8-Dec-2021 : LAB: 40 / CON: 34 / LDEM: 10 / GRN: 4

[graham]

Dec 26, 2021 23:23:25 GMT James E said:

Graham

Focaldata's constituency figures all show Labour making greatest progress wherever there was a large LD vote in 2019.

So in South Cambridgeshire (which was Con 46%, LD 42% in 2019), they show Labour up 14 points from 12% to 26%.

And in the LD/SNP contest of Orkney and Shetland, they show Labour nearly trebling their vote share from 6.7% to 18%.

If you can find a seat where the LD's did well in 2019, and Focaldata do NOT show Labour making more than 7 points* progress, then please let us know.


* the total vote shares from Focaldata show Lab up from 33% at GE2019 to 40%

Much of the increase in the Labour vote until the last couple of months is likely to have come from voters who switched to the LDs over Brexit and Corbyn in 2019. More recently,though, there is a fair bit of direct switching from the Tories - with some who had switched to the Greens returning too. If the methodology used is simply a crude formula which takes no account at all of local circumstances, then I do agree it will simply not work much of the time.

Re - Kingston & Surbiton. There is some evidence of a campaigning machine there in that - for some reason - Labour strongly outperformed in 2005!

[shevii]

Dec 26, 2021 23:08:43 GMT graham said:

Dec 26, 2021 22:43:34 GMT shevii said:

Labour has polled circa 15% there at recent GEs- albeit lower in 2019. A 20% vote share there is not beyond the bounds of possibility.

I think that without an LD sitting MP it is not impossible that there would be a latent Lab support of 25% there (and more with a two horse race Con/Lab squeeze) in much the same way that Wimbledon was won by Labour in 1997 and 2nd place is currently being contested.  Realistically though it seems very unlikely Lab would be anywhere near 20% on current optics and Davey incumbency.  In 2017 there was clear water between Lab and LD on policy but currently probably less than at any time apart from pre 1997. LD having been in coalition and the Corbynista factor probably led to more stickiness in the Labour vote in 2017 that got them that 15%.

At best it seems that focaldata MRP is predicting that sort of result if no-one had a clue who was the best ABT or had no reason to vote ABT because of PR or whatever.

[tancred]

Dec 26, 2021 19:24:56 GMT domjg said:

Dec 26, 2021 11:57:23 GMT tancred said:

I listen to Classic FM - not as often as I used to though. The yearly "Hall of Fame" seems to have become a brexiter's dream, with huge numbers of entries by Elgar and Vaughan-Williams, which, though undoubtedly composers of great talent, do not compare with the very greatest of composers. Furthermore, the "Lark Ascending" winning the competition repeatedly year after year seems to confirm how parochial and nationalistic even fans of classical music are in this country. I suspect that those years when Rachmaninov and Tchaikovsky won it was probably the handiwork of Putin's agents!

tancred If you’re into classical music in this country Radio 3 is the only game in town. I’ve a lot of time for Vaughn Williams though he does seem to have been hijacked of late. He was unconventional and curious. My choir did his Fantasia on Christmas carols as part of our Christmas concert. Mr syrupy/bombast Elgar on the other hand can go do one..

Well, we all have different tastes.  Radio 3 often focuses on more obscure classical works, which is fine, but I prefer to listen to what I like.  I actually like much of Elgar but he has been hyped up enormously just because of his Englishness, which grates on me.  Karajan once said that Elgar was second rate Brahms, which was a bit cruel, but not too far from the truth.

[graham]

Dec 26, 2021 23:40:29 GMT shevii said:

Dec 26, 2021 23:08:43 GMT graham said:

Labour has polled circa 15% there at recent GEs- albeit lower in 2019. A 20% vote share there is not beyond the bounds of possibility.

I think that without an LD sitting MP it is not impossible that there would be a latent Lab support of 25% there (and more with a two horse race Con/Lab squeeze) in much the same way that Wimbledon was won by Labour in 1997 and 2nd place is currently being contested.  Realistically though it seems very unlikely Lab would be anywhere near 20% on current optics and Davey incumbency.  In 2017 there was clear water between Lab and LD on policy but currently probably less than at any time apart from pre 1997. LD having been in coalition and the Corbynista factor probably led to more stickiness in the Labour vote in 2017 that got them that 15%.

At best it seems that focaldata MRP is predicting that sort of result if no-one had a clue who was the best ABT or had no reason to vote ABT because of PR or whatever.

For whatever reason, Davey - despite being the incumbent - lost vote share in both 2005 and 2010. At both elections he underperformed his national party's result. He is ,of course, very much an Orange Booker- a label which perhaps alienates him from left of centre voters who might be prepared to consider voting tactically. In terms of macroeconomic policy there is actually quite a big gap between his approach and what Labour would be likely to offer. Indeed his insistence on fiscal prudence etc places him to the Right of Johnson and makes him appear much more in the mould of Sunak and Osborne et al. For that reason, I can well imagine that many local Labour voters would not be attracted to him.

[tancred]

Dec 26, 2021 23:41:38 GMT James E said:

Dec 26, 2021 23:35:31 GMT tancred said:

Major went the distance because he was not challenged and replaced - it seems that all the ones who could have challenged him stayed loyal.  

en.wikipedia.org/wiki/1995_Conservative_Party_leadership_election

Apologies - I had forgotten this.  Anyway, Redwood back then was about as appealing as Rees-Mogg is today so it's hardly surprising that he lost badly.

[graham]

Dec 26, 2021 23:59:08 GMT James E said:

Graham

The more you look at the constituency figures from Focaldata, the more obviously crude they look. A couple more examples:

1. There is not a single constituency where the Tories are shown as getting more than 49% of the vote.

2. The Green Party would not lose a single deposit in England per these figures as they are on 5% or more in every English constituency.

These two don't effect the number of seats for each party, but they do help illustrate just how formulaic the whole thing is. And where their methodology gives us results with Plaid losing 25% of their vote share in every single Welsh constituency and the SNP losing around 15% of their vote share in every single Scottish constituency, this does matter.

I wonder if The Times (who commissioned the Focaldata MRP) have noticed these anomalies.

That does look odd - and fails to instill confidence!

[eor]

While we're kicking Focaldata, I'd add that their MRP missed about 30 CON gains in their final call projection for GE 2019, and (worse IMO) they hedged its presentation massively by saying that 67 seats were too close to call. In projection terms that's akin to saying "Our model is extremely confident an election is happening".

www.reuters.com/article/uk-britain-election-focaldata-idUKKBN1YE2LA

[oldnat]

Thanks to those that have analysed the focaldata recent stuff and suggested that their constituency predictions outwith the PSR of England are crap.

I've been too busy celebrating the release of son and family from Covid jail.

Naturally, I am wholly devoid of political bias (as are all on here), but I do note that the NHS Scotland PCR tests conducted on Xmas Day show them to be devoid of the plague, while the private company, who are mandated by the corrupt regime based in Westminster to conduct PCR tests to visitors from these shores the day before, found the only one not holding a UK passport to be a threat to the realm.

While the 25th December's arrangements were disrupted, that day is but a selection by a religious cult within the 12 days of winter solstice celebration - and one that was spurned by my ancestral lineage a mere 2 generations back.

The real celebration of the rejuvenation of light has now restarted among us pagans!

[tancred]

Dec 26, 2021 22:28:23 GMT eor said:

Dec 26, 2021 11:57:23 GMT tancred said:

I listen to Classic FM - not as often as I used to though. The yearly "Hall of Fame" seems to have become a brexiter's dream, with huge numbers of entries by Elgar and Vaughan-Williams, which, though undoubtedly composers of great talent, do not compare with the very greatest of composers.

Not sure the data is on your side there! Beethoven alone has more top 20 entries than those two combined. And the (to me, utterly unfathomable) Lark ascended to the upper echelons about a decade before Brexit hove into view, with Elgar's popularity at best described as static over the years.

Didn't Elgar get something like 14 entries?  Seems bonkers to me - 5 or 6 I could understand. I'm not saying that classical music popularity is related to Brexit but there seems to be a parochialism in musical taste in England at least. 

[oldnat]

Opinium Scottish poll

"Do you think the Scottish Government or the Westminster Government have done a better job of dealing with the Covid-19 crisis?"


Scottish Government - 58%

UK Government - 10%

Both the same - 24%

Don't know - 8%

In terms of factors affecting VI, it really doesn't matter whether any of these perceptions can be objectively justified (and how would you do that anyway?)

It does suggest, however, that the Unionist (and particularly Tory) suggestion that Scotland should ape England in its response to Covid, is a less than brilliant stratagem.

[Deleted]

There is not enough information about the methodology of that large sample poll.

It is possible that they did it, but I would have tested the hypothesis of all parties have equal chance, adjusting the chances after using 5,000 samples, and then using those probabilities with the next lot and so on.

I know it is Bayesian, but using it would give a point at which we.can see if the result is.settled or mobile (Bayesian factor).

[eor]

Dec 27, 2021 0:35:13 GMT oldnat said:

Opinium Scottish poll

"Do you think the Scottish Government or the Westminster Government have done a better job of dealing with the Covid-19 crisis?"


Scottish Government - 58%

UK Government - 10%

Both the same - 24%

Don't know - 8%

In terms of factors affecting VI, it really doesn't matter whether any of these perceptions can be objectively justified (and how would you do that anyway?)

It does suggest, however, that the Unionist (and particularly Tory) suggestion that Scotland should ape England in its response to Covid, is a less than brilliant stratagem.

I'd love to see a poll asking Scottish voters whether they think the Scottish Government or the Westminster Government have done a better job of ensuring the sun rises in the East tomorrow. 

Tho if your point is that any conscious copying of Johnson's COVID approach would be politically unrewarding (regardless of the actual measures involved) then I could wholly believe that is shrewd advice!

[eor]

Dec 27, 2021 0:21:01 GMT tancred said:

Dec 26, 2021 22:28:23 GMT eor said:

Not sure the data is on your side there! Beethoven alone has more top 20 entries than those two combined. And the (to me, utterly unfathomable) Lark ascended to the upper echelons about a decade before Brexit hove into view, with Elgar's popularity at best described as static over the years.

Didn't Elgar get something like 14 entries?  Seems bonkers to me - 5 or 6 I could understand. I'm not saying that classical music popularity is related to Brexit but there seems to be a parochialism in musical taste in England at least. 

Nah. Elgar has 9 in the top 300 this year, but quite a few of those are lurking down with the video game tracks and that. Comparable number for Mozart is 16. 

It flits a little from year to year but this year Elgar is joint 5th with Chopin and John Williams, behind Mozart, Beethoven, Bach and Tchaikovsky. And in terms of say the top 20 it changes relatively little from year to year - Beethoven and Tchaikovsky will have about 8 entries between them, which considering each voter has only 3 (weighted) votes is remarkably broad support, which I suspect harms their chances of hitting the top spot. That most of the top spots have been claimed by Vaughan-Williams, Rachmaninov and (in the early years) Max Bruch reinforces the idea that the voting system favours those with one or two very identifiable pieces. 

Maybe if they could be like the Scottish folk and do all the game theory of how many entries to submit... ;-)

The biggest potential Brexit impact I've seen has actually been the other way - the first poll done after Brexit actually happened saw Beethoven's 9th (home of the EU anthem) surge almost to the top!   

[statgeek]

Graham / Oldnat

Get your predictions in for by-election in Preston, Seton and Gosford (East Lothian) on the 20th of January 2022. Provisionally held by Lab, I'll go for a Lab hold on a lower share, with the SNP getting closer than in 2017. What say you?

As for the SNP on 37% akin to 2017, meh! They went from 37% to 45% in 2019, while Lab went from 27% to 19% (both with Corbyn). Other than sympathy for the late Lab councillor, or ultra-Unionists flocking to Lab, I can't see what pull Lab have that they didn't have in 2017. Not run by Corbyn shouldn't matter in council by-elections. It'll maybe give a hint as to the local footprint changes (or not) of the people voting in an election that doesn't get had at by the media.

Turnout should be ... poor? January and Covid etc. Only the ultra-political? Should prove interesting. Election for one seat in an election that usually selects four, so it's really for Lab to lose.

Other things to consider.

East Lothian Constituency (Holyrood)
2011: Lab 39.0%; SNP 38.5%
2016: Lab 27.8%; SNP 34.8%
2021: Lab 36.7%; SNP 39.2%

East Lothian Constituency (Westminster)
2010: Lab 44.6%; SNP 16.0%
2015: Lab 31.0%; SNP 42.5%
2017: Lab 36.1%; SNP 30.6%
2019: Lab 29.5%; SNP 36.2%

Of course, we're not seeing ward-level data, but for what's it's worth, Ward 5 (Haddington & Lammermuir) had a wee by-election in 2019, and if we compare to 2017:

Ward 5 (1st prefs of combined candidates in 2017 - single candidate in 2019):
2017: Lab 33.7%; SNP 26.1%; Con 29.0%
2019: Lab 21.5%; SNP 29.5%; Con 35.0% (Con was elected at stage five with five parties participating - Basically came down to SNP vs Con)

We can take a guess that Boris was keeping the Con vote high, either as pro-Tory, pro-Brexit, or pro-Union. All of that is past. God know how they would vote now. Less likely to vote SNP, but some will.

So get yer predictions in.

[alec]

An effective attack by Labour this morning on the issue of energy prices. The government under pressure from all sides, and Labour highlighting the large (£3bn, allegedly) windfall accruing to Sunak from VAT on fuel bills, and calling for VAT to be scrapped on domestic gas and electricity bills.

Again, this is a sign of a more effective and nimble opposition, I feel. However, neither Labour nor the media appear so far to have related this to the great policy failure of Brexit as yet. The aboloshion of VAT on domestic energy bills was an overt promise made by Johnson in the referendum campaign, yet as with all the Brexit promises, when the time came to enact these great reforms, Sunak flunked it.

Another Brexit dream dismantled.

[neilj]

The cost of living crisis, which will become more dominant in terms of people's awareness, will certainly be fertile ground for Labour to attack the Conservatives in the coming couple of years

[hireton]

This looks like another MRP poll this time from Survation working for 38degrees:

mailchi.mp/survation/new-survation-research-projects-111-conservative-seat-losses

Bad news again for the Tories but unlike Focaldata they do not forecast significant gains for Labour in Scotland.

Not surprisingly, Scotland is the least convinced about the integrity of the UK Tory government as assessed by adherence to the 7 Nolan Principles.

Analysing the methodology is way beyond me but perhaps James E will take his scalpel to it to see if it is more convincing than Focaldata.

[steve]

EOR/OLDNAT

When a regime and its leadership are relatively popular polling normally tends to suggest that the public give them a pass on all aspects of their activities.The converse is equally true.
On a purely objective basis both the Westminster and Scottish government handling of the covid pandemic have been similarly relatively crap.
The primary difference is that the Scottish government have been both more transparent in their presentation and less mired in the obvious sleaze and hypocrisy of the Spaffer regime.

However in terms of effectiveness there is virtually no difference and in respect of approach in terms of restrictions the differences are marginal and virtually non existent when it comes to comparison between the SNP the Labour party suggestions.

Both the UK government and Scottish and other devolved administrations have spent the last 21 months following the pack in terms of response sometimes a few days or week apart and sometimes with different degrees of over reaction.All have inflicted substantial damage to their economies ,in addition to that caused by the stupidity of the brexitanian regime which I readily concede isn't the Scottish, or Welsh governments fault. All have put covid priorities above and to the detriment of other public health concerns.
All are relative failures.
All. have opted to restrict international travel to try and keep out a virus that's already in community circulation, a policy that the WHO rightly regards as pointless and all have seen a significantly higher mortality rate than many comparable countries.

While old nat may not agree with my analysis of the comparative performance I suspect that he would concede that there is an element in Scottish opinion that tends to assume a better response to virtually anything can be found in Scotland compared to Westminster.

[steamdrivenandy]

Dec 27, 2021 0:14:00 GMT oldnat said:

While the 25th December's arrangements were disrupted, that day is but a selection by a religious cult within the 12 days of winter solstice celebration - and one that was spurned by my ancestral lineage a mere 2 generations back.

The real celebration of the rejuvenation of light has now restarted among us pagans!

Perhaps I'm biased but I've always preferred the twelfth day. Somehow it seems lighter and more rewarding. Sadly though its music is fairly boring, all those carols etc based around the 25th, when all the 6th gets is 'Happy Birthday To You' for the 74th time.     

[crossbat11]

I'm second to nobody in my admiration of JamesE's and Graham's psephological prowess and, more seriously, their contributions on all matters polling are a valuable addition to the site, but I offer a thin defence of the pollster Focaldata. Their MRP, by the sound of it, is flawed but their headline VI figures look pretty much in line with other pollsters.

They must be getting something right in terms of overall methodology! I thought I'd come to their weak defence before we binned them entirely as a voodoo pollster!

[crossbat11]

Reference my comment above, it should be remembered that The Thunderer, aka The Times, don't often waste their money, or risk their reputation, on voodoo polling organisations.

Now the Daily Express in the other hand.........🥴😂

[Deleted]

Dec 26, 2021 13:48:38 GMT robert said:

Colin



What worries me is that, like the extreme left of Labour, the extreme right of the Tories believe that the reason for their woes is the fact that their policies are not extreme enough. Labour under Starmer and with a competent reshuffle behind him, is demonstrating that the extreme left is wrong, on that score. Competence is what counts too. Who is there in the Tory party to tread a similar path (there are obviously one or two candidates) but will they have a chance to be elected new leader, when a third of the party can be said to be more extreme than Thatcher ever was?

Yes - I share that concern.

Even if a Hunt type candidate got through a Leadership Challenge to the last two, Popularity scores on Con Home indicate the members will go for a Truss type.

The Steve Baker faction were not in line with public opinion on Covid controls-Johnson was.

That is the tragedy of the man-right there.

[crossbat11]

The reason I'm dividing my posts into separate segments is to accelerate my progress to Full Member status.

A trick I've learnt from....

[crossbat11]

......that resident site scoundrel......

[crossbat11]

......Crofty.

[Deleted]

Dec 27, 2021 8:27:43 GMT neilj said:

The cost of living crisis, which will become more dominant in terms of people's awareness, will certainly be fertile ground for Labour to attack the Conservatives in the coming couple of years

Absolutely-with Energy stage centre.

The boss of my utilities supplier has told the government to get its finger out:-


www.bbc.co.uk/news/uk-59797224

I think his phraseology-"enormous crisis": is on the money.

Look at any Fixed Deal offering just now-they are eye watering-huge increases on 2021. So everyone will be going to Variable Tariff and hoping the Cap protects them.  I am.

But how long before more failures start appearing in the more established companies ,if the Cap puts the whole burden of price increases on them?. 

One way or another Sunak has a problem here. How much pain can the cap protect consumers from , and how much will the support costs be for Sunak to pick up the pieces at the supllier end ?

[Deleted]

Dec 26, 2021 23:59:08 GMT James E said:

Graham

The more you look at the constituency figures from Focaldata, the more obviously crude they look. A couple more examples:

1. There is not a single constituency where the Tories are shown as getting more than 50% of the vote.

2. The Green Party would not lose a single deposit in England per these figures as they are on 5% or more in every English constituency.

These two don't effect the number of seats for each party, but they do help illustrate just how formulaic the whole thing is. And where their methodology gives us results with Plaid losing 25% of their vote share in every single Welsh constituency and the SNP losing around 15% of their vote share in every single Scottish constituency, this does matter.

I wonder if The Times (who commissioned the Focaldata MRP) have noticed these anomalies.

This sort of stuff is way beyond my competence. But assuming you are correct-can you reach  any conclusions  about the effect on their predicted  seat share?

And are you saying the flaws attach only to constituency outcomes-or to VI % as well ?

 

[steamdrivenandy]

It looks like Survation gave the Site Summary table to the office junior to do.

Somehow the SNP started off with 48 seats, gained 7 and lost 1, but ended up losing a net 6 and the total went up by plus 6. It seems a mystery floating minus sign is to blame.

Also by my arithmetic the Tory total should read 254 not 255 and the Labour 310, not 309.

As to LD 11 minus 6 has always equalled 5 in my experience, not 9.

And on 15th Nov we apparently would have had 631 MP's and on 22nd December we had 632, when, of course the HoC consists of 650. If the balance between 631/632 and 650 are 'others' why does the 'others' line in the table not have any figures?

[Deleted]

Dec 27, 2021 9:41:34 GMT crossbat11 said:

......Crofty.

Still on 96 though ??

The site computer must be on holiday .

[crossbat11]

Dec 27, 2021 9:43:30 GMT James E said:

@ Crossbat11

Yes, there is no obvious problem with the polling aspect of Focaldata's figures. Their 8 point lead is close to other pollsters, who are showing Lab leads of 5-9 points. But the big drops in support for the SNP and Plaid - despite the former having almost the highest vote retention (85% to Lab's 86%)- show an obvious methodological error.

Unless, of course, the polling really does show close to Zero tranfers from Con/Lab/LDs to the SNP (or to PC in Wales).

The points you and Graham have made about the Focaldata MRP, and the seemingly glaring flaws contained therein, do once again highlight how pollster methodologies still appear to be a moveable feast. We used to discuss this at some length on the old UKPR site, and Mr Wells offered his own erudite views too, but whereas pollsters can't control how layman pundits interpret their data and findings, one would have thought a fairly common methodology would have been settled upon by now that had learned from both past best practice and mistakes.

In other words, why would a seemingly reputable pollster still be making what appear to be such fundamental errors in methodology?