support in the 'Red Wall' falling to 38%.
It would be really useful to know what they mean by 'Red Wall' and what the comparitor is for the 38% - the Tories are, after all, only polling at around 33-34% in recent GB polls.
And there is other polling evidence (mostly from cross-breaks) which seems to me to show a swing of around 8-9% in the North of England, and much the same in the Midlands. It would be interesting to know if the More in Common findings show any more than this.
|
|
|
|
Post by James E on Apr 22, 2022 11:55:21 GMT
Damning figures for Johnson from YouGov.
"Thinking about how he has responded to the issue of parties being held at Downing Street during lockdown, do you think Boris Johnson has or has not lied?"
Has lied 78%
Has not lied 8%
DK 14%
2019 Con voters divide by 17-61 for 'Has lied', and even current Tory supporters are 25/51 for the same.
As another observation - I do think it's worth keeping an eye on the Green and RefUK VI in YouGov in particular; this poll puts the Greens on 8% and Ref on just 3%. I suspect that both would be squeezed in an actual GE. This would (IMO) be likely to offset the differential 'Don't Knows' from 2019 voters, which are 20% for Con and 9% for Lab. An ICM 50% reversion of these would give the Tories a net boost of 3 points.
|
|
|
|
Post by James E on Apr 22, 2022 8:47:31 GMT
As ever with Techne, their GE2019 cross-breaks look implausible. They show just 2% of Con2019 switching to Lab, and 0% the other way. They do show some LD>Lab movement (29%), but these cross-breaks cannot explain the overall 10% Con>Lab swing from GE2019.
The other oddity of Techne (to me) is just how stable their figures are. Every single poll they have released seems to be right in line with the pack. Surely all pollsters will get the occasional outlier - or at least a bit of random variation?
|
|
|
|
Post by James E on Apr 20, 2022 9:30:17 GMT
sheviiThanks for the comparative figures above from Ipsos in April 1992. One thing that should be added to those is that their headline VI figures are Con 35, Lab 43 - so they were some 15 points adrift from the actual result. So for example, where the poll shows the Tories more trusted on the economy by 36/31, it is fair to asume that they really had a far larger lead.
|
|
|
|
Post by James E on Apr 18, 2022 19:55:28 GMT
James re ''The Tories' vote is now more efficiently spread than Labour's in England, too. looking at the last 3 GEs, it seems that lab now need to be around 2 points ahead to be level in seats, and probably 7-8 points to get a majority.'' More a suggestion than a challenge but I wonder if UNS understates the winners bonus as any LTV advantage, and tactical voting, is more likely in marginals. I would not be surprised if a lower Lab margin over the Tories GE than 7-8% delivered an OM; or more likely similar vote shares a produced similar number of sets. I'd agree with your first point: if Labour has a significant lead in a GE, the effects of (anti-Tory) tactical voting are greater. This was a large part of the reason why the Tories were reduced to so few seats in 1997 and 2001. But in the event of something like 39/39 between Con and Lab, the Tories would (IMO) still be the largest party. UNS sugests that they would do so by around 20-30 seats. |
|
|
|
Post by James E on Apr 18, 2022 17:26:47 GMT
I suspect the smaller swing in London is down to the fact Labour is already dominant there and so it is more difficult to pick up even more votes. Indeed. But in any case, the 3 London polls we have had this year from YouGov all show a sizable further advance on the 16% Lab lead at GE2019, with leads of 28,32 and 32 points. These are probably right at the top end of what Lab might be able to achieve. en.wikipedia.org/wiki/Opinion_polling_for_the_next_United_Kingdom_general_election#London |
|
|
|
Post by James E on Apr 18, 2022 17:11:36 GMT
guymonde Its's London sub-sample of just 174, so I would not pay a lot of attention to that. My own analysis of YouGov's London sub-samples suggests a modest swing to Labour in London compared to GE2019. A comparison of the last 3 YouGovs, compared to GE2019 show the following 'regional' average swings London 3.5% South England 11% Midlands + Wales 9% North England 7.5% Scotland 3.5% By Social Class, it's 8.5% Con>Lab among ABC1s and 6.5% among C2DEs Probably not quite what Labour strategists would want, as there are relatively fewer marginals in the South of England. But worth noting that the Midlands swing is still above the average, and the North of England is exactly in line with the overall GB average.
|
|
|
|
Post by James E on Apr 18, 2022 16:14:33 GMT
As JimJam has pointed out, that 8% lead with R&W is welcome news for Labour, given that the previous R&W looked rather too high. The 7 most recent polls now average a 7 point Lab lead. birdseye - It isn't just the rise of the SNP which has affected they way in which Con and Lab translate votes into seats at Westminster. The Tories' vote is now more efficiently spread than Labour's in England, too. looking at the last 3 GEs, it seems that lab now need to be around 2 points ahead to be level in seats, and probably 7-8 points to get a majority. |
|
|
|
Post by James E on Apr 18, 2022 15:54:45 GMT
Be interesting to see if the Rwanda situation will have an effect. Last week's Yougov had leave voters with 20% D/K, 10% WNV and 3% refused so could be a plus for Conservatives. Those figures are fairly typical for YouGov, who always seem to pick up a lot of Leavers who say they won't vote in a GE. Their previous three polls had 10, 10 and 9% for Leavers who say they would not vote, and 21,17 and 21% who don't know which way they will vote. |
|
|
|
Post by James E on Apr 17, 2022 18:53:48 GMT
Moving on further from that, what happens if the leaders of a GE winning party don't actually win their own seats? Has it ever happened? An example of something of the sort would be Patrick Gordon Walker in October 1964. He was expected to be the Foreign Secretary in Harold Wilson's incoming Labour administration but lost his seat in Smethwick against the tide, on the back of an extremely racist campaign by the Conservative candidate (slogan "If you want a n***** neighbour, vote Labour"). Wilson, in an unusual move, appointed Gordon Walker as Foreign Secretary anyway, even though he was not in parliament. The Labour MP for Leyton was pressured to stand down and given a peerage to create a by-election (by all accounts he was not very happy about it). The by-election was held in January 1965 and the local electors, annoyed at being used in this way, returned the Conservative candidate. Gordon Walker resigned as Foreign Secretary. Gordon Walker went on to win Leyton in the 1966 GE, but by then his moment had passed and although Wilson did bring him back into the cabinet it was in more junior posts. Fair point about Gordon Walker in 1964, but i think that short answer to SDA's question is 'not yet'. It is just about conceivable, though that it could happen at the next GE if johnson is still Tory leader. As I've mentioned before, Johnson's Uxbridge constituency has been trending to Labour at successive General Elections as part of the wider trend in London. As such, it might just be possible to get the required 7.5% there in the context of a Con>Lab GB swing of around 4%, leaving the Tories with around 320-325 seats, but Johnson out of Parliament. I'm not quite sure what would happen, but I guess the Tories would then need to quickly nominate another MP as their preferred PM. |
|
|
|
Post by James E on Apr 9, 2022 19:18:55 GMT
I don't think that Opinium are providing comparative figures any more, but when they did the difference that their new methodolgy made was to take 6-7 points off the Labour lead. So this might have been a 10-11 point lead a few months ago..
Sunak's personal rating has slumped to -15.
|
|
|
|
Post by James E on Apr 9, 2022 16:56:42 GMT
Good early afternoon from the seats of Ellwood and Kinnock. On the long drive between my seats my wife read to me the piece by Mr Parris who states today that Tony Benn, who became the inspiring voice of the new Left from early 1971 onwards, legally avoided the inheritance tax trap while handing on his not inconsiderable wealth to his children. Does anyone know if this calumny is correct? 1. How can a 'calumny' be correct? 2. Are you aware of google? A 2 minute search reveals that Benn settled a potential £210,000 Inheritance tax* bill by making a donation of artifacts worth around £500,000 www.todayswillsandprobate.co.uk/tony-benns-estate-avoids-inheritance-tax-using-acceptance-lieu/* for comparison, this sum is 1/10th of the sum that Murty avoids every year through non-dom status. And her potential Inheritance Tax bill is around £280 million. |
|
|
|
Post by James E on Apr 6, 2022 18:13:03 GMT
@mark ….. their arms around teenage girls - I don't just mean Saville - all of them. Nobody batted an eyelid at the time, but watching it now you think "wow....creepy much?". Yes, you’re right - but I have never ever used that odd turn of phrase to describe my feelings. I am fascinated by the ways in such language can become popular and why some people use it and most don’t. I sort of wonder about such things (as with the now ubiquitous “absolutely” as the correct response to “would you like a cup of coffee?”) who said it first and why did it take off? I think it's a bit older than 'AbFab' (which was first broadcast in Nov 1992). Reading this reminded me of the wonderful comedy series 'Absolutely', which ran from 1989-93. Stoneybridge Town Council was a particuar favourite. www.comedy.co.uk/tv/absolutely/videos/7765/macglashan_inventions/ |
|
|
|
Post by James E on Apr 5, 2022 9:37:03 GMT
Savanta ComRes @savantacomres · 9m 🚨New Westminster Voting Intention🚨 📈Labour 7pt lead 🌳Con 33 (-2) 🌹Lab 40 (+1) 🔶LD 11 (=) 🎗️SNP 5 (=) 🌍Green 4 (+1) ⬜️Other 8 +1) 2,220 UK adults, 1-3 April We now have 4 of the last 5 polls showing a somewhat wider Lab lead (averaging around 5-6 points), with Techne being the only exception. en.wikipedia.org/wiki/Opinion_polling_for_the_next_United_Kingdom_general_electionI doubt that this is the return of partygate, but more likely cost-of-living issues, and also a reversion after the bounce Johnson got in the early weeks of Russia's war in Ukraine. |
|
|
|
Post by James E on Apr 2, 2022 8:32:32 GMT
Some interesting discussion above on Local Elections and the French Presidential Election.
Thanks to all who contributed to the that - it's just the kind of thing which makes UKPR2 worth reading.
|
|
|
|
Post by James E on Mar 31, 2022 17:00:09 GMT
...in October 1964 a very small lead over the Tories in the House, due mainly (GRAHAM may confirm) that the Liberal Party had a big increase of votes and helped Lab to win. Labour's popular vote hardly shifted at all from 1959-64. What you fail to mention is that the the Liberal Party stood in 149 more seats in GE1964 than in GE1959. In GE1959, they achieved 5.9% of the vote, but this was actually 17.2% in the 216 seats where they fielded candidates. Standing in more seats would have brought the Liberals around 4% of the 5.3% increased vote they achieved at GE1964. In this context, Labour's overall increased vote share was a useful achievement. en.wikipedia.org/wiki/1964_United_Kingdom_general_election |
|
|
|
Post by James E on Mar 31, 2022 10:04:02 GMT
Some interesting figures from Opinium's most recent poll to the question: "Has brexit gone better or worse than you expected?' Responses from Remainers are predictably negative, but the answers from Leavers (sample of 845) are more revealing. They divide as follows: Better than expected 22% Worse than expected 28% I expected it to go badly and it has 16% I expected it to go well and it has 18% Don't Know 16% So overall, a larger number think it is going 'worse' or 'badly' (44%) than 'better' or 'well' (40%). But the figure that really intrigues me is the 16% who now state 'I expected it to go badly and it has'. Some on UKPR may remember that I've highlighted some exensive pre-referendum polling by YouGov (taken in the week before the referendum) showing that just 4% of Leave supporters expected negative economic consequences and only 3% thought it would make them worse off. If we take YouGov's contemporaneous polling as a guide, it looks like the around three-quarters of those Leavers who now say they 'expected it to go badly' are suffering from some kind of false recall. And there is an interesting comparison between the proportion of Leavers who thought Brexit would make Britain ecenomically better-off (48%) and the proportion who see their own expectation as being fulflled (18%). d25d2506sfb94s.cloudfront.net/cumulus_uploads/document/atmwrgevvj/TimesResults_160622_EVEOFPOLL.pdf |
|
|
|
Post by James E on Mar 30, 2022 18:32:17 GMT
|
|
|
|
Post by James E on Mar 30, 2022 11:14:19 GMT
@ NeilJ - re your 11:59am
It seems that despite an underwhelming 3-4 point polling lead, Labour now have the advantage on a number of polling indicators, on things such as trust and leadership. In some ways, this is the reverse of the position in the 2010-15 (or 1987-92) Parliament, where there were big polling leads for Labour, but negative indicators on the wider questions.
|
|
|
|
Post by James E on Mar 28, 2022 20:47:01 GMT
A couple of things to bear in mind re 'Swingback'. 1. Isn't the 2017 General Election the best example of when 'swingback' did not apply? 2. Some current polls - most notably Opinium's - are now constructed in such as way that the opposite may apply. For those who missed it, this is how they described the effect of their new methodology, 2 months ago, with changes which took 6-7 points off the Labour lead. (On a like-with-like basis, Opinium show very little movement since January.) "However, it will mean a shift is required in the way our polls are analysed. For the past few months people have generally considered the high number of Conservative don’t knows to be a caveat against the high Labour poll leads. In our case, the opposite is now true. The existence of the high number of 2019 Conservative undecided voters is more likely to increase the Labour lead at a later date rather than decrease it." www.opinium.com/resource-center/uk-voting-intention-27th-january-2022-2/ |
|
|
|
Post by James E on Mar 27, 2022 10:26:35 GMT
As far as I can see, Opinium have not yet published tables, nor comparative figures to shw the effect of their new methodology. However, as others have pointed out, this made about a 6-point difference in their first 3 'new' polls, so it's reasonable to guess that they might have shown a Labour lead of 8 points. Prior to their change, Opinium were consistently producing the 'best' results for Labour, and in fact their first 'new' poll turned a 10-point lab lead to one of just 3 points. On this basis, they have shown very little movement in the past 2 months. |
|
|
|
Post by James E on Mar 18, 2022 19:34:58 GMT
TechneUK appear to weight back to 52/48% leave/remain which would not be correct now as, sadly for their families, many more Leave voters would have passed away by now. It should be around 50.5/49.5% now. Parity between living 2016 Remainers and living 2016 leavers is likely to happen towards the end of 2024. |
|
|
|
Post by James E on Mar 18, 2022 19:11:49 GMT
TechneUK in 4 polls, I think it is, have had Tory 2019-Lab as 1-3% whereas the lowest on any other pollster this year that I can recall is 7% and typically it is higher ... YG tend to show quite high DKs compared to others and TechneUK show extremely low %s (eg 2% of CON'19 and 0% of LAB'19 are 'uncertain'?) It looks to me like Techne follow a similar methodology to Survation in asking first about likelyhood to vote, and only including those who say they are 'likely' in their tables. They then describe those who say that they would vote, but are not sure who they's vote for, as 'uncertain'. It's somewhat different to the 'Don't Knows' category for most pollsters. |
|
|
|
Post by James E on Mar 18, 2022 18:11:21 GMT
|
|
|
|
Post by James E on Mar 18, 2022 18:02:02 GMT
@jimjam
Those Techne figures for VI by GE2019 vote are very odd, and show staggeringly high levels of voter-retention.
Given the much-lower levels of voter retention found by the likes of YouGov and Opinium who have panels and asked the VI question soon after GE2019, I would be inclined to largely disregard Techne.
|
|
|
|
Post by James E on Mar 7, 2022 20:33:01 GMT
James E It’s an established fact that once parties get within 2-3% the MoE means that cross over can take place at anytime. Of course we all know that MoE can occur with any poll but we don’t usually mention it until the polls close enough for cross over to take place. Hope that clears it up for you. You clearly don't know what 'Margin of error' means, and seem to be incapable of learning. Can I ask if you accept that 'Margin of Error' applies to an individual poll or sub-smaple? There is no such thing as the collective margin of error of polls. |
|
|
|
Post by James E on Mar 7, 2022 19:29:20 GMT
|
|
