Much better. Good to have the GE now based on 'a bird in the hand'.
Which is how I put it the other day IIRC but from the Labour viewpoint. Two in the bush beats one in the hand if the one in the hand smells like shit. Hence why they would not have gone for a GE now if not forced into it by the LD/SNP initiative.
Haha even a 11% lead isnt enough for a tory majority in your model? Really? What sort of lead would it take?
Con 36.6%, Lab 23.4% would give the Tories 327 seats. So a 13% lead would do it.
Remember there are a lot of SNP and LD seats. It's not just the lead over Labour.
Even with a 13% lead in my example, the Tories are still 8% points down on 2017 where they got 43.4%. Lead over Labour isn't everything.
Show you working please.....otherwise zero marks will be awarded.
That's OK
I'm happy to disclose my assumptions and and what they produce, in case they are of interest to anyone.
I'm not happy to publish my spreadsheet (don't know how) and anyway, your eyes would glaze over.
So I'll take zero marks.
Please can you disclose your assumptions? It genuinely doesn't make any sense to me when compared with 2015, which the polls are increasingly similar to, but better for the Tories.
Con 36.6%, Lab 23.4% would give the Tories 327 seats. So a 13% lead would do it.
Remember there are a lot of SNP and LD seats. It's not just the lead over Labour.
Even with a 13% lead in my example, the Tories are still 8% points down on 2017 where they got 43.4%. Lead over Labour isn't everything.
13% a bare majority? With a 13% lead there is hardly any tactical voting going on against the tories, you haven't taken into account that?
Con 36.6% gives them a 0.5 higher share than they got in 2015 With Labour 5.6% down on 2015
But you've got the Tories winning fewer seats than they did in 2015, despite a more than 3% swing from Labour to Tory? How?
He must be predicting masses of libdem gains from the Tories because of tactical voting.
But that doesn't make any sense because Boris has a massive lead on Best PM And although the survation constituency polling is in line with his models, that's only because they have much smaller national leads.
We need to see some constituency polling from the ones showing double digit leads.
As his numbers mean the Tories are winning Chorley, it would also mean that all of the Labour money spent on defending seats would have been spaffed up a firewall.....
My numbers show Labour winning Chorley by 1,400, as I said.
An interesting set of splits, and I wonder about the samples on that. Some particularly striking features:
1. Brexit Party have an enormous problem (but also potentially an opportunity) with female voters - close to 20% with middle aged men, close to zero with middle aged women.
2. Labour still doing well with younger groups, and Lib Dems less well than you might expect with a relatively youthful leader (albeit not in the age bracket) and a youth-friendly Brexit policy. I wonder if we might see that shift a bit over the campaign - Corbyn is not likely to be the "flavour of the month" choice he was in 2017 and there's an opportunity for Tories and Lib Dems there.
3. Interesting to see the Lib Dems leading (narrowly) with middle aged women. Perhaps not surprising given their leader is a fully paid up member of that group. Again, it will be fascinating to see if the Lib Dems build on that or other parties can bite into it - the retirements of some senior female politicians in that group does play into a narrative that might move the group towards the Lib Dems and other smaller parties.
If it were data from any other field than politics I'd be saying poor sampling to get such spiky data. I suppose it is just about believable for voting intentions, but particularly with the Brexit Party numbers it is surprising to see so much variability between groups.
That is interesting. I am pretty sure that is a reversal of 20+ years ago (anyone with data able to advise?), when the less educated voted Labour and more educated Liberal or Tory.
The LDs still do best with the most educated percentage of vote wise, the last time the Tories won graduates was 2015 but even then they did better percentage wise with non graduates.
Remember in the 1980s only 10% went to university ie becoming mainly higher earning, professional Tory or LD voters, now 30 to 40% go plenty of middle income or even low income Labour graduates too
Possibly, and also as Tories tend to be in the older age group the percentage of grads decreases with age. It could also be that educated people realise Brexit is a moronic policy, and the party now most associated is the party that used to describe itself as conservative.
1. "You're jealous". 2. "Stop being so old-fashioned. Next thing you'll be causing a winter of discontent and bringing back flared trousers". 3. "Stop being so class-prejudiced." 4. "Abolishing Eton etc. will cost the country billions."
I can understand some of the sniffy comments about Boris but the idea that he is "lazy" and "unfocused" is barmy. The guy is bursting with energy and utterly focused, even if the focus is entirely on him but, hey, he is a politician. Just look at the discipline of his leadership campaign.
If he managed to pull off a GE win (after 9 years of less than scintillating Tory Govt) that would objectively be a hell of an achievement. If he delivers Brexit then he becomes the most consequential PM since Thatcher.
If you add to that two times winner of the London mayoralty, the man who won the Brexit referendum, then you have the most successful electoral track record since Blair (at least).
I'm not the greatest Boris fan but its not a bad record so far - even if he does blow up at some point in the future. (But don't count on it.)
I think Boris will be looked back on as more influential than even Blair eventually, and won't have Iraq to his name.
Don't forget to add to his electoral track record winning the referendum itself!
Seems good to me. There is nothing there anyone sane in the UK could conceivably object to, unless they don't want US medical advances to be available here. I suppose Corbynistas would rather that people weren't treated well if the alternative is that - God forbid - a US company supplied the treatment.
There is plenty to be concerned about there if you know anything about how pharmaceuticals or medical devices are regulated, evaluated and reimbursed in the UK.
I would just add that Chorley is target seat number 86 for the Conservatives.....
I wouldn't fancy their chances against Hoyle, he is well respected. You'd have to be getting into serious landslide territory to see Chorley go.
Also, what if he wins the Speaker election, would they stand against him? Last I heard it was still on for Monday.
I remember a decent procfit on that one last time
Chorley was considered a 2017 target and in the end, it wasn't even close.
if there are odds worth taking I'd so so again
It looks like Barnesian has run the numbers, then looked at the Labour casualites and said "well, that's never going to fall, nor that, nor that, nor that....."
Perhaps he'd like to share with us his "special pleadings" - and tell us where the votes go instead.
No special pleadings or "I don't like the look of that". That would be completely pointless.
I'm not doing this for spin or to wind anyone up.
I'm trying to develop an objective credible model by constituency that might shed some light on individual constituencies and on the overall result that might be of interest to PBers. That's all.
You can simply ignore it as spin if you want. Or I could just keep it to myself for when the constituency betting markets open up.
Based on 2017 that gives JC lots of space to win by some margin.
What are you talking about? On those numbers, the Cons are holding 83.8% of their 2017 vote - Labour just 60.7%.
So for example in Chorley, which Labour won with Lindsay Hoyle by c7,500, on the same turnout as 2017 the Conservatives win with a majority of 800. Now, there was a big UKIP vote, and that might go back to Labour if a) that is where it came from and b) there is no Brexit Party candidate. But Barnesian can tell me if he's got the Cons winning Chorley in his model.....
Labour retains by 1,400 but it's close
Additive model (UNS) gives Con 19,339. Lab 22,405 Multiplicative gives Con 19,494, Lab 19,366
I use 80/20 additive/multiplicative and this gives Con 19,370, Lab 21,794
EDIT: This includes a 400 tactical vote from LDs for Labour. It also assumes BXP stand (no UKIP last time) and get 5,400 so that is a critical assumption.
Which is a bit of a problem. You have two models, combine them in a way that may well be valid but has no apparent justification, then make post-modelling adjustments. It's not unheard of, and many modellers behave like this, but it becomes progressively more difficult to defend.
Excuse me while I rant for a moment.
Modellers do this and it drives me nuts. Every time one gets good, they start publishing more than one. Nate Silver (or 538, I forget which) had six simultaneously. Ashcroft had three MRP models. Hix had two or three. It doesn't help and arguably makes things worse. If both of your models are wrong, why does combining them make it better? If one is correct, why do the other model? Aaargh!
End rant ...
Neither are correct. Both have flaws at the boundaries (negative shares for UNS, >100% share for multiplicative). My combination isn't correct either but hopefully is a better predictor of uniform swings and "lumpiness".
You have a point though. I shouldn't publish the result of the two sub-models - only the combination as my best guess.
Luckily the model isn't for flying an aircraft so I don't need to "prove" it or even defend or justify it, though sometimes I try to do the latter.
Seems good to me. There is nothing there anyone sane in the UK could conceivably object to, unless they don't want US medical advances to be available here. I suppose Corbynistas would rather that people weren't treated well if the alternative is that - God forbid - a US company supplied the treatment.
"Procedural fairness" means that NICE can't use NHS clout to negotiate drug prices.
There is a reason that US health costs are at least twice as high as ours per capita.
There's lots of reasons, but that's not one of them. Of course the NHS can get a good price, it's a big purchaser. Same as any big purchaser. There's nothing 'procedurally unfair' in that, any more than there is in Apple getting good prices from its Chinese suppliers.
The idea that we will suddenly pay more for drugs is for the birds. Anyone who has read a cancer story about someone going to America for treatment would be aware that NICE evaluates on a cost benefit model which would lead to drugs being not recommended for use if the price was higher. Fair access probably means transparency on how these decisions are made.
1. "You're jealous". 2. "Stop being so old-fashioned. Next thing you'll be causing a winter of discontent and bringing back flared trousers". 3. "Stop being so class-prejudiced." 4. "Abolishing Eton etc. will cost the country billions."
If the opponent doesn't like the issue...
(*) (Insert joke about Winchester-Eton match.)
It will appeal to the cult, it will not appeal to higher earning Remainers who will go LD rather than Corbyn Labour (and it obviously will not appeal to 2017 Tory voters either)
Which is a bit of a problem. You have two models, combine them in a way that may well be valid but has no apparent justification, then make post-modelling adjustments. It's not unheard of, and many modellers behave like this, but it becomes progressively more difficult to defend.
Excuse me while I rant for a moment.
Modellers do this and it drives me nuts. Every time one gets good, they start publishing more than one. Nate Silver (or 538, I forget which) had six simultaneously. Ashcroft had three MRP models. Hix had two or three. It doesn't help and arguably makes things worse. If both of your models are wrong, why does combining them make it better? If one is correct, why do the other model? Aaargh!
End rant ...
In my experience the best inoculation against the disease of believing too much in models is having written models oneself.
That's so true! I remember going through the modelling of the economic benefits of the third runway at Heathrow and thinking: Jeez - this is a dressed up PR exercise. Sure enough it was "the computer says" by the media and MPs. Didn't look at the assumptions or entity definitions.
I built my first model in 1965, in Fortran, to help determine how much ICI should spend on advertising Dulux Paint using Markov chains on brand shares.
That is interesting. I am pretty sure that is a reversal of 20+ years ago (anyone with data able to advise?), when the less educated voted Labour and more educated Liberal or Tory.
The LDs still do best with the most educated percentage of vote wise, the last time the Tories won graduates was 2015 but even then they did better percentage wise with non graduates.
Remember in the 1980s only 10% went to university ie becoming mainly higher earning, professional Tory or LD voters, now 30 to 40% go plenty of middle income or even low income Labour graduates too
Possibly, and also as Tories tend to be in the older age group the percentage of grads decreases with age. It could also be that educated people realise Brexit is a moronic policy, and the party now most associated is the party that used to describe itself as conservative.
Graduates tend to be Remainers which may also be a factor, though the Tories still do better with graduates than the Brexit Party do
My tactical voting situation is easy luckily, it's a Lab remainer held seat where the tories came close in 2015 but way out again in 2017. LDs going nowhere, and its not quite a safe enough seat to risk voting for them instead of Lab.
Don't envy those who have a less clear choice but it seems to me that the the safest choice is to go by 2017 results because that is what the average person will do, the one who doesn't look and compare 75 different voting guide websites, and is the only concrete data we have. Euro election results don't count in a GE, its a different system with different stakes so misleading to accord it the same importance as a GE. Also, the argument. That 2017 can't be used as a base because Labour's support has slipped is a circular argument - the people who don't want to vote Labour now are largely the remainer tactical voter types moving to LDs. The bulk of whom will look at their constituency on polling day and cross the Labour box because in this country we vote against the one we want to lose, not for the one we want to win. It sucks for the LDs of course but from a tactical voting system it will almost always be best to just vote for second place party in 2017.
Corbyn framing this election as "whose side are you on?" (yacht owners or disabled children etc). He seems to be re-running what seemed to work in 2017.
Colour me sceptical but do we really expect that the 500bn spending spree is going to be paid for by the 50000 yacht owners
Even if that is individuals who own the yachts it is 1 million per head in tax due.
There's yachts and yachts, and the median value of a uk owned yacht is probably about £25000. I think he means superyachts.
Well quite. I would honestly like to live in the world Corbyn has planned in his head, but if you want to mega tax the mega rich they just move abroad. That’s what happened in France not so long ago.
That is easier in France where many just move to Belgium and get a similar culture with a quick train to Paris. Leaving London to another English speaking place isn't the same. You become too far from family or have to live in a backwater like Dublin.
You just go to New York or Singapore and take your family with you
Haha even a 11% lead isnt enough for a tory majority in your model? Really? What sort of lead would it take?
Con 36.6%, Lab 23.4% would give the Tories 327 seats. So a 13% lead would do it.
Remember there are a lot of SNP and LD seats. It's not just the lead over Labour.
Even with a 13% lead in my example, the Tories are still 8% points down on 2017 where they got 43.4%. Lead over Labour isn't everything.
Show you working please.....otherwise zero marks will be awarded.
That's OK
I'm happy to disclose my assumptions and and what they produce, in case they are of interest to anyone.
I'm not happy to publish my spreadsheet (don't know how) and anyway, your eyes would glaze over.
So I'll take zero marks.
Please can you disclose your assumptions? It genuinely doesn't make any sense to me when compared with 2015, which the polls are increasingly similar to, but better for the Tories.
I have disclosed my assumptions at length throughout this thread. You'll need to go back and read them if you interested. If not, just ignore.
Corbyn framing this election as "whose side are you on?" (yacht owners or disabled children etc). He seems to be re-running what seemed to work in 2017.
Colour me sceptical but do we really expect that the 500bn spending spree is going to be paid for by the 50000 yacht owners
Even if that is individuals who own the yachts it is 1 million per head in tax due.
There's yachts and yachts, and the median value of a uk owned yacht is probably about £25000. I think he means superyachts.
Well quite. I would honestly like to live in the world Corbyn has planned in his head, but if you want to mega tax the mega rich they just move abroad. That’s what happened in France not so long ago.
That is easier in France where many just move to Belgium and get a similar culture with a quick train to Paris. Leaving London to another English speaking place isn't the same. You become too far from family or have to live in a backwater like Dublin.
If you earn the rarified heights of £10M, the difference between 45% tax and a French style 75%, as I believe was introduced, is around £3m or just over 8 grand a day. So you can stay in Corbyn's workers' paradise and be charged 8 grand a day before you buy a tin of beans or leave. Dublin sounds great. They also sell in various places around the world lots of sunny passports for the price of a nice villa or a few hundred grand in govt bonds, and if you invest something like £5M.countries will fall over themselves to open their doors.
Notwithstanding personal ties (and these days there's the internet) there comes a point where the calculation for these kind of people really isn't that hard. It's a question of where each individual tipping point is, and it doesn't take many to start making a dent in the income tax take.
With the stated objective of keeping the Tories out.
Tactical voting advice is inherently about keeping a particular party out, no? For instance I'm I'm certain that a Unionist tactical voting wheel of SNPbad is being fashioned in the orc pits of Barad-dûr as I type.
Sure - but the context is Gina Miller the doughty champion of Parliament’s rights.
That was all bollocks. She was just a wealthy, articulate anti-Tory Remainer who saw an opportunity to overturn the people’s wishes
It is possible to be a champion of parliamentary rights, and be either wealthy or poor, pro or anti tory and be in the majority or minority views on various issues of the day. Articulate certainly helps though.
I believe she used “parliamentary rights” as a convenient tool - it wasn’t a point of principle for her. She wanted her own way.
Corbyn framing this election as "whose side are you on?" (yacht owners or disabled children etc). He seems to be re-running what seemed to work in 2017.
Colour me sceptical but do we really expect that the 500bn spending spree is going to be paid for by the 50000 yacht owners
Even if that is individuals who own the yachts it is 1 million per head in tax due.
There's yachts and yachts, and the median value of a uk owned yacht is probably about £25000. I think he means superyachts.
Well quite. I would honestly like to live in the world Corbyn has planned in his head, but if you want to mega tax the mega rich they just move abroad. That’s what happened in France not so long ago.
That is easier in France where many just move to Belgium and get a similar culture with a quick train to Paris. Leaving London to another English speaking place isn't the same. You become too far from family or have to live in a backwater like Dublin.
Im pretty sure that Billionaires could find a way to make themselves happier.
Which would mean lower drug prices? Access doesn't mean much
It would mean higher drug prices.
In the US, federal agencies are not allowed by law to negotiate drug prices. In the UK, NICE uses the NHS clout to negotiate down drug prices. The US want that removed and a move to the US system.
NICE does nothing of the sort
They have an explicit value based pricing model
They have no problem approving high prices provided that the science justifies it
It is then up to the various NHS bodies to determine whether they will reimburse that drug
This all points to the Tories becoming the party of the uneducated and the Lib Dems becoming the party of the educated. Bad news for Labour.
No, the educational qualifications profile results are more indicative of age. Few oldies have degrees. As the youthful graduates get older they will become more Tory.
I was about to write the same thing. It's a bit unfair to think of these people as thick; it's just that hardly anyone was encouraged to go on to higher education before the 1960s.
Which would mean lower drug prices? Access doesn't mean much
It would mean higher drug prices.
In the US, federal agencies are not allowed by law to negotiate drug prices. In the UK, NICE uses the NHS clout to negotiate down drug prices. The US want that removed and a move to the US system.
NICE does nothing of the sort
They have an explicit value based pricing model
They have no problem approving high prices provided that the science justifies it
It is then up to the various NHS bodies to determine whether they will reimburse that drug
Fortunately here in East Ham I don't have to worry about tactical voting considerations as Stephen Timms battles to hold on to his wafer thin 39,883 majority.
@HYUFD hasn't included East Ham in his list of key East and North East London marginals so while I suspect we'll see Labour activists aplenty in the High Street over the next few weekends, I suspect effort from the other parties will be more limited.
As an aside, interesting to hear the Conservatives talking about "conversations about democracy". It's likely to show largesse when you have all the power and permit everyone else to think they are having a say when they aren't. There is something similar here in Newham with Citizens' Assemblies all organised and steered by the Council - Labour 60, all other parties NIL.
Which would mean lower drug prices? Access doesn't mean much
It would mean higher drug prices.
In the US, federal agencies are not allowed by law to negotiate drug prices. In the UK, NICE uses the NHS clout to negotiate down drug prices. The US want that removed and a move to the US system.
NICE does nothing of the sort
They have an explicit value based pricing model
They have no problem approving high prices provided that the science justifies it
It is then up to the various NHS bodies to determine whether they will reimburse that drug
If NICE approve during its Technology Appraisal process it is not up to NHS bodies, it is mandatory to reimburse.
Revealing the price in a 'transparent' process is highly likely to increase prices for the UK as the discount the NHS achieves currently (based on value), through secrecy, will be hammered by every other country demanding the same.
So the NHS will have the choice to pay more or go without. Most new drugs blows the cost effectiveness threshold NICE use without a discount.
I am wondering how much reliance should be placed on opinion polls this time without some very material adjustments for local circumstance. In the past local effects have been subdued or cancelled out. This time they look more material. Obvious examples are the SW oddities of Wollaston and Wright, which would be Conservative according to any poll - but Wollaston will hold Totnes and Wright is an interesting situation. Then there are the London seats with Jewish votes. Add in the really local effects, like the uproar over local plans in places like Guildford and Mole Valley, and the scandals of O'Mara and Vaz.......and there seem a number of seats which are not well predicted by polling. I am looking forward to some good odds.
Comments
Anyway, phew.
How well will the stock Tory responses fare?
1. "You're jealous".
2. "Stop being so old-fashioned. Next thing you'll be causing a winter of discontent and bringing back flared trousers".
3. "Stop being so class-prejudiced."
4. "Abolishing Eton etc. will cost the country billions."
If the opponent doesn't like the issue...
(*) (Insert joke about Winchester-Eton match.)
Don't forget to add to his electoral track record winning the referendum itself!
I'm not doing this for spin or to wind anyone up.
I'm trying to develop an objective credible model by constituency that might shed some light on individual constituencies and on the overall result that might be of interest to PBers. That's all.
You can simply ignore it as spin if you want. Or I could just keep it to myself for when the constituency betting markets open up.
Neither are correct. Both have flaws at the boundaries (negative shares for UNS, >100% share for multiplicative). My combination isn't correct either but hopefully is a better predictor of uniform swings and "lumpiness".
You have a point though. I shouldn't publish the result of the two sub-models - only the combination as my best guess.
Luckily the model isn't for flying an aircraft so I don't need to "prove" it or even defend or justify it, though sometimes I try to do the latter.
https://twitter.com/robwattsf1/status/1189904522946207744
I built my first model in 1965, in Fortran, to help determine how much ICI should spend on advertising Dulux Paint using Markov chains on brand shares.
Don't envy those who have a less clear choice but it seems to me that the the safest choice is to go by 2017 results because that is what the average person will do, the one who doesn't look and compare 75 different voting guide websites, and is the only concrete data we have. Euro election results don't count in a GE, its a different system with different stakes so misleading to accord it the same importance as a GE. Also, the argument. That 2017 can't be used as a base because Labour's support has slipped is a circular argument - the people who don't want to vote Labour now are largely the remainer tactical voter types moving to LDs. The bulk of whom will look at their constituency on polling day and cross the Labour box because in this country we vote against the one we want to lose, not for the one we want to win. It sucks for the LDs of course but from a tactical voting system it will almost always be best to just vote for second place party in 2017.
Notwithstanding personal ties (and these days there's the internet) there comes a point where the calculation for these kind of people really isn't that hard. It's a question of where each individual tipping point is, and it doesn't take many to start making a dent in the income tax take.
They have an explicit value based pricing model
They have no problem approving high prices provided that the science justifies it
It is then up to the various NHS bodies to determine whether they will reimburse that drug
https://www.england.nhs.uk/2018/11/nhs-set-to-save-record-300-million-on-the-nhss-highest-drug-spend/
Fortunately here in East Ham I don't have to worry about tactical voting considerations as Stephen Timms battles to hold on to his wafer thin 39,883 majority.
@HYUFD hasn't included East Ham in his list of key East and North East London marginals so while I suspect we'll see Labour activists aplenty in the High Street over the next few weekends, I suspect effort from the other parties will be more limited.
As an aside, interesting to hear the Conservatives talking about "conversations about democracy". It's likely to show largesse when you have all the power and permit everyone else to think they are having a say when they aren't. There is something similar here in Newham with Citizens' Assemblies all organised and steered by the Council - Labour 60, all other parties NIL.
Revealing the price in a 'transparent' process is highly likely to increase prices for the UK as the discount the NHS achieves currently (based on value), through secrecy, will be hammered by every other country demanding the same.
So the NHS will have the choice to pay more or go without. Most new drugs blows the cost effectiveness threshold NICE use without a discount.