|
|
Post by James E on Mar 26, 2023 16:42:13 GMT
Labour's relative strength in rural Norfolk in the first half of the 20th century was based on agricultural trade unionism among farm workers. When they were replaced by mechanisation it was game over and they became 'normal' Tory voting rural areas - albeit in Norfolk North the Lib Dems were able to capture some of the old anti-Tory feeling for a while. Indeed so - though Labour remained competitive in rural Norfolk seats until 1970. Today only Norfolk North West - which includes Kings Lynn - remains a possibility for the party. Pedant Alert - the most viable Labour target in rural Norfolk is now South Norfolk, where a 17-point swing is needed. That 17-point swing compares to 20% for Great Yarmouth and 21% for Norfolk North-West - both Labour seats back in 1997. South Norfolk was one of a handful of seats which the Tories could easily have lost in 1997 (with 40%), but the LDs and Labour were then almost equal-second. In recent elections (post-coalition) it has trended to Labour, as their vote is up 6% since GE2015 (compared to +2% in GB vote) while the Tories are up 4% (compared to +8% for GB). I'd tentatively suggest that the areas bordering Norwich may have become more promising for Labour. en.wikipedia.org/wiki/South_Norfolk_(UK_Parliament_constituency)www.electionpolling.co.uk/battleground/targets/labour |
|
|
|
Post by James E on Mar 24, 2023 21:22:14 GMT
James E - just one final thought on multi-member wards and voting. I will try and illustrate what is the problem with your suggested method of looking at all the votes and why it is not used via a thought experiment. Suppose there is a three member ward. 1000 voters turn up to vote, half of whom support the Conservatives and half support Labour. However, for whatever reason, the Conservatives have fielded three candidates and Labour only one. The Conservative candidates each receive 500 votes. The Labour candidate also received 500 votes. If you look at all the votes since the Conservatives got 1500 votes and Labour only 500, this would look like a 75% Conservative ward. Only of course it isn't. The Highest vote method will take account of just one Conservative candidate on 500 votes and the Labour candidate on 500 votes and treat this as a 50/50 ward, which is what it is. As it happens, in this case my averaging method would also call it as 50/50. That analysis is good for your suggested scanario, as there is no-one for whom the Labour voters can cast their second or third choice votes. But it seems highly unlikely that Lab would field only one candidate where they might expect an even contest. On the other hand, where the Tories are strong favourites (as in Rushden South) it makes tactical sense to field just three candidates spread across the LoC parties, as they will be very likely to each receive the two parties' combined vote. Hence the strategy of Labour fielding two candidates and Greens one in a three member contest. - and the collapse of the Green vote when the same ward is re-fought in a single vote by-election. |
|
|
|
Post by James E on Mar 24, 2023 19:53:58 GMT
I am really surprised by your defending this way of analysing results, as. It flies in the face of common sense. While 'not everyone' uses all of their three votes, surely the reality must be that the majority do. In particular, I would expect that most of those minded to vote Conservative would have voted for their three candidates rather than just one or two. Also with only 2 Lab candidates and one each for the LD and Greens, it is obvious to me that all of Lab, Green and LD would have benefitted from picking up votes where they were second choices. So the apparent 'collapse' in the LD and Green vote shares in this by-election is obviously an artefact of the voting system, and number of candidates who stood 2 years ago rather than reflecting any change in voters intentions. And the Tory vote is really down by 13 points (not the 11 I quoted earlier, and certainly not up 5 points). Incidentally, the turnout for the local election in 2021 was somewhere between 23% (if all those who voted used all three votes) and 69% (if voters each cast just one vote). Which do you think is the more likely? It is not so much that I am defending it as stating a fact. All the professional psephologists who do this sort of thing for a living use the method of taking the leading candidate of each party and calculating the percentages based on that, which means that's how the percentage changes in results are then given in public forums. Councils often duck the issue by just publishing the votes and skipping percentages. It seemed odd to me when I first encountered it and I had a different method I used of adding the votes of each party together and dividing by the number of candidates to get a mean vote for each party and then using those to calculate the overall percentages by party. However, that gave me different results to the generally published ones, so I gave it up. Anyway, it is not perfect as you will often find that one candidate is vastly more popular than the others on that ticket (due to a personal vote for example) and the mean party vote becomes misleading. In the absence of a detailed breakdown of the voting or a universal move to single member wards there is no perfect solution to this. Thanks for clarifying this. As far as I can see, the comparative percentages quoted for Rushden South are in fact 'your' method of dividing the total votes by the number of candidates. But to me, this is an absloute nonsense, and calls into question all comparative figures where we have a local by-election compared to a previous multi-member Local Election. To my mind, the only worthwhile comparisons which can be made are be for parties who fielded candidates for all available seats in the previous local election- as the Conservatives did in this case. Their vote share went down by 13% (on average) in the by election. For Labour and the Green Party, the practice of fielding three candidates between them enables both to pick up each others' second and/or third choice votes, and looking at the 2021 result it is obvious that that happened. And they probably picked up a lot of LDs' 2nd and 3rd choice votes, too. There is no way of disentangling the individual parties' "first preference" votes from the three which each voter was entitled to cast. |
|
|
|
Post by James E on Mar 24, 2023 19:12:29 GMT
oldnat - "Probably starting gently, with bomb attacks on the TransPennine Express." Perhaps I better explain, for those in the south? The TransPennine Express is a railway service across the north of England, popular with people from Lancashire who like to dress as Yorkists, and vice versa. It gets quite spicey at weekends, I've been told. Anyone from south of Birmingham best keep away. Northern Ireland's intergrated transport system is another likely target. Not sure what the DUP policy on this is: www.translink.co.uk/ |
|
|
|
Post by James E on Mar 24, 2023 16:46:47 GMT
I think that it due to the way that the results are presented. Here are the figures for when the seat was last contested in May 2021. northnorthants.moderngov.co.uk/mgElectionAreaResults.aspx?ID=23The Tories (collectively) achieved 68% of the vote then, so in my view, were down 11 points. The two Labour candidates achieved 20% between them, with the Greens taking 7% and LDs 6% - each with a single candidate. Because each of them fielded fewer than 3 candidates, I would suggest that those Lab, LD and Green candidates who did stand last time benefitted from being the 'second choice' from each others' voters. It is a standard agreed way of doing the percentages in multi-member wards - just take the top one for each party. People have more than one vote but not everyone uses them all, so can't really add all the candidates together. I am really surprised by your defending this way of analysing results, as. It flies in the face of common sense. While 'not everyone' uses all of their three votes, surely the reality must be that the majority do. In particular, I would expect that most of those minded to vote Conservative would have voted for their three candidates rather than just one or two. Also with only 2 Lab candidates and one each for the LD and Greens, it is obvious to me that all of Lab, Green and LD would have benefitted from picking up votes where they were second choices. So the apparent 'collapse' in the LD and Green vote shares in this by-election is obviously an artefact of the voting system, and number of candidates who stood 2 years ago rather than reflecting any change in voters intentions. And the Tory vote is really down by 13 points (not the 11 I quoted earlier, and certainly not up 5 points). Incidentally, the turnout for the local election in 2021 was somewhere between 23% (if all those who voted used all three votes) and 69% (if voters each cast just one vote). Which do you think is the more likely? |
|
|
|
Post by James E on Mar 24, 2023 14:38:24 GMT
Rushden South LG by-election result Con 54.6% (+5.2) Lab 28.8% (+6.6) LD 7.1% (-6.3) Breakthrough 5.4% (new) Green 4.2% (-10.8) Turnout 19.3% Edit: Breakthough are a far left party btw All the usual caveats but interesting that Con actually increased their percentage. This has happened quite aLooking at the figure few times since we've been looking at these results. I think that it due to the way that the results are presented. Here are the figures for when the seat was last contested in May 2021. northnorthants.moderngov.co.uk/mgElectionAreaResults.aspx?ID=23The Tories (collectively) achieved 68% of the vote then so, in my view, were down 13 points in this by election. The two Labour candidates achieved 20% between them, with the Greens taking 7% and LDs 6% - each with a single candidate. Because each of them fielded fewer than 3 candidates, I would suggest that those Lab, LD and Green candidates who did stand last time benefitted from being the 'second choice' from each others' voters. |
|
|
|
Post by James E on Mar 21, 2023 18:13:31 GMT
Westminster Voting Intention:
LAB: 46% (-2)
CON: 31% (-1)
LDM: 8% (=)
SNP: 4% (+1)
RFM: 4% (+1)
GRN: 2% (=)
Via
@survation
, 17-20 Mar.
Changes w/ 13-15 Mar.
15 point lead with Survation - slightly down on their previous polls since the start of Feb (16, 19, 16, 16)
|
|
|
|
Post by James E on Mar 21, 2023 11:47:43 GMT
shevii "Lib Dems main problem in their lumpy vote translating into more seats is that they aren't really in that strong a position in the seats you might think they had a chance of the ABT vote (I guess mostly the 50 seats they held pre Clegg meltdown plus some new blue wall) and in many cases having to jump from 3rd to 1st even if they are the only ones likely to beat the Tories in those seats..." The LibDems have had a number of by-election wins from 3rd place, but I can't recall them ever doing so in a General Election. And it seems to me highly unlikely that they will in 2024. Polling cross-breaks show them losing 30-45% of their 2019 voters to Labour, while regional cross-breaks and polling such as R&W's 'Blue Wall' show the largest Labour progress in exactly the places where they have traditionally been weakest - especially the South-West. R&W South West England average 12 polls from 3 Jan to 19 March Lab 47% (+24) Con 26% (-27) LD 15% (-3) To give some examples of seats that I think now are realistic Labour targets, I'd go for Weston-Super-Mare, which was an LD seat from 1997-2005. The current Conserative majority of 31% is right in line with their lead in the whole SW region in 2019. Other 'non-traditional' targets in the SW region would be St Austell and Newquay (also needing a 15% swing) and Somerset North (14%). www.electionpolling.co.uk/constituencies/uk-parliament/weston-super-mare
|
|
|
|
Post by James E on Mar 20, 2023 17:52:16 GMT
When considering those Deltapoll figures, it should be remembered that their previous poll showed an unusually high Labour lead (for Deltapoll) of 23 points. Per the figures I posted yesterday, that 10 point lead is 'only' 7 points lower than their 2023 average of 17.5%.
Today's R& W with a Lab lead of 21% is 2 points lower than their 2023-average, while PeoplePolling's 25-point Lab lead is entirely normal (for them).
It's also worth noting that those seat figures are based on a simple Uniform Swing, on the new boundaries. Tactical voting is bound to be a factor, but more importantly, all the polling companies' regional figures show very non-uniform swings - typically 3-5 points higher in the South of England (outwith London), around 2-3 points higher in the Midlands, but 5 or more points lower in London itself.
TO ADD:
Deltapoll themselves show that 'non-uniform' swing very clearly. Their nine previous GB polls average at Con 29%, Lab 46%, and so around a 14-15 % Con> Lab swing. But taking their figures for England region-by-region, the swings are:
London 3%
South 19%
Midlands 22%
North 14%
Lower swings in Wales and Scotland too (which is consistent with full polls in each case), but far larger than average swings in both the South and Midlands. And this is the same with YouGov, R&W and Opinium's figures.
|
|
|
|
Post by James E on Mar 20, 2023 10:12:52 GMT
Some 'comic relief' from People Polling. Lab 45% (+3) Con 20% (-3) Green 13% (+3) LibDem 9% (+1) RefUK 6 % (-1) SNP 5% (+1) fieldwork 16-17 March peoplepolling.org/tables/202303_GBN_W11_full.pdf#subsection*.12No other recent poll has the Green higher than 7%, and their average is around 5-6% (and 2% with Survation). PP's cross-break for South England shows Lab 26 points ahead, which on UNS would give the Tories zero MPs in the SE and SW England regions. (This compares to a 5 point Lab lead in the South with YouGov or 2 points with Opinium) |
|
|
|
Post by James E on Mar 19, 2023 20:02:32 GMT
There's a staggering difference between Survation's figures of Green 2%, Ref 3%, and the likes of YouGov .. I think the reason for this is in Survation's polling question, which is: " If the Westminster Election was taking place tomorrow and there was a candidate from all political parties standing, which party do you think you would vote for?" They do not prompt with the names of any political parties... Interesting that it appears to make little difference to the Lab lead (technically you could argue this poll shows a lower Lab lead than the average although there's been a bit more variance in polling recently anyway). Maybe the ones who were given the choice of other parties filter into don't knows or less likely to vote? I'll be mildly interested in Opinium tonight- some of the recent polls from other polling companies have been edging closer to their "predictive" figures. I assume doing an average of recent polling will show a small drop in Lab lead although nothing for Lab to worry about, but there's no sign yet of Opinium to date showing lower leads. If the lead does remain the same with them then maybe all we're seeing is the predictive element becoming the actual position and the "lets face it they were always going to vote Tory, even if they were embarrassed by some of the goings on recently". Also presumably a Kantar is due soon who I think operate methodology similar to Opinium. Before they changed their methodology 12 months ago, Opinium were producing the most favourable figures for Labour. As a result, figures for the early months of 2022 were generally in the range of 1-5 points lower then other pollsters, even though their headline figures were moved by 6-7 points by their methodological change. As the Tories became more unpopular in mid 2022, their 'Don't Knows' grew and so the gap between Opinium and other pollsters grew - up to about 10 points when Truss was briefly PM. Since then, as Sunak has clawed back more of the DKs, the gap has narrowed again. But all through this, it has been noticable to me that Opinium and YouGov's polling findings are actually very similar: the difference in their headline figures is all due to Opinium's re-weighting. As you say, other companies are now showing very similar figures to Opinium, so may also be re-weighting in the same way. Looking across the polling industry, there are now two distinct groups, with a 'high Lab lead' group averaging around a 23% and a 'low lead' group averaging more like 17%. Only Techne seem to fall in the middle. Average leads by polling orgnaisation so far in 2023 are: Ipsos Mori (2 polls) 25.5% People Polling (11) 25% Redfield and Wilton (11) 23% Omnesis (11) 22.5% Focaldata (2) 22.5% YouGov (12) 22.3% Techne (11) 20% Deltapoll (10) 17.5% Kantar (1) 17% BMG (3) 17% Survation (4) 16.8% Savanta (7) 16% Opinium (5) 15.7% |
|
|
|
Post by James E on Mar 18, 2023 19:54:47 GMT
As Electoral Calculus currently has Labour on course to win 484 seats to the Conservatives 84, I think we can agree that it overstates the Labour gains. The question should be whether Labour will perform as well in Devon Central as in similar-range seats such as Graversham or Burton. But I do think that the Local Elections can provide much of an answer to this..... etc I don't disagree with that and I'm happy to adjust whatever national vote share projection that is made after the May elections for it, but it is a different point to the one I am making, which is about extreme historic weakness of Labour in certain types of seats (those outlined above) which makes them less likely to be won than the polls would indicate. I would suggest that the likeliest outcome in those areas is that the anti-Conservative vote will be split between Labour (which will get some boost from a general uplift, as happened in 1997) and the Liberal Democrats (the more traditional opposition in those areas) resulting in the Conservatives holding those seats on a minority of the total vote. Several of the constituencies which you outlined above ( Devon SW, Totnes, Tiverton and Honiton, Wiltshire SW, Salisbury, Devizes, Huntingdon, Cambridgeshire NW, SE and South and Suffolk West) don't even make it to the top 250 Labour targets, and could only be Lab-gains on the 20+ point leads we have in some polls. A more realistic approach would be to look at those top 250 Labour targets (specifically the 213 which are Tory-held), and see where Lab and LDs could be competing against a Tory incumbent in the 30-45% range at the next election. www.electionpolling.co.uk/battleground/targets/labourOf the 213 Conservative-held seats listed here, Labour were in 2nd place to the Tories in all but around 30,, despite their poor showing in GE2019. Those with Labour in 2nd include Devon Central, incidentally. But many of those 30-or-so show very close figures for the LDs in 2nd and Lab in 3rd (e.g. Salisbury) , and given that Labour are still around 15 points up on GE2019, it still seems reasonable to target these. The same applies wherever Labour are within 15-20 points of the Tories (such as in Wimbledon or Finchley), who of course are polling 17-18 points lower than 3 years ago. This leaves us with just a few seats on the top 250 targets where it could be argued that Labour should 'unprioritise' in favour of the LDs, on the basis that the latter may still be better placed. These are: Woking Cambridgeshire SE Wantage Totnes Hitchin and Harpenden Sutton and Cheam. Coincidentally, all 6 of these are among the 42 constituencies regularly polled as part of R&W's so-called 'Blue Wall' of Southern Tory held seats - and this polling has generally been far more encouraging for Labour than the LDs. But notwithstanding this, it looks to me like there are only a relatively small number of realistic target seats where it is hard to predict who will be better placed at the next GE. |
|
|
|
Post by James E on Mar 18, 2023 15:39:24 GMT
pjw1961 As Electoral Calculus currently has Labour on course to win 484 seats to the Conservatives 84, I think we can agree that it overstates the Labour gains. The question should be whether Labour will perform as well in Devon Central as in similar-range seats such as Graversham or Burton. But I do think that the Local Elections can provide much of an answer to this. As I have detailed before, there is a long-established pattern in English Local Elections of the LDs doing a lot better than their (Westminster) polling, and Labour a lot worse. The 2022 Local Elections in England which I've quoted above, were consistent with this. In some places, people vote differently in Local and Westminster elections. As a often-quoted example, these were the vote shares in the local elections for Watford in 2004: LibDem 47% Conservative 27% Labour 17% Green 8% But the following year's General Election for the Watford constituency ended up: Labour 34% LibDem 31% Conservative 30% Green 3%
|
|
|
|
Post by James E on Mar 18, 2023 12:25:08 GMT
.....the little polling of (truly) 'rural' voters that we have suggests a Con>Lab swing much in line with the rest of Britain... etc Hi JamesE. Even as I typed the word 'rural' I was aware that it was the wrong word, but I don't think there is a collective term for the seats I am talking about, so I will try and describe them. I mean seats where the Labour party's history of electoral success is very poor or non-existent and which don't contain a specific set of locations - such as a town sized urban area or a set of old mining villages - that would provide a bedrock of Labour voters that can be built on, but which the polls are predicting will fall to Labour.(generally on MRP rather than uniform swing, but sometimes the latter too). To take one example, Electoral Calculus have Devon Central falling to Labour by a 4,200 majority, with Labour having a 70% chance of winning the seat. The seat contains no substantial settlements, no history of industrial activity, no strong Labour areas in local elections and Labour have never won in this area. I am much more of a psephologist than a polling expert (polls have a statistical element that is not my forte) and from that perspective my instinct is that the polls - or at least the interpretation of the polls - are getting these areas wrong. I would cite three pieces of evidence: 1) There is no evidence - literally none - from regular local elections or by-elections of a huge surge in Labour voting in these types of areas. The Conservatives are losing seats, but they are losing them to Independents, Lib Dems and Greens. We have elections in May that include a large number of English District Councils that should provide a further test of this. 2) The three parliamentary by-elections that have come up in this type of area have all been won by the Liberal Democrats not Labour 3) Psephological history - with a tiny number of exceptions, these types of area were not won by Labour in 1945, 1966, 1997 or 2001. Why are folk in these areas with no personal or family history of voting Labour even in its great years and under its strongest leaders going to do so next time? I don't have the same issues with urban areas. I find it perfectly credible to believe Labour can win in Bournemouth for example, when you look at how somewhere like Brighton and Hove, once safely Tory, has become a left-wing bastion. To take your three pieces of evidence in turn: Firstly, local by-elections are notoriously unreliable in predicting General Elections. I have linked this before, but to summarise the findings of our former UKPR host, for the 1997-2005 GEs they pre-GE by-elections overestimated the LDs by an average of 10 points, and underestimated Labour by 9 points. And the results are hugely variable locally, dependent on local organisation and getting the vote out. pollingreport.uk/articles/can-local-by-elections-predict-general-electionsAnd taking the most recent Local Elections in June 2022, with Net Equivalent Vote of Lab 35%, Con 30%, LD 19%: this is again typical of historic performance. Labour were about 10-12 points ahead of the Tories in polls at that time, but there are good historic precedents for the LDs overperforming by 5-10 points, and a similar underperformance by Labour (and a little by the Tories, too) in Local Elections. Parliamentary by-elections are another well-known area of LD overperformance. For the past 50 years we have had Liberal (or LD) surges in by elections for seats where they are perceived as likelier to beat the Tories. And this happened even in the 1992-97 Parliament, in seats such as Christchurch, Eastleigh and Littleborough and Saddleworth. The three LD gains during the current Parliament came at times when the general polling position was far less promising for Labour then it is now. As for psephological history. To take the example you quote of the Bournemouth seats: these have already seen some significant improvement for Labour at the last 2 GEs*, in that they have made net gains against the Tories of 8 and 12 points respectively in Bournemouth West (Con+5, Lab+13) and East (Con +4, Lab +16). But the pattern is much the same in the Central Devon seat (Con +3, Lab +12), or Somerset North (Con +0, Lab +10) and is easlily exceeded down in Truro and Falmouth (Con -1, Lab +23). So there has been recent Labour advance in places which are either rural, or lack a previous Labour tradition, just as much as in your example of Bournemouth. A further point to consider in the 'regional' pattern in recent polling. I have highlighted many times and across 4 different polling organisations how the higher swings are higher in the South of England, and lower in the North and (especially) London. But there is also a general pattern of Labour gaining more in exacly the places where they were previously weakest. Using the figures from R&W, these can be summarised as follows: Large Lab lead in 2019 (16%): London. Con>Lab swing in 2023 6% Small Lab leads in 2019 (4-8%) Wales, North-East, North-West Con>Lab swings of 10%-14% Small Con lead in 2019 (4%) Yorkshire & H Con>Lab swing of 17-18% Large Con leads in 2019 (21-32%) South East, South West, East, W Mids, E Mids) Con>Lab swings of 21-26% The largest swing of all by region per R&W is the South West of England, which is also probably the area with the largest concentration of rural or 'not traditionally Labour' places. * comparisons are from GE2015 to GE2019 in all cases. |
|
|
|
Post by James E on Mar 17, 2023 18:29:43 GMT
There's a staggering difference between Survation's figures of Green 2%, Ref 3%, and the likes of YouGov (6% for each) or at the top end, Omnesis ( 7% and 9%) and People Polling ( 10% and 7% ) who have these two parties taking 16% and 17% of the vote between them. I think the reason for this is in Survation's polling question, which is: " If the Westminster Election was taking place tomorrow and there was a candidate from all political parties standing, which party do you think you would vote for?" They do not prompt with the names of any political parties. Survation's approach seems to me more likely to reflect what will happen in an actual election when the smaller parties are squeezed by FPTP. And it means that we have at least one pollster when the Ref>Con (and Green>Lab) reversion is already factored in. |
|
|
|
Post by James E on Mar 17, 2023 9:58:47 GMT
pjw1961 Re Cottenham by election in South Cambridgeshire. Thanks for that - interesting stuff. As South Cambridgeshire is my local area, I feel well placed to comment on this one. The LibDems control South Cambs council, so were in the position of defending their record in power in this by-election. As such, the swing against them is unsurprising. The proposed Cambridge congestion charge is unpopular with some, and the Tories are likely to have picked up votes on this. I agree with you regarding the likely outcome of the South Cambs constituency: it is clearly not a Labour target, as they start miles behind the LDs, who took 42% last time - almost their highest share in any constituency they failed to win. In fact, I would expect a larger LD majority than you predict. However, I don't think there is much polling evidence to support your contention regarding 'rural seats'. Firstly, few seats are truly 'rural'. Much of South Cambs seat consists of large 'villages' which are part of the Cambridge conurbation, but outside the formal boundaries of the city. And the little polling of (truly) 'rural' voters that we have suggests a Con>Lab swing much in line with the rest of Britain. See this from YouGov last October. yougov.co.uk/topics/politics/articles-reports/2022/10/06/conservatives-are-now-13pts-behind-among-rural-votAnd this from Dec 2022. www.mirror.co.uk/news/politics/rural-voters-just-likely-back-28803967The 16.5% swing in the Mirror's article is right in line with the general polling average for Dec 2022.
|
|
|
|
Post by James E on Mar 15, 2023 12:49:15 GMT
Another 22-point lead with YouGov from a few days ago, but no details or tables at present: (from twitter) "Latest YouGov in The Times Red Box this morning: Labour 45 (-2) Tory 23 (-2) LibDem 10 (-) Greens 7 (+1) Reform 7 (+2) SNP 4 (-) No polling dates provided, but likely to have been Tuesday and Wednesday. If so, no Tory small boat bounce as yet." ... and 6 days later, YouGov have now finally released the full figures and dates. These figures were for 7-8 March, and we now have 6 more recent GB polls, averaging a 20-point Labour lead. yougov.co.uk/topics/politics/articles-reports/2023/03/15/voting-intention-con-23-labour-45 |
|
|
|
Post by James E on Mar 14, 2023 21:35:32 GMT
A poll like that could encourage Sunak to go to the country early. A brilliant outcome. I reckon October 2024- any earlier and they won't have much chance of a recovery in the polls (not that they probably do anyway), any later and no doubt there will be another NHS winter crisis. I would not be surprised if Sunak (or whoever is Tory leader) took it to the latest possible date, as other PMs have done when behind in the polls. As to when that is : "...if it has not already been dissolved at the request of the prime minister, Parliament automatically dissolves five years after it first met (17 December 2024) and polling day occurs no more than 25 working days later taking into account Christmas Eve and bank holidays in any part of the United Kingdom. I make that Tuesday 28th Jan 2025, or if he sticks to the tradition of Thursday elections, Thursday 23rd Jan. [Edit - or 24th Jan 2025, according to Wikipedia ] en.wikipedia.org/wiki/Next_United_Kingdom_general_election |
|
|
|
Post by James E on Mar 14, 2023 19:03:15 GMT
Re that R&W 'Blue Wall' poll.
Compared to GE2019, it's:
Lab 36% (+15)
Con 34% (-16)
LD 23 % (-4)
Ref 4% (+4)
Grn 3% (+1)
That's still a 15.5% swing compared to GE2019, and equivalent to a Labour lead of around 19 points in a full Westminster poll. Previous polls of this set of constituencies have averaged around a 19% Con>Lab swing, a couple of points higher than thse of all GB, whereas this one is a couple of points lower than the prevailing polling across GB.
Meanwhile, JLP's 'Red Wall' poll shows a 17% Con>Lab swing, so is right in line with the average of Westminster polls, with Labour around 21 points ahead.
|
|
|
|
Post by James E on Mar 10, 2023 17:04:46 GMT
Results of the three elections yesterday, nit sure they tell us much One thing I noticed was that in two of the three the Labour percentage fell by more than that of the Tories. All the usual caveats about local by-elections of course, but it does continue the theme that the public aren't exactly enthused by Labour. It would be very interesting to see what equivalent elections were doing in 1996. This might give an idea of how Labour will actually do in the GE. For instance, if Labour results back then were similarly lukewarm then a massive landslide could indeed be on the cards but if Labour were doing very well then the next GE could be a lot closer than the polls are suggesting. On the old site, Anthony Wells provided data back in 2009 on Labour's performance in pre-General Election by-elections. "Sum of local by-elections Jan-May 2005: CON 33%, LAB 26%, LDEM 31% Actual general election result 2005: CON 33%, LAB 36%, LDEM 23% Sum of local by-elections Jan-Jun 2001: CON 32%, LAB 30%, LDEM 26% Actual general election result 2001: CON 33%, LAB 42%, LDEM 18% Sum of local by-elections Jan-May 1997: CON 28%, LAB 37%, LDEM 30% Actual general election result 1997: CON 31%, LAB 44%, LDEM 17%" So a substantial underperformance by Labour and overperformance by the LDs is the norm. |
|
|
|
Post by James E on Mar 9, 2023 19:41:53 GMT
Another 22-point lead with YouGov from a few days ago, but no details or tables at present:
(from twitter)
"Latest YouGov in The Times Red Box this morning:
Labour 45 (-2)
Tory 23 (-2)
LibDem 10 (-)
Greens 7 (+1)
Reform 7 (+2)
SNP 4 (-)
No polling dates provided, but likely to have been Tuesday and Wednesday. If so, no Tory small boat bounce as yet."
|
|
|
|
Post by James E on Mar 8, 2023 23:20:37 GMT
mercian"Which ONE of the following will be most important to you in deciding how to vote in the referendum?" "Britain's right to act independently, and the appropriate level of co-operation with other countries" (which is what I mean by sovereignty) - 45% Leave voters said that, followed by 35% about immigration. Only 8% said that the economy and jobs was the most important issue. Yes, that's a good point, as that question about the most important issues illustrates the very different priorities that Leave and Remain voters had (with Remainers stressing the economic issues more). But the poll also shows clearly that very few Leavers expected any negative economic consequences, with most seeing either no change or an upside to all the questions asked.
|
|
|
|
Post by James E on Mar 8, 2023 22:48:04 GMT
I would suggest remainers saw they would be richer staying in so put up with it, and leavers that they would not be poorer leaving, so as well be shot of it. Corbyn giving the EU 7/10 was probably dead on the money for most remainers. Few people love the EU from either side, but also few hate it. Nobody on the Leave side that I knew was motivated in any way by economic considerations. It was about sovereignty. I expect that there were some Leave voters who thought they would have a better chance of a better paid job with less competition, but I never met one. As I have mentioned before there was some good 'eve of Referendum' polling by YouGov on 20-22 June 2016 on what Leave and Remain voters expected the effects of the decision to be: d25d2506sfb94s.cloudfront.net/cumulus_uploads/document/atmwrgevvj/TimesResults_160622_EVEOFPOLL.pdfThose intending to vote Leave believed by a margin of 48% to 4% that Leaving would make Britain economically better off (as opposed to worse at 4%). Those intending to vote Leave believed it would have a good effect of jobs by a margin of 44% to 2%. Those intending to vote Leave believed it would make them better off, rather than worse off, by 20% to 3%. Many said 'no difference' or Don't Know to this. Those intending to vote Leave believed it would be good for the NHS rather than bad by 69% to 2%. Those intending to vote Leave thought there would be less (rather than more) immigration by 84% to 1%. |
|
|
|
Post by James E on Mar 8, 2023 20:36:45 GMT
|
|
|
|
Post by James E on Mar 8, 2023 19:55:31 GMT
With YouGov belatedly releasing their figures to 1 Mar today, I have taken the chance to re-visit my English regional cross-break analysis of their polls. These are the averages of the 9 YouGov polls with fieldwork solely in 2023. I have included the LibDem figures for London and the South too, as these are the areas where they are more competitive.
All figures are compared to GE2019, and come in the context of polls showing typically a 22-23 point Lab lead, so a swing of 17.5% since the 2019 General Election.
North of England
Lab 56% (+13)
Con 22% (-17)
Con>Lab swing of 15%
Midlands
Lab 48% (+15)
Con 29% (-25)
Con>Lab swing of 20%
South of England
Lab 41% (+18)
Con 31% (-21)
LD 13.5% (-3)
Con>Lab swing of 20.5%
(and, arguably, Con>LD 'swing' of 9%)
London
Lab 57% (+9)
Con 19% (-13)
LD 9% (-6)
Con>Lab swing of 11%
So a very similar picture to when I last checked YouGov's cross-breaks, and indeed to the other pollsters for whom I've done this, such as R&W and Opinium. YouGov's cross-breaks seem to be more consistent that R&W's, so to me this might suggest greater accuracy.
|
|
|
|
Post by James E on Mar 8, 2023 16:07:55 GMT
A couple of details on EU-related polling released in the last 24 hours: Deltapoll's 2-6 Mar poll includeded a Stay out V Rejoin question, which was 58% Rejoin, 42% Stay Out. YouGov (28Feb-1Mar) asked their usual "Hindsight" tracker which was 62% 'Wrong to Leave' and 38% 'Right to Leave' after excluding the DKs. Both of these are very typical of recent findings. But more interesting than these is the new YouGov question: " As far as you are concerned, is Brexit 'done'?" The headline figures are: It is done 18% It is not 'Done' 60% Don't know 21% The fact that 2016 Remainers, Labour and LD voters overwhelmingly think it is 'not done' is unsurprising. And the figures show only a little variance beyween the different age-groups. The really interesting figures are for 2016 Leave voters and Tories. For 2016 Leavers, it's a 26/57 majority for 'Not done'. So barely a quarter of those who voted for it think that Brexit is 'done'. And of 2019 Tories, only 30% think it is 'done' against 55% 'not done'. The highest response for 'done' is among the current Tory VI, but even there, only 40% regard Brexit as 'done' compared to 44% who say it is 'not done'. docs.cdn.yougov.com/saca6dp8tm/TheTimes_VI_Brexit_230301_W.pdf(see page 4 of 6 in link) |
|
|
|
Post by James E on Mar 7, 2023 19:05:38 GMT
pjw1961"..the "red wall" will not be the only battleground in the next election. Prosperous but liberal and remain-ish middle class areas are a definite Labour (and Lib Dem in some cases) target next time. " Indeed. The longer distance battleground looks very interesting for the next GE, and there is also some evidence (per Deltapoll's cross-breaks) that the Tories may be having more difficulty in holding their Leave supporting 2019 voters than the smaller number of those who backed Remain. Looking at the most recent R&W polls for the Northern and Southern English battleground seats ('red' and 'blue' walls), the Con2019 cross-breaks are: Northern marginals: Con 57% Lab 19% LD 2% Ref 8% DK 13% Southern Marginals: Con 53% Lab 21% LD 6% Ref 4% DK 13%
|
|
|
|
Post by James E on Mar 7, 2023 17:11:04 GMT
re the R&W "Red Wall" poll. Compared to GE2019 its: Lab 51% (+13) Con 29% (-18) Ref 9% (+2.5) LD 6% (+1.5) Green 2% (+0.5) Swing is a little lower than R&W's most recent GB polling at 15.5% compared to 19%. Typically, these 'Red Wall' polls have shown a swing 1-2 points lower than their GB polling; this might be explained by the low LibDem2019 vote in these constituencies. This latest poll is at the low end of the range for their 'Red Wall' polls so far in 2023, which have had Labour leads of 22,26,23,28 and now 22 again. en.wikipedia.org/wiki/Opinion_polling_for_the_next_United_Kingdom_general_election#Other_polling |
|
|
|
Post by James E on Mar 7, 2023 16:22:03 GMT
A couple of polls released this afternoon - albeit that the fieldwork for Ipsos is quite old:
Ipsos:
LAB: 51% (=)
CON: 25% (-1)
LDM: 9% (=)
GRN: 5% (=)
RFM: 3% (+1)
Via
@ipsosuk
, 22 Feb - 1 Mar.
Changes w/ 18-25 Jan.
Deltapoll:
LAB: 47% (+1)
CON: 31% (=)
LDM: 8% (=)
[no other details quoted]
Via
@deltapolluk
, 2-6 Mar.
Changes w/ 24-27 Feb.
Average Labour lead now 21 points for the 8 polls conducted wholly or partly in March. This compares to an average of 22 points for the 9 polls conducted in the first week of February.
|
|
|
|
Post by James E on Mar 4, 2023 19:06:28 GMT
A 2-point narrowing of the Labour lead with Find Out Now. Seems fairly consistent with the average of a 3-4 point narrowing with the other polls so far in March.
Westminster Voting Intention:
LAB: 48% (=)
CON: 25% (+2)
LDM: 9% (-2)
GRN: 6% (+1)
RFM: 5% (=)
Via
@findoutnowuk
, 1-3 Mar.
Changes w/ 27 Jan - 5 Feb.
|
|
