|
|
Post by James E on May 15, 2023 17:51:47 GMT
Deltapoll's latest 16% Lab lead is much in line with their average for the year to date.
Opinium's recent 14% and Techne's 17% Lab-leads are each just 1 point below their 2023 averages.
The outliers are R&W, whose 14% is still 6 points below their previous 2023 average, and Omnesis, whose 27% is 5-6 points above theirs.
|
|
|
|
Post by James E on May 15, 2023 17:03:30 GMT
I think a lot of Conservatives vote LD in local elections. But go home and hold their noses in a GE. The Lib Dems have won Mayoral and local elections in my constituency Watford for well over a decade. Labour has 1 seat on the council. However, the 2019 Election had Con on 26,421, Lab on 21,988, LD on 9,323. As an interesting aside Watford's MP has been in the ruling government in every election since 1979. I do not think a tactical vote is possible in Watford as its so split. I can see this happening for the Conservatives across several LD gains. So they are in with a shout as the red wall will not be enough for a majority I think. Also, I think turnout in an election is key. 1997 71.4% - Labour Landslide 2001 59.4% - Things going well voters were not motivated to change government 2005 61.4% - Labour had such a majority whilst Iraq really cost them the required swing was too high for the Conservatives to win, but they gained a decent foothold. 2010 65.1% - An appetite for change motivated a change in government 2015 66.1% - A very good conservative manifesto and a previously hung Parliament, Con voters came out and delivered. A majority that Cameron was not expecting. 2017 68.8% - Brexit dominated and both sides of the argument came out in numbers. If the con vote had stayed at home Corbyn likely would have won, that additional 2.7% vote was vital to the outcome 2019 67.3% - Again Brexit brought out the Con vote in numbers. Next year I expect for Labour to win a majority the turnout will be key. I don't think you can consider the 'Red Wall' in isolation from the rest of the country. Polling there, such as Redfield and Wilton's (see link below) suggest that the Con>Lab swing in those constituencies polled (all recent Con gains) is much the same as in polling of all GB (around 15%). So these seats do not need to be 'enough for a majority'. en.wikipedia.org/wiki/Opinion_polling_for_the_next_United_Kingdom_general_election#Redfield_&_Wilton_StrategiesTurnout looks likely to be low - at least if you compare those telling pollsters they are 'certain to vote'. These are down by around 4 points compared to previous pre-election years. Not sure it will be 'key' though. The very similar results in 1997 and 2001 happened in the highest and lowest turnouts of the past 30 years. |
|
|
|
Post by James E on May 15, 2023 10:35:51 GMT
"Flynn said: "It's increasingly clear that the SNP can hold the balance of power after the next general election - putting Scotland in prime position to pull the strings of a minority UK government."
It looks to me like this is another of those judgements based on UNS - and the false narrative that Labour need a lead of 13% in the overall vote to form a majority.
My own claculations suggest that the point that Labour can get an overall majority is around 4% lead - such as 39/35. Electoral Calculus suggests a little lower with tactical voting. I am working on an assumption of just 5-10 Lab MPS in Scotland: the same dynamics which show Labour doing far better than UNS in England on current polling also suggest that they would underperform against the SNP in Scotland unless they can draw very near to being level in Scottish vote share with them.
It's harder to tell what sort of figures would make it possible for Labour + LDs to get to 326 seats, or near enough to get an effective combined majority. But my best estimate is that this would happen where Labour have a small overall lead (0%-3%). This is based on the LDs winning 15-25 seats.
This leaves us with the scenario that the Tories have too few seats to govern, even with the DUP. And at the same time, Labour cannot command a majority even with LD support. I think this is what would happen if the Tories have a small lead in the overall vote of 0%-3%.
The Tories have reduced Labour's polling lead from 21-22% 3 months ago to an average of 17% now - and of course just 7% in the Local Election NEV figures. But that is still a long way from the kind of result which would give the SNP the balance of power.
|
|
|
|
Post by James E on May 13, 2023 18:30:28 GMT
I suspect ITV was reporting results before declarations had actually been made. Shortly after the Portsmouth North gain BBC did report that Labour was claiming to have won Crosby - and Rochdale -but the results had not actually been announced. My source for the sequence of events is "Were you still up for Portillo?", Brian Cathcart, (then deputy editor of the Independent Sunday), Penguin Books, 1997. You are correct that ITV generally report results more quickly that the BBC (even now) because the beeb wait for the returning officer to read the declaration out and ITV report when they know the result, but in this case Cathcart devotes two paragraphs to the Crosby timing issue (p43-44) and seems convinced Crosby declared first. I suppose it could only be resolved by finding out the timing of the Crosby declaration (the Edgbaston one was caught on camera of course). This supports your (and Cathcart's) version of events: the ITV coverage from around midnight on 2nd May 1997. www.youtube.com/watch?v=FLMJ86mVCLcCrosby is shown as a Labour gain, at 14 mins, in the 10% swing column (as that is what was needed for a Lab gain). However, they do not report the actual figures, nor the 18% swing achieved. At 18 mins ITV go to Edgbaston for the live declaration. |
|
|
|
Post by James E on May 13, 2023 17:10:11 GMT
James E - the fact of bigger swings in safer Tory seats in 1997 caught the broadcasters out in one respect. The first Labour gain of the night was actually in Crosby and the swing was over 18%, but no one had expected it to be in play and all the attention (and cameras) was on Birmingham Edgbaston, which was won sure enough but on a more typical swing of 10%. As the sensational results then poured in, Crosby never got its moment in the limelight. The normal swing in a Lab gain was around 13% in 1997 compared to the UNS of 10%. In fact, the Edgbaston swing of only 10% was really a weak performance in the context of the election as a whole. I don't think any lab gain in 1997 had a swing of less than 8.5%. It's worth noting too that varying swings between Con and Lab have become even more common in recent elections. The highest we had in GE2019 was (I think) Bassetlaw, which was 18.4% Lab>Con compared to UNS of 4.5% while Putney defied the 'national' swing the most by going 6.4% Con>Lab. |
|
|
|
Post by James E on May 13, 2023 16:23:34 GMT
Re 'Proportionate Swing' There is a good illustration in how this works from the latest YouGov tables (below). While it's never wise to rely on a single poll, let alone single cross-breaks, I will just say that these are fairly typical for YouGov (and most pollsters), although the swing in London is actually somewhat lower in most. docs.cdn.yougov.com/9yxa6z6y1e/TheTimes_VI_230504_W.pdfHeadline VI Lab 43% (+10) Con 26% (-19) Swing Con>Lab 14.5% London sample: Lab 53% (+5) Con 19% (-13) Swing Con>Lab 9%North of England sample: Con 23% (-16) Lab 53% (+10) Swing Con>Lab 13%Rest of South England sample: Con 32% (-22) Lab 38% (+16) Swing Con>Lab 19%So Labour are gaining more, and the Tories losing a lot more in the South outside London. But if you look at the Conservative losses on a proportionate basis, this pattern all makes sense. In London they are on 19% compared to 32% in 2019. That's 59% of their 2019 vote In the North, they now have 23% compared to the 38% they won in 2019. And that's 59%, too. And in the South, where they are down the most, they now have 32% compared to the 54.5% . And once again, that's 59% of their 2019 voters Obviously, this does not fully explain the differential Lab gains, but these are clearly greater where the Tories are losing the most; the Lab>Green movements seems to be stronger in London, although of course that too is probably a proportionate effect. This is actually just simple Maths - multiplying fractions, as children are expected to be able to do before they leave Primary School. Which makes it all the more surprising that the media (e.g Sky News last week) continue to present forecasts based on Uniform swing. |
|
|
|
Post by James E on May 13, 2023 15:27:56 GMT
@lululemonmustdobetter "I watched a bit of the coverage of the '97 election (I was so young then) the other day, and some of the swings to Lab were very high. With an overall swing of 8.8% they gained 146 seats. Matching that would give Starmer a comfortable majority." To be pedantic, the swing in 1997 was 10%. Labour were up by 8.8%, and the Conservatives down by 11.2%. But I fully agree that a similar movement* would deliver a very comfortable majority. However, a Uniform Swing of 10% would in fact leave Labour short of a majority. And in various parts of the media, this is being used to raise the spectre of a 'coalition of chaos', even though it is a wildly inaccurate predictor of seats changing hands - especially when there is a large swing. Looking at the 1997 results, Labour achieved around 40 gains in excess of the number predicted by a 10% uniform swing. And this was mostly due to 'proportionate swing' with a bit of help from tactical voting. For this, please see page 14 of 18 of the attached paper on Constituency system and the 1997 landslide. The differential movements between all different types of seat are provided, but as a preliminary point, the Conservatives lost most (13.1%) and Labour gained most (13.8%) in safe Conservative seats where Labour were second. The average swing in all Conservative held seats with Labour second was around 13% - so a 3% 'bonus' on top of UNS. And note that the Tory losses were all greater than the average of 11.2% where they were defending. Labour also did well in marginal seats they were defending, but the lowest swings (av 7.9%) came in safe Labour seats. Again, this is simply how a 'proportionate swing' works. www.dannydorling.org/wp-content/files/dannydorling_publication_id1318.pdf* This looks plausible enough. If we have identical movements in 2024/5, it would be: Lab 41.8% (+8.8) Con 33.5% (-11.2) LD 10.7% (-1) Others up by 3.4% UNS puts Lab on 310 seats (15 short) on this. Electoral Calculus says Lab 368 seats and a majority of 86 (with 50% tactical voting) or 46 seat majority with no tactical voting. |
|
|
|
Post by James E on May 13, 2023 10:12:37 GMT
petere Stay out 41% (+4), Rejoin 59% (-4) - "Surely, looks better for leave as they jumped 3%(4%)?" It's a reversion to the norm after a rather high 'Rejoin' lead in the previous Omnisis poll. The recent average is Stay Out 42%, Rejoin 58%. www.whatukthinks.org/eu/opinion-polls/poll-of-polls-uk-eu/There are now significant numbers of those who voted Leave in 2016 and/or Conservtive in 2019 ( "Get Brexit Done") who now say that they would vote to Rejoin.
|
|
|
|
Post by James E on May 12, 2023 17:53:11 GMT
Omnisus, wow [ Lab 51% (+3), Con 24% (-3) ] Meanwhile, in this satirical parody known as the Daily Telegraph... "Rishi Sunak is the unexpected winner from last week elections. The spectre of a Labour coalition of choas...will terrify voters." /photo/1  |
|
|
|
Post by James E on May 12, 2023 10:51:38 GMT
I posted this a week ago, but here are the comparative figures for the LibDems' Local Election vote shares since 2017 (per BBC NEV) compared to their performance in the 6 preceding GB Westminster polls - updated the reflect the 2023 figures. 2023 20% in Local Elections v 11% in GE polls (+9) 2022 19% in Local Elections v 11% in GE polls(+8) 2021 17% in Local Elections v 8% in GE polls (+9) 2019 19% in Local Elections v 9% in GE polls(+10) 2018 16% in Local Elections v 8% in GE polls (+8) 2017 18% in Local Elections v 7% in GE polls (+11) The polls have generally predicted the LDs' vote share accurately for previous General Elections, so there is no reason to think that the customary LD overperformance in the Local Elections is a pointer for the next General Election. EDIT: To add the words of our former host on UKPR1 : "As you can see, the contrast between what the Liberal Democrats manage in general election voting intention polls and what the Liberal Democrats achieve in local elections is not new - it has been consistent for the last twenty years." (Anthony Wells - writing in 2012) pollingreport.uk/articles/lib-dem-support-at-local-elections |
|
|
|
Post by James E on May 12, 2023 9:00:20 GMT
|
|
|
|
Post by James E on May 11, 2023 8:04:28 GMT
Pjw1961 I had a quick look at the ward by ward analysis for Redditch and, to be honest, found it a little baffling at first sight. Big jumps in Labour percentage vote share from 2019, particularly in the five wards they gained from the Tories, but no reciprocal reduction in the Tory percentage vote share. That only dropped a few points in most cases and, at first glance, it wasn't immediately obvious where the Labour increased share had come from. Maybe some independent candidates had stood down from 2019. There was clear evidence, however, of the Lib Dem vote being squeezed by Labour. Put another way, a static Tory horse being overtaken by a turbo charged Labour horse that had been given additional legs by a hitherto rival stable. That is the same pattern as I found when comparing the Rushmoor (Aldershot) results to the 2019 LEs. Con 37% (-5) Lab 47% (+17) LD 14% (+3) Others 2% (-15) The reason was mostly the decline of UKIP who were taking around 14% in GB polls in early May 2019. And it's worth remembering that the Tories were at a low ebb at that point, too, taking an average of around 28% in Westminster polls, and the same share per NEV in the local elections. Of course, it is highly unlikely that these net movements reflect the actual churn in each party's 2019 voters. So it does not mean that Labour won their voters from UKIP or others. |
|
|
|
Post by James E on May 10, 2023 18:11:32 GMT
On the Britain Elects/New Statesman 'ward by ward' analysis. Having read the article and, in particular, studied the table of 69 currently Tory constituencies, it looks like we are to some degree heading back toward pre-Brexit voting patterns, although not totally so. For example in Thanet South (Labour target 111, but not a 'traditional' Labour seat, won only in the Blair landslides) Labour lead by 4%, but in Great Grimsby (target 114, but a traditional Labour area) Labour are 10% ahead. Likewise Labour have stonking great leads in other traditional Labour seats such as West Bromwich East (26%), Bolsover (19%), WAkefield (25%), Burnley (19%). Labour has a 9% lead in Cannock Chase - target number 275! However, it is clear that some change is also going on. From example Labour 'won' Aldershot 41% to 35%, a seat even Blair couldn't take. Of the 69 seats examined Con win 33 (they are not all marginals, there are many safe seats included), Labour win 24, the Lib Dems 9 and the Greens 3. In case you are wondering where the Green ones are: they are 35% to 35% with the Tories in Bury St Edmunds, 28% to 25% ahead in Folkestone and Hythe and 31% to 25% in Forest of Dean, so probably unlikely they would actually win any of those in a GE. I did post the Aldershot result here on Saturday, pjw1961 - see page 49  . Labour won 60% of the votes cast in Aldershot itself, to the Tories 34%, but the other parts of the constituency are more favourable for the Conservatives. The nearest the Labour have been before was 17% behind the Tories in 2001. It was one of those seats where the LDs and Lab were competing to be the main challengers to the Tories in both 1997 and 2001. The 2019 Tory majority was 35%, so the LE result would be a swing of slightly over 20%. |
|
|
|
Post by James E on May 10, 2023 9:39:50 GMT
Re: 'Actual' vote shares in the Local Elections.
In 2019, when the same seats were fought, the actual shares of Con 32%, Lab 29% were reflected in the BBC's NEV shares as 28% for each of Con and Lab.
So if the same pattern were applied to the BBC's advertised NEV, it would have been Con 30%, Lab 36% in actual votes cast.
|
|
|
|
Post by James E on May 9, 2023 18:14:53 GMT
Some details from Deltapoll as to where Conservative voters from 2019 now stand:
They are still retaining more of their Remain-voting supporters from 2019 than their Leavers. However, they are also losing more Remainers to Labour. The different cross-breaks, averaged from the past 5 Deltapolls are:
Con-Leavers
Con 64%
Lab 18%
LD 3%
Ref 10%
Grn 1%
Con-Remainers
Con 70%
Lab 22%
LD 5%
Ref 1%
Grn 1%
So they're retaining a net 48% of Con-remainers against Labour, as opposed to 46% of Con-leavers.
AND:
"And if there were a second referendum on British membership of the European Union, how would you vote?"
Figures for Conservative 2019 voters (sample of 525) excl DK and WNV.
Stay Out 70%
Rejoin 30%
|
|
|
|
Post by James E on May 9, 2023 16:11:25 GMT
R&W's so-called Blue Wall is a key battleground between Con LDs and Lab . The 42 seats polled include nearly all the LDs' targets in London and the South, plus about 15 Lab targets.
R&W's last 'Blue Wall' poll was unusually good for the LDs and this rather masks the fact that their 23% still an improvement the 20%ish they had been averaging in the same seats. Fieldwork date is given as 7May (Sunday) so should reflect any post Local Elections effect.
Compared to GE2019, this one is:
Lab 36% (+15)
Con 32% (-18)
LibDem 23% (-4)
Swing Con>Lab 16.5% compared to around 13% in recent GB polls.
|
|
|
|
Post by James E on May 9, 2023 10:46:20 GMT
Shevii, Personally I think 35% is unlikely but early 30s likely and 33% possible. What would Labour need to achieve to get an OM with Tories on 32% for example? UNS not that accurate and will it be 7 or 15 gains in Scotland? Using Electoral Calculus , I have tried figures of Con 33, Lab 36, LD 16, Ref 4, Grn 6. I have added in tactical voting of half of each party's voters. This produces an 18-seat Labour majority on a 3 point lead (and a 7% overall swing). Labour would be 103 seats ahead of the Tories in this scenario, whereas UNS would put the two main parties close to level in seats. I think EC is somewhat generous to Lab in Scotland at the moment, as it shows 20 SNP losses, presumably to Lab. While EC clearly has its faults, it looks to me far closer to the likely outcome than the UNS used by Polling Reports or Sky News's recent forecast of a hung parliament. The effectiveness of tactical or 'ABT' voting is reduced somewhat as the Con vote rises, so it is not possible to give a firm figure for the kind of lead Lab need for an OM. My best guess, based on a Tory VI close to 35% would be around 4-5 points. |
|
|
|
Post by James E on May 8, 2023 13:28:17 GMT
For a post Local Elections boost for the LDs, see the polls before and after those of 2 May 2019 - linked below. LibDem VI Before Local Elections: 7, 10, 8 6, 8,11,13: Average 9% Lib Dem VI after Local Elections: 11,16,14,18, 11,13,15 : Average 14% Most of this came from Labour who were down by 3 points over the same polls. en.wikipedia.org/wiki/Opinion_polling_for_the_2019_United_Kingdom_general_electionLooking at more recent years, they got a 2 point boost in 2021, and 1 point in 2022. |
|
|
|
Post by James E on May 8, 2023 12:35:42 GMT
@trevor My view is based on talking to my neighbours, people in the area and election candidates. I've lived here for 30 years and it's clear which way the wind is blowing. edit: everyone here knows Labour have zero chance so the ABT vote won't be split. - your view that local people understand the tactical situation and take it into account is psephologically sound. EC's calculations, being polling based, are much too generous to Labour in areas where the Lib Dems are clearly the more likely competitor to the Tories, as I have been saying for some time. Here for example, is EC's 'prediction' for Tiverton and Honiton: Con 38.9%, Lab 34.1%, LD 11.1%, Green 6.2%. In reality, when put to the test in the by-election the Lib Dems won it at a canter. Here is an even dafter example: They have Central Devon being won by Labour 37%, Con 35.9%, LD 11.8%. In the Mid Devon council elections the Lib Dems won 33 of the 42 seats and Labour won none. There is simply no basis of Labour strength in that area on which to produce a result like EC's prediction, which is frankly garbage. Your gut instinct is likely to be much closer to the truth than EC's flawed methodology. Mole Valley looks like a decent LibDem longshot, as they are in a strong 2nd place as of 2019. Devon Central is different though, as Labour held 25% to the LDs' 15% last time. I seem to remeber Davey saying that he would not be targeting those seats where the LDs were behind Labour in 2019. |
|
|
|
Post by James E on May 7, 2023 19:42:22 GMT
I understood that they took 6% of the Local Election vote, in the 6% of the seats where they stood. This would mean that they received about 0.36% of all the votes cast. |
|
|
|
Post by James E on May 7, 2023 10:50:35 GMT
re: Sky news projection - Lab 36%, Con 29%, Lab 7 points ahead, Lab 298 seats, Con 238. That looks very much like what you get with a uniform national swing from the Tories 11% ahead to Lab 7 % ahead. A simple 9 point swing delivers 94 seats from Con to Lab. But this seems to overlook the very varied pattern in Thursday's Local Election results, with Lab and LibDems vastly outstipping UNS in seats where they are best placed against the Conservatives. Perhaps I should clarify my point here - especially in the light of Trevor's comments about "assumptions". I am not making assumptions here. What I am doing is commenting on how votes/swings can translate into seats. In the Local Elections, the BBC's NEV percentage show swings of just 4% from Con>Lab and 1.5% Con>LD when compared to the 2019LEs. The council seats changing hands would have been in the low hundreds if these movements had applied uniformly, so in these elections UNS totally fails as a forecasting tool. So when Sky news offer us a seats projection which is clearly based on UNS, I think we're entitled to query it. Recent General Elections results have diverged widely from UNS too, and I expect 2024 to be no different. And any calculation which anticipates the LDs doing as well in a General Election as in Local Elections is absurd - their LE performance is typically around 7-8 points above their GE polling - though this makes very little difference to the Lab seats total as there are so few Lab/LD contests. There are two further factors to bear in mind: regional differences in polling results, and (especially) tactical voting. Luckily, Electoral Calculus already reflects the regional pattern of current polling, and can be used to incorporate tactical voting. I've used it to show a 9% swing, using figures of Lab 42%, Con 35%, LD 12%, Green 5%, but with half of the Lab/LD/Green supporters being willing to switch tactically. EC produces figures of Lab 372, Con 204, LD 23 on these figures. Very different from the Lab 298, Con 238 produced by Sky news. Both reflect a 7% Lab lead, and a 9% Con>Lab swing from 2019, but I'd be very confident that EC will be closer to the actual seat totals if that actually happened at the GE. |
|
|
|
Post by James E on May 6, 2023 21:22:10 GMT
James E, what we can't ever know is what size the 'discounts' are from the current GE levels of support for the 2 main parties. I asserted with no evidence that the Lab discount would be lower than the typical 10% and the Tory one larger. That 36/29, though, seems consistent enough with Lab 43/45 and Tories 28/30 in current polling. Omnisis is not the best benchmark I believe but some uptick in VI polls for all but the Tories, and a fall for them, may well occur as it often does in these situations but typically wanes over a month or so. NB) Smug about my 6-8% prediction a couple of weeks ago. When I've looked into the LibDems' polling performance around Local Elections, it's been obvious that they have had a boost every year, following the news of their good LE performances. It was because of this that I made my comparative figures reflect their VI figures before the LEs in each case rather than after. I'd certainly expect the same this time, but would not be surprised if that boost came as much from Lab as Con. |
|
|
|
Post by James E on May 6, 2023 20:43:20 GMT
Very interesting results in South Norfolk where Labour has performed far better than it did in 1995 in the peak Blair year. The LDs have not recovered their former strength there , and in many wards are no longer the obvious anti-Tory challenger. , you may remember that I pointed out 6 weeks ago that South Norfolk has become increasingly promising for Labour over the past couple of elections, and was now their most viable target in rural Norfolk. See from 26 March 5:42pm and your own response at the bottom of the same page. ukpollingreport2.proboards.com/thread/57/nov-2022-lab-con-ldem?page=402 |
|
|
|
Post by James E on May 6, 2023 20:02:15 GMT
re: Sky news projection - Lab 36%, Con 29%, Lab 7 points ahead, Lab 298 seats, Con 238.
That looks very much like what you get with a uniform national swing from the Tories 11% ahead to Lab 7 % ahead. A simple 9 point swing delivers 94 seats from Con to Lab.
But this seems to overlook the very varied pattern in Thursday's Local Election results, with Lab and LibDems vastly outstipping UNS in seats where they are best placed against the Conservatives.
|
|
|
|
Post by James E on May 6, 2023 9:26:44 GMT
I have checked the aggregated the results for my old home-town of Aldershot (Rushmoor Borough) where each of the 13 wards elected a councillor on Thursday. The council stays as a Con hold, as it elects by thirds, but it must count as one of Labour's greater successes in a seat which which has been held by the Tories since it was created in 1918. They currently hold it with a 35% majority, and it's Lab target number 217. Here's how it voted on Thursday ,with comparison to the GE2019 percentage votes. Lab 9,886 46.8% (+23) Con 7,822 37.1% (-21) LibDem 2,997 14.2% (-) Oth 1.9% www.rushmoor.gov.uk/your-council/elections-and-voting/local-elections-4-may-2023/election-results/EDIT I should have remembered that Aldershot Constituency also includes two wards from Hart Council. These voted heavily for the LDs with Lab in 3rd place. With those added in, Aldershot Constituency voted like this: Lab 40.8% (+17) Con 35.3% (-23) LD 21.9% (+7) Others 2% Swing Con>Lab 20.2% (in comparison to GE2019) |
|
|
|
Post by James E on May 5, 2023 16:16:56 GMT
As I write the number of Conservative losses is 666, which sounds a bit ominous for Sunak! Now 741 losses. That's 75 gone in 15 minutes. |
|
|
|
Post by James E on May 5, 2023 14:50:48 GMT
If Labour's Net Equivalent vote is 35%, then that's 9 points below their current polling average of 44%.
This is not unprecedented. Labour's NEV was 10% below the polls in 2013, when they recorded 29% against poll-ratings of 39%. And as I mentioned earlier today, Labour's NEV was 9% below the polls in 1995 and 10% below in 1996. The pattern is that the higher Labour are polling, the wider the difference to English Local Election results. And the average, across good and bad years, is around 4-5% below.
Nevertheless, 35% is a bit disappointing. I had expected around 37%, as per the Omnisis LE poll a few days ago.
|
|
|
|
Post by James E on May 5, 2023 14:03:40 GMT
To me, the relatively low Lab lead over Con is most likely caused by the lack of any real competition on the right. So far, neither RefUK nor any other alternative party on the right was won a single seat, and indeed they do not appear to have contested many. Residents' Assocs have won 17. Contrast this to the 200+ for the Greens.
I don't think the exclusion of London, Scotland and Wales from these elections is a factor, though. This would affect the actual total Labour vote, but not the 'NEV'. Labour's polling success over the past year-or-so has been larger in 'England-outwith-London' than in any of those places.
|
|
|
|
Post by James E on May 5, 2023 9:53:12 GMT
pjw1961 Labour under-performance in Local Elections - here are the comparative figures: 2022 35% v 39% in GE polls (-4) 2021 29% v 34% in GE polls (-5) 2019 28% v 29% in GE polls (+1) 2018 35% v 39% in GE polls (-4) 2017 27% v 29% in GE polls (-2) I should add that the Conservatives underperform in LEs, too. Sometimes by more than Labour. And there is clearly a pattern that both Labour and the Conservatives underperform more when they are doing well. As an historical example of this, Labour underperformed by 9% and 10% in 1995 and 1996 when enjoying those huge polling leads under Blair. Thanks James. So I won't over-egg it, but it would be not unreasonable on those figures to add - say - 4% to the notional national vote based on the local results. Since that looks likely to be in the region of 36%-ish (we won't get the figures until tomorrow), that would suggest Labour on c40%. There is also a more recent precedent for Labour underperforming by 10 points - the 2013 Local Elections where they got 29% NEV v 39% GE poll average. en.wikipedia.org/wiki/2013_United_Kingdom_local_electionsThis might have been an early sign that the polls were overstating Labour - as they did by 3 points in GE2015 polls. But even if we allow for that error, it would still be a performance 7 points below Labour's (adjusted) polling level. So an NEV of 37-38% would be compatible with the current Average Lab VI of 44%. |
|
|
|
Post by James E on May 5, 2023 9:20:08 GMT
pjw1961Labour under-performance in Local Elections - here are the comparative figures: 2022 35% v 39% in GE polls (-4) 2021 29% v 34% in GE polls (-5) 2019 28% v 29% in GE polls (+1) 2018 35% v 39% in GE polls (-4) 2017 27% v 29% in GE polls (-2) I should add that the Conservatives underperform in LEs, too. Sometimes by more than Labour. And there is clearly a pattern that both Labour and the Conservatives underperform more when they are doing well. As an historical example of this, Labour underperformed by 9% and 10% in 1995 and 1996 when enjoying those huge polling leads under Blair.
|
|
. Labour won 60% of the votes cast in Aldershot itself, to the Tories 34%, but the other parts of the constituency are more favourable for the Conservatives.
