|
|
Post by James E on Oct 3, 2023 18:04:38 GMT
I wondered who might leap on that. Attack the source but ignore the data Interesting use of the word attack. All I have done is literally provide links to examples of his previous work, without commenting.... Is this an attack on Daniel Hannon?... reaction.life/britain-looks-like-brexit/ |
|
|
|
Post by James E on Oct 3, 2023 17:22:03 GMT
|
|
|
|
Post by James E on Oct 3, 2023 16:54:56 GMT
I'm not trying to make some snide comment about the title of this thread (honest) but I do find it quite amusing the way some Torys have jumped on this Opinium poll like it is the start of a new era. Take for example the Tory MP on Westminster Hour on Sunday. What I found even more incredible was that the pundit agreed that it was remarkable Labour had gone down below 40% for the first time in months. Have these people never heard of averages and outliers? Or are they just pretending to be ignorant, because that's better than the alternative? And what's more, let's just stand back for a moment and think about the fact that the Tories are celebrating a poll where they are still 10 points behind Labour. There's some very wishful thinking going on in some quarters. It also shows ignorance of Opinium's post Jan 2022 methodolody. Opinium's fieldwork findings of Lab 42%, Con 26% were nearly identical to YouGov's Lab 43, Con 27 of 22 Sept. These two polls reflect the same underlying position, but have different headline figures due to Opinium's re-weighting. Perhaps the most surprising thing is that, despite what is normally around a 6 or 7-point reduction to the Labour lead, Opinium's polls rarely look like outliers. |
|
|
|
Post by James E on Oct 2, 2023 19:00:59 GMT
As is my wont, Just For Fun, I plugged those headline VI figures into Electoral Calculus, using the new boundaries, and got a LAB OM of 66. I did the same, but EC's figures for Scotland rather than 'SNP 3%', and got a 56 Lab majority. But if you add in Tactical Voting at 50% of Lab, LD and Greens, this magically turns into a majority of 128. (Con 29, Lab 39, LD 12, Grn 7 Ruk 7) www.electoralcalculus.co.uk/fcgi-bin/usercode.py?scotcontrol=Y&tvcontrol=Y&CON=29&LAB=39&LIB=12&Reform=7&Green=7&UKIP=&TVCON=&TVLAB=50&TVLIB=50&TVReform=&TVGreen=50&TVUKIP=&SCOTCON=16&SCOTLAB=32.3&SCOTLIB=7&SCOTReform=2.1&SCOTGreen=3.4&SCOTUKIP=&SCOTNAT=36.7&display=AllChanged®orseat=%28none%29&boundary=2019nbbase |
|
|
|
Post by James E on Oct 2, 2023 18:40:02 GMT
A little milestone was reached a few days ago: it is now more than 12 months since any poll showed a Labour lead of less than 10 points. The last such poll was from Kantar with fieldwork of 22-26 Sept 2022, and was Lab 39, Con 35. Just a few days later (28-29 Sept) , we had Lab 54, Con 21 with YouGov.
Looking back at previous Leaders of the Opposition, neither Jeremy Corbyn nor Ed Milliband ever managed even an entire month of 10%+ leads - although the latter came close at times. And even David Cameron managed to get "all 10%+" leads for just 4 consecutive months, from May-Aug 2008, and later another 3 months in June-Aug 2009.
|
|
|
|
Post by James E on Oct 2, 2023 10:46:32 GMT
Taking the 5 most recent polls, there does seem to be a bit of narrowing.
I have compared each of the most recent 5 polls to the same pollster's previous average during August and September.
28-29 Sep We Think Lab lead 20% (v 19.8 in 8 polls):+0.2
27-29 Sept Opinium Lab lead 10% (v 14.5 in 4 polls):-4.5
26-27 Sept YouGov Lab lead 21% (v 19.6 in 8 polls):+1.4
26-27 Sept Techne Lab lead 18% (v20.2 in 5 polls):-2.2
22-25 Sept Deltapoll Lab lead 16% (v20.2 in 6 polls):-4.2
So overall, with this set of pollsters, the averaged lead is down 2 points from 19% to 17%.
Incidentally, with Opinium having published their latest tables, their unadjusted figures show Lab 42%, Con 26%, so the pattern of the (new) methodology reducing about 6 points from the Labour lead continues. And their methodology also again boosts Reform UK - in this poll from 5.5 to 7%.
|
|
|
|
Post by James E on Sept 30, 2023 22:48:29 GMT
The previous Opinium poll two weeks ago was 41/26 per their methodology, but would have been 45/23 witout re-weighting. So a net 7-point 'adjustment.
We don't yet have their tables, but a 5-7 point adjustment is normal.
Incidentally, the 'Don't Knows' are not re-allocated on the basis of their past vote; the positive responses of each demographic group are re-weighted to reflect the turnout of GE2019, to counteract the disproportionate number of DKs in some group (esp Con 2019) . This is why Opinium often show high VI for Reform UK as well as the Conservatives.
|
|
|
|
Post by James E on Sept 28, 2023 23:01:25 GMT
These figures from James E are a reminder that class based politics is well and truly dead. ABC1 Lab 47% (+14) Con 23% (-20) Con2019 ret 54% C2DE Lab 41% (+8) Con 28% (-20) Con2019 Ret 58% Is it? I think those figures show that the posh Labour people are crushing the poor working-class Tories under their well-healed jackboots. 🤣 Social Class cross-breaks and analysis vary massively from one pollster to another. Most pollsters other than YouGov show Labour faring better with C2DE voters - for instance the Deltapoll linked below is Con 33, Lab 40 wih ABC1s, and Con 24, Lab 55 with C2DEs. deltapoll.co.uk/wp-content/uploads/2023/08/Deltapoll-230829_trackers.pdfFor the purposes of my own analysis above, I'm comparing YouGov's own 2019 election analysis with their current polls. Incidentally, where YouGov, and other pollsters, divide voters up into 4 social groups (AB, C1, C2, DE), the Conservatives' strongest demographic is with the C2s, and Labour's strongest the C1s. |
|
|
|
Post by James E on Sept 28, 2023 20:41:27 GMT
A small landmark for my (and YouGov's) favourite tracker question - "In hindsight, do you think Britain was right or wrong to vote to leave the EU?". This week's YG poll marks the 250th time they have asked the same question, going back to 2 August 2016, when a net 4% thought it was 'right' (46/42) totally in line with the result of the referendum itself. These days, though, the typical result is a net 23% for 'wrong' (32/55), as seen in the poll released today, and several others recently. The YouGov poll of 26-27 Sept also completes the ninth successive calendar quarter in which the majority for 'Wrong to Leave' has grown. Here are the quarter-by-quarter averaged net results and movements for the past 3 years. Jan-Mar 21 (11 polls) Net 'Wrong' : 4.6% Apr-June 21 (8 polls) Net 'Wrong': 2.5% (-2.1) July-Sept 21 (7 polls) Net 'Wrong': 6.4% (+3.9) Oct-Dec 21 (8 polls) Net 'Wrong' : 9.8% (+3.4) Jan-Mar 22 (5 polls) Net Wrong: 10.6% (+0.8) Apr-June 22 (7 polls) Net Wrong: 11.2% (+0.6) July-Sept 22 (8 polls) Net Wrong: 15.5% (+4.3) Oct-Dec 22 (10 polls) Net Wrong: 18.9% (+3.4) Jan-Mar 23 (8 polls) Net Wrong: 20.6% (+1.7) Apr-June 23 (10 polls) Net Wrong: 22.1% (+1.5) July-Sept 23 (8 polls) Net Wrong: 23.4% (+1.3) Most of the movements we have seen to this tracker reflect the changing narrative around Brexit. There was a 9.8% lead for 'Wrong' in the 4th quarter of 2020, but this was reduced to just a few points in the 'vacccine-bounce' of mid-2021. By the end of 2021, that 9 point lead was back again as 'partygate' broke. And the growth in the 'Wrong' lead in the final quarter of 2022 matches up with the chaos of the Truss era - although the impact on the 'hindsight' question was far less dramatic than on Voting Intention polls. To me, the most interesting thing is what has happened since that 4th quarter of 2022. The lead for 'Wrong' has continued to grow steadily -by a total of 4.5%. This has happened during a period when the Conservatives have reduced their polling defecit by about 5 points, and the signing of the 'Windsor framework'. And with support for Brexit now very concentrated into the 65+ age-group (who have moved far less than all other groups) it is likely that age-demographics alone will continue to move opinion towards 'Wrong to Leave' - this in itself is adding a little over 1% per annum to 'Wrong'. www.whatukthinks.org/eu/questions/in-highsight-do-you-think-britain-was-right-or-wrong-to-vote-to-leave-the-eu/?pollster%5B0%5D=yougov |
|
|
|
Post by James E on Sept 28, 2023 16:28:08 GMT
With YouGov finally up to date with their tables, I've calculated my usual cross-break analysis for their past 10 polls stretching back to 26 July. Overall, their polls have averaged Lab 45%, Con 25% - which is just short of a 16% swing on 2019. I have also analysed each cross break by the Conservatives' Vote Retention from GE2019 ('Con2019 Ret') for which the overall average is 56% (25/44.7). LondonLab 55% (+7) Con 19% (-13) Swing 10% Con2019 Ret 59% South EnglandLab 38% (+15) Con 30% (-25) Swing 20% Con2019 Ret 55% MidlandsLab 44% (+11) Con 31% (-24) Swing 17.5% Con2019 Ret 58% North EnglandLab 54% (+11) Con 22% (-17) Swing 14% Con2019 Ret 56% RemainLab 57% (+8) Con 13% (-6) Swing 7% Con2019 Ret 68% LeaveLab 24% (+10) Con 44% (-30) Swing 20% Con2019 Ret 59% ABC1Lab 47% (+14) Con 23% (-20) Swing 17% Con2019 ret 54% C2DELab 41% (+8) Con 28% (-20) Swing 14% Con2019 Ret 58% I have not anlysed the Scotland or Wales cross-breaks, as their is better data in full Scottish and Welsh polls. The two full YG Westminster polls for each of those covering the same period show the Tories getting a 2019 Vote retention of 62% in Scotland (15.5/25), and 53% in Wales (19/36). However, note that the high Tory losses do not always correlate with the larger Lab gains in vote share - Wales is a good example of this due to PC and RefUK receiving increased support. So it's a fairly consistent picture of proportionate vote loss for the Tories, and in fact their vote seems to be holding up better in their weak demographics - Remain voters, London, Scotland. This is the opposite effect to Uniform National Swing, and explains why those MRP projections show them losing so many seats. Link to their tables is below: this includes at least one (11 Aug) which isn't shown in the Wikipedia polls page as it was published so late. yougov.co.uk/topics/politics/explore/issue/Voting_Intention?content=surveys |
|
|
|
Post by James E on Sept 28, 2023 15:43:14 GMT
Yougov poll, always liked them 😀 Westminster Voting Intention: LAB: 45% (+2) CON: 24% (-3) LDM: 11% (+1) RFM: 8% (=) GRN: 7% (=) SNP: 3% (=) Via @yougov , 26-27 Sep. Changes w/ 21-22 Sep. They have now published the tables on time - something they had been neglecting to do in recent months. d3nkl3psvxxpe9.cloudfront.net/documents/TheTimes_VI_HS2_230927_W_yzXCaLz.pdfThe Brexit hindsight tracker works out as the usual 63/37 for 'Wrong to leave' (excl Don't Knows). People support the HS2 project by 47/33, but think it's bad value for money by 15/58 |
|
|
|
Post by James E on Sept 28, 2023 14:39:29 GMT
Yougov poll, always liked them 😀 Westminster Voting Intention: . W ell a year out from the likely date of the next GE and Lab's lead over the Tories looks to be fairly stable in the high teens. A double digit % lead will ensure an OM (I am not aware of GE in the UK where that has not been the case). Currently, the question does look like what size of OM Starmer will have. I doubt, given the political landscape, it would be as large as the ones Blair achieved in '97 and '01.
I was wondering if the Tory tactic of going on the attack over environmental policies was tapping into something in the electorate, but perhaps not. While the electorate as a whole don't seem to have become besotted with Starmer, given the choice between him and Sunak, currently they seem to have opted for Starmer.The highest lead any party has achieved the past 100 years without getting an overall majority was the Conservatives' 7% lead in 2010. If you apply UNS to the 2019 results, they show Labour needing to be around 11-12% ahead for an OM, but it should be clear from all the detailed data highlighted on UKPR2 that the real target is certain to be lower. Yesterday's Survation MRP with Labour lead of 46/29 provides some sort of a guide as to how this might work out: if you apply a 3.5% 'swing-back' to the Tories and adjust for the norm of each 1% of Con/Lab movement flipping around 15 seats, then this suggests that a 10-point Lab lead might now result in around an 80-90 seat majority. If you extend this further, with Labour 17 points ahead and 101 seats 'over the line' at 426 (per Survation), the Tories would need around 6-7% swing-back to deprive them of a majority. This would mean Labour being around 3-4 points ahead. |
|
|
|
Post by James E on Sept 27, 2023 17:09:17 GMT
Survation polling predicts a majority of over 100 for LabourI still think they have the Lib Dems too low (typical for Survation) as John Curtice said at our Party Conference he expected 34 Lib Dem MPs after the election and he only thought we would be that low because people still were not sure what we stood for. EDIT: It looks like this is the Survation polling for Greenpeace (fieldwork 18th August-1st September) where they have delayed release of the MRP results. It shows a Labour majority of 202 - which is of course 'over 100'. Seat totals are Lab 426, Con 141, LD 25, SNP 36. Vote shares are Lab 46%, Con 29%, LD 11%, SNP 3%, Grn 3%, Ref 4%. Overall sample was just over 20,000, so this does look like a worthwhile MRP, rather than just a projection added to a standard-sample poll of 1,000 to 2,000. The seat figures are interesting. Firsly, note that they still show the SNP dominant in Scotland, though they make a net loss of 12 seats. So these figures show a Labour landslide in England and Wales, but only a slight Labour recovery in Scotland. The Lab 426 is also far higher than UNS would imply. Using a 14.35% swing against the Tories and an estimated 10% against SNP would bring Labour just 355 seats per UNS on the 2019 boundaries - and most likely a similar number on the new ones. So this shows Labour actually gaining 71 more seats than UNS would produce, turning an estimated majority of 60 seats into one of 202. |
|
|
|
Post by James E on Sept 27, 2023 16:44:20 GMT
leftieliberal"I suspect that the reason why the swing against the Tories by the 2016 Remainers, Londoners, and the under 34 group is so low is that there were not many of them to start with. " They were certainly weak demographics for the Tories in 2019, but my analysis of the past 10 YouGovs suggests that they are performing 'above par' with at least two of these groups, even on a proportionate basis. The corollary of this is large/proportionate losses where they were strong. The last 10 YouGovs average at Lab 44.6%, Con 25.1%. So the Tories are down by nearly 20 points on a liner basis, or are retaining 56% of the 2019 Vote Share overall. In London, YG's figures are averaging Lab 54.7%, Con 19%. So that's a slightly better retention at 59% of the 32% they won in 2019. This compares to just 55% retention in the South of England, where YG show the Tories averaging 30.1% compared to 55% in 2019. And with Remain voters, it's an average of Lab 57%, Con 13%. So the Tories are retaining 68% of the 19% share they won in 2019. This compares to 59% retention with Leavers - the Conservatives appear to be faring disastrously with those who did not or could not vote in the 2016 referendum.
|
|
|
|
Post by James E on Sept 26, 2023 17:14:04 GMT
|
|
|
|
Post by James E on Sept 26, 2023 11:21:01 GMT
|
|
|
|
Post by James E on Sept 20, 2023 20:13:50 GMT
|
|
|
|
Post by James E on Sept 19, 2023 21:17:52 GMT
|
|
|
|
Post by James E on Sept 19, 2023 14:21:45 GMT
Tories 100,000 asylum seekers pledge bounce Westminster Voting Intention: LAB: 47% (+1) CON: 23% (-5) LDM: 10% (=) GRN: 7% (=) RFM: 6% (+1) SNP: 4% (+1) Via @deltapolluk , 11-15 Sep. Changes w/ 1-4 Sep. This will go down like a bag of sick in some Tory circles, I would think. Two polls short of the four poll run rule but that's two polls today putting the Tories on 23% and 24%. I know this isn't a time for cliches, and a week's a long time in politics, with very few of us possessing crystal balls, but are we looking at something existential going on here, I wonder 🤣 I don't think we can count today's Ipsos and Delta polls as part of a run, as they are not the most recent fieldwork - that was R&W's 26/44. But they do suggest that any possible tightening of the polls was just normal variation rather than polling movement. The 10-poll average Labour lead now stands at 19.3%, which is slightly higher than it was 6 months ago (18.4% average for 43 polls in Mar 2023). en.wikipedia.org/wiki/Opinion_polling_for_the_next_United_Kingdom_general_election |
|
|
|
Post by James E on Sept 15, 2023 9:45:38 GMT
The Yougov Scottish poll Scottish Westminster Voting Intention: SNP: 38% (+2) LAB: 27% (-5) CON: 16% (+1) LDM: 7% (+1) Via @yougov , 8-13 Sep. Changes w/ 3-8 Aug. As pjw1961 and others have said, this looks like a bit of an outlier, but there can be little doubt that the SNP are still ahead in Scotland - and have been throughout recent months despite the odd poll showing a tie. And it's not the only bit of evidence for a large SNP lead. The FindOutNow MRP Scottish sample of 900 showed SNP 40% and Lab 29% released in early August, although this detail was largely overlooked at the time in favour of the seats figures. Looking at YouGov's last two polls, we have SNP leads of 4 and 11, while the cross-break analysis oldnat has provided shows a 6 point lead (37/31). I'd be tempted to go down the middle and accept their average of a 7% lead. But that's higher than the 3%-ish we've seen from Survation recently. I don't have a lot of faith in R&W's polls for Scotland, Wales or London, because in each case they are seem to be an outlier, and as they have only started doing such polls recently they do not have a track-record for comparison. Among other things, R&W consistently show higher Tory VI in Scotland than any other pollster, and with so many recent elections there may be an issue of false-recall for weighting specifically to people's recalled GE2019 vote. My best guess is that the underlying position is an SNP lead of around 6%. en.wikipedia.org/wiki/Opinion_polling_for_the_next_United_Kingdom_general_election#Scotland |
|
|
|
Post by James E on Sept 15, 2023 9:18:02 GMT
steve "...the massive gains in vote share you might have expected for Labour in these local by elections just isn't happening..." The comparison in this case is to the Local Election in May2023. So Labour might expect around the same vote share, but not a massive increase. On the basis that they performed bettter in 2022 and 2023, Labour might now expect some improvement (+5-6%?) where the comparison is to the 2021LEs, but not from the last 2 years' Local elections.
|
|
|
|
Post by James E on Sept 13, 2023 11:40:04 GMT
graham I think the bookies assumption and their punters was that Labour would allow the lib dems a relatively free run at taking the seat. This doesn't now look as likely, which is up to the parties in the area and the leadership , personally I think it would have been better for them to reach a local consensus but neither seem particularly interested . This does of course leave a window for an unpopular Tory to win. I think it would take a very evenly split vote between LDs and Lab for the Tories to retain Mid Beds. Looking at the Selby & Ainsty and Somerton & Froom by elections in July, the Tories retained 57% of the 2019 vote against Labour in the former and just 47% in the latter. If this were repeated, they would get 34% or 27% of the overall vote in Mid Bedfordshire. It is entirely normal for the Tories to lose more votes in a Parliamentary by-election when faced with a LibDem challenge, but even at the higher figure of 34%, they would stand little chance of a hold. Looking at the GB-wide polling figures, the Conservatives' 10-poll average has gone from 26.3% in polls immediatly prior to those two July by-elections to 26.7% most recently. |
|
|
|
Post by James E on Sept 8, 2023 17:22:36 GMT
I've no idea who Malcom Muggeridge is/was. Will look him up. I suddenly feel old, so old 😀 Malcolm Muggeridge died in Hastings in 1990. No one noticed at the time... en.wikipedia.org/wiki/Malcolm_Muggeridge |
|
|
|
Post by James E on Sept 8, 2023 11:32:17 GMT
Westminster Voting Intention:
LAB: 45% (-1)
COM: 29% (=)
LDM: 10% (=)
RFM: 6% (+1)
GRN: 3% (-1)
SNP: 3% (=)
Via
@savanta_UK
, 1-3 Sep.
Changes w/ 25-27 Aug.
Savanta's average in 9 polls from June-Aug is around a 16.5% labour lead, so same as usual from them.
It will be interesting to see the raw figures. The last time I checked these for Savanta, they needed to adjust heavily for having too few Con2019 (and far too many Lab2019) voters by recalled vote.
|
|
|
|
Post by James E on Sept 7, 2023 10:33:06 GMT
@redfieldwilton SNP and Labour are TIED in Scotland. Scotland Westminster VI (2-4 September): SNP 35% (-2) Labour 35% (+1) Conservative 15% (-2) Lib Dem 8% (+1) Green 4% (+2) Reform 2% (–) Other 1% (–) Changes +/- 5-6 August I wonder how long we will have to wait for the first poll showing Labour ahead in Scotland? Despite the criticism on here of Labour as Tories-lite; it doesn't seem to be bothering the people of Scotland. Scottish Westminster polls in the past 3 months have all been in a narrow range from a TIE to a 4-point SNP lead. So the best approximation is that the SNP are still a couple of points ahead. This particular poll may well be simply MoE movement. en.wikipedia.org/wiki/Opinion_polling_for_the_next_United_Kingdom_general_election#ScotlandWe might see a slight Lab lead, but we could also see a 5% SNP lead - either would be consistent with normal MoE variation. I also think that the seat projections shown by Election Maps based on this poll look rather simplistic. The Tories would be most unlikely to retain 6 seats based on this figures, as they are losing far more of their overall 2019 vote-share (c35%) than the SNP is (c20%). |
|
|
|
Post by James E on Sept 3, 2023 11:43:22 GMT
Best PM rating from Opinium Starmer extends his lead over Sunak by 3 points to 27% Best PM is a funny one because it gives the incumbent a bit of a boost for actually being PM. So nothing there to stop Starmer becoming PM and he only needs to be evens to move onto other issues. However that 27% really isn't very impressive when "neither" is the winner. Also singularly Starmer is 28% approval- not sure if this is normal or not but he's getting less than 2/3rd of the Labour current VI approving of him. It's noticable that where pollsters give a 'none of these' option, the figures for both Starmer and Sunak as 'Best PM' are lower. Starmer tends to get a larger lead when 'neither' is not a prompted option - incidentally, in this Opinium his lead is just 4% (up from 3%) whereas most others seem to average around 7-10%. |
|
|
|
Post by James E on Aug 24, 2023 13:19:05 GMT
@mark Re those 'demographic changes'. Interesting to divide the recent YouGov (8 Aug) figures on 'Rejoin/stay out' by age (excl DK and WNV, which are similar across all ages, in any case). Age 18-49 79% Rejoin, 21% Stay Out Age 50-64 56% Rejoin, 44% Stay Out Age 65+ 36% Rejoin, 64% Stay Out docs.cdn.yougov.com/7myo59zisk/Internal_Brexit_230809_W.pdf |
|
|
|
Post by James E on Aug 23, 2023 22:06:34 GMT
oldnat"I haven't seen the detail of Curtice's seat projections..." Neither have I, but the seat figures match up exactly to a simple UNS, and he is using the 2019 boundaries. So SNP and Con each down 8%, LD down 3%, Lab up 16%. So, for example, Glasgow Central becomes: Lab 49% (+16) SNP 41% (-8) LD 2% (-3) Con 1% (-8)
|
|
|
|
Post by James E on Aug 23, 2023 19:42:58 GMT
Survation for Scotland SNP 37%. 24MPs Lab 35%. 24MPs Con 17%. 6MPs LD. 6%. 5MPs Oth. 5% Highest Labour vote by Survation for Labour This is a poll with a sample of 1,022, not an MRP. Seat figures are as estimated by John Curtice*. Curtice is also quoted as saying that every seat in Scotland is a maginal. Perhaps overstating it a bit, but it does seem likely that there will be a lot of very close contests - maybe even more than in 2017. I would be surprised by the Tories keeping six, given that they are shown as retaining a significantly lower share of their 2019 vote-share (68%) than the SNP. (82%). *It looks to me like those seats figures are simple UNS, as this produces 24 for the SNP. It would also explain Con 6 seats, as both they and SNP are down 8 points, while LDs gain, as they are only down by 3**. The effects on the SNP-held seats are very strange, as they would keep their 2nd-most marginal (Gordon) of their current 48 seats, while losing their Labour-facing seats all the way down to No 39, Paisley and Renfrewshire North. **UNS would still show the LDs retaining their 4 seats against the SNP or Tories, even if they too were down by 8 points to 1.5% |
|
|
|
Post by James E on Aug 22, 2023 21:38:48 GMT
A 25 point Labour lead with Deltapoll.... (etc) Genuine question for you and other deep knowledge sorts.
There seems to have been massive noise in the last few polls (R+W national and red wall polls moving in opposite directions, this one). Is there any evidence that August polls are more volatile and less reliable (in the technical sense). My hypothesis being more people on holiday and uncontactable leads to less representative samples. Has there ever been a study on this?
Thanks in advance
I'm afraid that I'm a 'Don't Know' on that one. Looking back at polls from previous Augusts, I don't see unusually variable polls. And I would expect that the lack of everyday Westminster politics in August would be more likely to lead to stagnation rather than volatility. My best guess is that we've just had a series of polls at the outer limits of Margin of Error, which have all come at once. |
|
