Another tin-foil hat lunatic airs his populist conspiracy theories. Ignore.
Surely Nigel doesn't actually believe that the 'establishment' is controlling YouGov's polling methodology? I think he just wants to instil his followers with a lot of anger, a lot of hate and a lot of victimhood. Expect much more of this.
> @Stark_Dawning said: > > @HYUFD said: > > > https://twitter.com/Nigel_Farage/status/1134365557325799424 > > > > > > Another tin-foil hat lunatic airs his populist conspiracy theories. Ignore. > > Surely Nigel doesn't actually believe that the 'establishment' is controlling YouGov's polling methodology? I think he just wants to instil his followers with a lot of anger, a lot of hate and a lot of victimhood. Expect much more of this.
Or he's just been spending too much time around the Donald.
> > > Another tin-foil hat lunatic airs his populist conspiracy theories. Ignore.
> >
> > Actually, it's a fair point, isn't it?
> >
> > Much as I hate and loathe the Brexit Party, and would like to see its leader soaked in milk shakes for the rest of his natural life, It does seem a bit peculiar to categorise them under "Others" when they have just topped the poll in a national election.
>
> Hardly, given that Brexit polled higher in the EU polls than they actually got in the EU election itself.
Why does that make it fair to class them among Others?
The aim of an opinion poll is to get an accurate poll reading, not to massage the ego of the party's leader. YouGov are entitled to use whatever methodology they believe works best. There's no evidence as yet that their poll rating is being underrecorded.
> Final salary is an abomination in this day and age, 10+10 defined contribution superannuation along the lines of Oz is the way to go.
Don't even get me started on public sector pensions. People talk about unfairness. How can in be fair that "The Few" can retire in their mid 50s while "The Many" have to carry on working until they are in their 70s. Why won't it change? Because the powers that be (including the Union Barons) like their own arrangements too much. This includes Nigel-man-o-the-people-Farridge
Final salary pensions have not been available to Civil Servants since about 2005/6. The only ones still extant are those who worked before then. If you think jobs in the civil service are so easy and lucrative, the job site is/was jobs.civilservice.gov.uk. but you have to sit one test (two for the professions) an aptitude test, a panel interview, a security check, an enhanced background check for some jobs, and if you are successful the pay is crap and the annual payrises zero. I earn about £18k plus bonus more in the private sector than I would get in the public one. It is not as easy or as lucrative as you think.
The real divide is between defined benefit (DB) and defined contribution (DC) schemes, rather than just final salary (which are a form of DB) and everything else. The reason being that on DB schemes, investment risk is on the sponsor while with DC it gets transferred to the employee. In addition, since civil service pension arrangements aren't funded, the only risk remaining with the employee is that the government refuses to pay up when the time comes, which is assumed to be vanishingly unlikely.
As I understand it, most new civil servants/doctors/teachers etc get some form of career average pension, which has been set up to be slightly less generous than the previously available final salary, but is still vastly superior to anything available in the private sector. This is largely due to the hugely increased level of security that the eventual payments will cover the individual's lifestyle following retirement, but is also expected to be better in pure expected value terms as well (although obviously this is difficult to determine since it involves making assumptions about long term annuity rates/life expectancy improvements etc).
I'd still bite your hand off for a government sponsored pension arrangement, but agree that I probably wouldn't want to accept the shorter term drop in salary that goes with it.
> > Another tin-foil hat lunatic airs his populist conspiracy theories. Ignore.
>
> Actually, it's a fair point, isn't it?
>
> Much as I hate and loathe the Brexit Party, and would like to see its leader soaked in milk shakes for the rest of his natural life, It does seem a bit peculiar to categorise them under "Others" when they have just topped the poll in a national election.
Hardly, given that Brexit polled higher in the EU polls than they actually got in the EU election itself.
I thought you gov explained their approach very clearly up thread in NF doesn’t understand it it’s his problem
I hadn't realised YouGov had been quoted on this, but thank you for letting me know about it.
But really I don't think "we're still testing whether this makes any difference" is sufficient justification for concluding it doesn't make any difference. When a party has been registering in the 20s both in opinion polls and actual elections, it makes no sense to exclude it from the list of named parties.
> @Chris said: > > I hadn't realised YouGov had been quoted on this, but thank you for letting me know about it. > > But really I don't think "we're still testing whether this makes any difference" is sufficient justification for concluding it doesn't make any difference. When a party has been registering in the 20s both in opinion polls and actual elections, it makes no sense to exclude it from the list of named parties. ---------------
YouGov explained in their statement. For the European Election polling, they did name them, and it resulted in them significantly overestimating their support. YouGov had them at 37% in those polls.
> @kamski said: > > @anothernick said: > > > @kamski said: > > > Anyone know why sterling has fallen again against the euro the last 2 days? I would have thought with Hammond's clear anti-no-deal statement the pound might have perked up slightly? > > > > Market beginning to realise that Brexit uncertainty is not going to be resolved anytime soon. > > > > City types have been extraordinarily optimistic about Brexit - I have lost count of the number of presentations I have been to (as part of my job) at which managers have opined that sterling is set for a rise as soon as the markets have clarity. Up until recently the comforting assumption has been that clarity was just around the corner, it is only now that they are beginning to realise that it may be delayed for a decade or two. > > Interesting, tho I assume it would depend on the type of "clarity" - wouldn't No-Deal clarity lead to sterling falling further? And conversely No-Deal being ruled out would lead to a rise from current levels? > Or was the assumption up to now that clarity would definitely involve No-Deal being ruled out? > > Maybe the calculation is that a competitive Conservative leadership election will lead to contenders all promising No-Deal (if Unicorn doesn't appear), and it now looks more likely to be competitive than a few days ago?
Another tenet of city thinking has been that the chances of no deal were minimal and in that they have been proved correct. But although we have avoided no deal we are no closer to defining our long-term relationship with the EU.
No one likes to talk about John Redwood, understandably.
Actually from being the go-to Tory loon nonpareil, he's been somewhat eclipsed by the current rich array. He must feel a bit miffed.
The list of beauty parade contestants has not been closed yet. We might be treated to up to four more, one of whom is an old hand in a swimsuit. Though I'm not sure that any of us are yet ready to see Sir Graham in a mankini.
> @AlastairMeeks said: > > @eek said: > > > > @Chris said: > > > > > @ah009 said: > > > > > > @HYUFD said: > > > > > > https://twitter.com/Nigel_Farage/status/1134365557325799424 > > > > > > > > > > > > Another tin-foil hat lunatic airs his populist conspiracy theories. Ignore. > > > > > > > > Actually, it's a fair point, isn't it? > > > > > > > > Much as I hate and loathe the Brexit Party, and would like to see its leader soaked in milk shakes for the rest of his natural life, It does seem a bit peculiar to categorise them under "Others" when they have just topped the poll in a national election. > > > > > > Hardly, given that Brexit polled higher in the EU polls than they actually got in the EU election itself. > > > > Why does that make it fair to class them among Others? > > The aim of an opinion poll is to get an accurate poll reading, not to massage the ego of the party's leader. YouGov are entitled to use whatever methodology they believe works best. There's no evidence as yet that their poll rating is being underrecorded. > > Their statement today seems entirely reasonable.
No evidence - apart from all the evidence we already have that if you prompt people with certain alternatives and exclude others, the alternatives that people are prompted with will benefit.
Disappointing to see you being so naive - or disingenuous.
> @williamglenn said: > > @Chris said: > > > > I hadn't realised YouGov had been quoted on this, but thank you for letting me know about it. > > > > But really I don't think "we're still testing whether this makes any difference" is sufficient justification for concluding it doesn't make any difference. When a party has been registering in the 20s both in opinion polls and actual elections, it makes no sense to exclude it from the list of named parties. > --------------- > > YouGov explained in their statement. For the European Election polling, they did name them, and it resulted in them significantly overestimating their support. YouGov had them at 37% in those polls.
You say "it resulted in them significantly overestimating their support". But as YouGov are saying there's no evidence it makes any difference, I suspect you're making that up.
> > Final salary is an abomination in this day and age, 10+10 defined contribution superannuation along the lines of Oz is the way to go.
>
>
>
> Don't even get me started on public sector pensions. People talk about unfairness. How can in be fair that "The Few" can retire in their mid 50s while "The Many" have to carry on working until they are in their 70s. Why won't it change? Because the powers that be (including the Union Barons) like their own arrangements too much. This includes Nigel-man-o-the-people-Farridge
>
> Final salary pensions have not been available to Civil Servants since about 2005/6. The only ones still extant are those who worked before then. If you think jobs in the civil service are so easy and lucrative, the job site is/was jobs.civilservice.gov.uk. but you have to sit one test (two for the professions) an aptitude test, a panel interview, a security check, an enhanced background check for some jobs, and if you are successful the pay is crap and the annual payrises zero. I earn about £18k plus bonus more in the private sector than I would get in the public one. It is not as easy or as lucrative as you think.
There is no evidence that the equivalent public sector position (if it is possible to draw comparisons) pays less than private, and most studies that I have seen point to the opposite. What is a clear fact is that the public, while not so much "job for life" (unless you are a doctor) as it used to be, is massively safer than private and still carries much better pension arrangements. And that is fundamentally unfair on most workers
Thanks for the job link, but no I have no need of a job in the civil service, and the suggestion was a bit silly. I was not suggesting it is easy to be in the civil service, or any other public sector job.
Yes they do. Cohort stratified public/private comparisons clearly show private pays more than public. It is only when you add in the pensions that the total compensation becomes equitable.
> @Stark_Dawning said: > Surely Nigel doesn't actually believe that the 'establishment' is controlling YouGov's polling methodology? I think he just wants to instil his followers with a lot of anger, a lot of hate and a lot of victimhood. Expect much more of this.
Exactly. It's the old "poor me" control drama. Covers up for the fact that they have only one policy, keeps the spotlight on him, fuels that nasty populist conflagration the Kremlin has been nursing.
> @HYUFD said: > > @nichomar said: > > > @brokenwheel said: > > > > > Back in the 2010 GE campaign when the LDs were polling at a similar level their support was 25% of ABC1 and 23% of C2DE. It’s now 29% ABC1 and 16% C2DE. > > > > The Brexit Party meanwhile are 29% C2DE and 17% ABC1. > > > > We thought class politics was dying in 2017 with the Tories and Labour with the Tories having just a 2% higher lead with ABC1s over Labour than with C2DEs, if it is Brexit Party v LDs then class politics is back with a vengeance > > > > I don’t really understand why the group who stand to suffer most by no deal brexit are the most keen on it. > > Immigration and sovereignty and they have less to lose anyway ________________________
Some of 'the other half' considered that there was nothing to lose, so they might as well vote that way as a gigantic protest at all manner of things which they thought were less good than before. That's partly what 40 years of Thatcher and Sons does to a country.
I'm not saying that they all voted that way for that reason. But enough did to give the unexpected referendum result. Douglas Alexander ex-MP got at least part of the message via the usual drastic route, i.e. he lost his seat! Still, he makes a reasonable radio presenter, just like Rory Stewart made a good TV series https://www.bbc.co.uk/programmes/m0001d9y
No evidence - apart from all the evidence we already have that if you prompt people with certain alternatives and exclude others, the alternatives that people are prompted with will benefit.
Disappointing to see you being so naive - or disingenuous.
You're missing the point completely. The Brexit party have had one outing. In that one outing, their poll ratings were generally overrecorded rather than underrecorded. This suggests that pollsters are finding too many Brexit party supporters. Anything that increases their poll rating runs the risk of increasing that error.
Now it may well be that the correct action is to take a different action to compensate. But the idea that the Brexit party is being hard done to by pollsters (to what purpose, goodness only knows) is absurd.
> @HYUFD said: > > The Brexit Party won Epping Forest, Buckingham, Wellingborough, Tonbridge, Thurrock, Mansfield, Ryedale etc even a few Tory Remain seats like Tunbridge Wells and Maidenhead last week (Stoke already has a Tory seat, also lost to Farage). > > In fact the Tories lost more areas to the Brexit Party than the LDs last week and far more of their voters ----------
The attempt to win back votes from the Brexit Party comes at the cost of losing the votes of people like Richard Nabavi. In addition, by splitting the combined Brexit vote, you help Remain parties to win.
No one likes to talk about John Redwood, understandably.
Actually from being the go-to Tory loon nonpareil, he's been somewhat eclipsed by the current rich array. He must feel a bit miffed.
The list of beauty parade contestants has not been closed yet. We might be treated to up to four more, one of whom is an old hand in a swimsuit. Though I'm not sure that any of us are yet ready to see Sir Graham in a mankini.
I am waiting for an appropriate moment to declare my candidacy for Tory Leader and PM. I have a couple of minor imperfections on my CV such as not being an MP or even a member of the Party but I still think I have a lot more to offer than most of the field. If anyone ever had any doubt about how deluded our political class is they are proving it in spades.
> @Chris said: > > You say "it resulted in them significantly overestimating their support". But as YouGov are saying there's no evidence it makes any difference, I suspect you're making that up. ------
No, YouGov are saying that there is substantial historical evidence that prompting for minor/new parties results in overstating their support, and on the face of it that's what happened last week.
> @AlastairMeeks said: > No evidence - apart from all the evidence we already have that if you prompt people with certain alternatives and exclude others, the alternatives that people are prompted with will benefit. > > > > Disappointing to see you being so naive - or disingenuous. > > You're missing the point completely. The Brexit party have had one outing. In that one outing, their poll ratings were generally overrecorded rather than underrecorded. This suggests that pollsters are finding too many Brexit party supporters. Anything that increases their poll rating runs the risk of increasing that error. > > Now it may well be that the correct action is to take a different action to compensate. But the idea that the Brexit party is being hard done to by pollsters (to what purpose, goodness only knows) is absurd.
Finding too many, or overestimating the chance that they will actually get off their arse and vote?
> @HYUFD said: > > @noneoftheabove said: > > > @HYUFD said: > > > > @Nigel_Foremain said: > > > > > @HYUFD said: > > > > > > @brokenwheel said: > > > > > > Back in the 2010 GE campaign when the LDs were polling at a similar level their support was 25% of ABC1 and 23% of C2DE. It’s now 29% ABC1 and 16% C2DE. > > > > > > > > > > The Brexit Party meanwhile are 29% C2DE and 17% ABC1. > > > > > > > > > > We thought class politics was dying in 2017 with the Tories and Labour with the Tories having just a 2% higher lead with ABC1s over Labour than with C2DEs, if it is Brexit Party v LDs then class politics is back with a vengeance > > > > > > > > Only if you define class by intellect, then you might be correct. There are a lot of bright "working class" people I have met who think Brexit is for nutters. Now I have lit the touch paper, I had better go to lunch!! > > > > > > There are also some rich Brexit Party voters, I suspect the divide is more based on the class divide in terms of education than wealth and income though the divide is still there even with the latter (wealthier Leave voters are more likely to be soft Brexiteers and thus less likely to support the No Deal Brexit Party) > > > > Given the age profiles it would be interesting to see the net worth profiles of Brexit v LD, I would not be surprised if the mean Brexit net worth > mean LD net worth despite the class differences above (even excluding Putin). > > No way. > > LDs will be far wealthier on average than Brexit Party voters if they lead with ABC1s including lawyers, doctors and managers but the Brexit Party lead with the working class. > > Thus the LDs won Richmond Park, Islington, Kensington and Esher and Walton last week but the Brexit Party won Bolsover, Merthyr Tydfil, Wigan and Stoke. > > The middle class battle at the moment is LD v Tory, the working class battle is Brexit Party v Labour with the former winning currently in each case
Again you underestimate the impact of the last twenty years on those under 45 which is half the LD vote but less than a third of the BP vote. Most of the LD ABC1 <45s will have negative net worth due to cost of going to university and high housing costs. Only 5% of under 35s are in the top 40% of net worth and 44% are in the bottom 20%. Net worth is so skewed by age that it seems unlikely that the age effect is offset by the class impact on voting.
No one likes to talk about John Redwood, understandably.
Actually from being the go-to Tory loon nonpareil, he's been somewhat eclipsed by the current rich array. He must feel a bit miffed.
The list of beauty parade contestants has not been closed yet. We might be treated to up to four more, one of whom is an old hand in a swimsuit. Though I'm not sure that any of us are yet ready to see Sir Graham in a mankini.
I am waiting for an appropriate moment to declare my candidacy for Tory Leader and PM. I have a couple of minor imperfections on my CV such as not being an MP or even a member of the Party but I still think I have a lot more to offer than most of the field. If anyone ever had any doubt about how deluded our political class is they are proving it in spades.
One MP's observations that some of the candidates might usefully reflect on:
> @AlastairMeeks said: > No evidence - apart from all the evidence we already have that if you prompt people with certain alternatives and exclude others, the alternatives that people are prompted with will benefit. > > > > Disappointing to see you being so naive - or disingenuous. > > You're missing the point completely. The Brexit party have had one outing. In that one outing, their poll ratings were generally overrecorded rather than underrecorded. This suggests that pollsters are finding too many Brexit party supporters. Anything that increases their poll rating runs the risk of increasing that error. > > Now it may well be that the correct action is to take a different action to compensate. But the idea that the Brexit party is being hard done to by pollsters (to what purpose, goodness only knows) is absurd.
That's a completely unscientific attitude. So what if YouGov did overestimate Brexit Party support? The way to correct that is to find out why and make appropriate adjustments, not to adopt a method that results in an underestimate of the Brexit Party vote for an entirely unrelated reason, and to hope that that error will cancel out the other error that they don't understand.
No wonder opinion polling is in such a mess if that's the way the pollsters operate,.
> @Chris said: > > That's a completely unscientific attitude. So what if YouGov did overestimate Brexit Party support? The way to correct that is to find out why and make appropriate adjustments, not to adopt a method that results in an underestimate of the Brexit Party vote for an entirely unrelated reason, and to hope that that error will cancel out the other error that they don't understand. > > No wonder opinion polling is in such a mess if that's the way the pollsters operate,. --------
Something doesn't become unscientific just because you say it is. The whole reason why they have an 'other' category in the first place is because of the evidence that prompting for minor parties distorts the results.
> @Dura_Ace said: > > @Dura_Ace said: > > > https://twitter.com/thebeakbrewery/status/1134002379765010432 > > > > > > > > > > > > > > We are 12-18 months away from Sarajevo '92. > > > > That is ludicrous > > There were plenty of people in SFRY expressing similar sentiments right up until the day they were scavenging for snow dusted firewood in sniper alley.
> No evidence - apart from all the evidence we already have that if you prompt people with certain alternatives and exclude others, the alternatives that people are prompted with will benefit.
>
>
>
> Disappointing to see you being so naive - or disingenuous.
>
> You're missing the point completely. The Brexit party have had one outing. In that one outing, their poll ratings were generally overrecorded rather than underrecorded. This suggests that pollsters are finding too many Brexit party supporters. Anything that increases their poll rating runs the risk of increasing that error.
>
> Now it may well be that the correct action is to take a different action to compensate. But the idea that the Brexit party is being hard done to by pollsters (to what purpose, goodness only knows) is absurd.
That's a completely unscientific attitude. So what if YouGov did overestimate Brexit Party support? The way to correct that is to find out why and make appropriate adjustments, not to adopt a method that results in an underestimate of the Brexit Party vote for an entirely unrelated reason, and to hope that that error will cancel out the other error that they don't understand.
No wonder opinion polling is in such a mess if that's the way the pollsters operate,.
This is not a new problem. Pollsters have had this problem with the Greens and with UKIP in the past too. Part of the problem relates to the lack of party infrastructure, as well as voter recognition, the ability to get out the party vote. The question of when to move the minor party from others to the big boys' list is not straightforward and isn't necessarily a question of how much they are polling.
> @williamglenn said: > > @Chris said: > > > > You say "it resulted in them significantly overestimating their support". But as YouGov are saying there's no evidence it makes any difference, I suspect you're making that up. > ------ > > No, YouGov are saying that there is substantial historical evidence that prompting for minor/new parties results in overstating their support, and on the face of it that's what happened last week.
If they are saying that, please can you quote them?
All I see in what they were quoted as seeing earlier is that they are still trying to determine whether the Brexit Party's support was affected by prompting, and this on their previous policy about UKIP: "We regularly tested the effect of prompting on UKIP support, and, once it seemed it was no longer giving them an artifical boost, we started including UKIP in the main prompt alongside Labour, the Conservatives and Liberal Democrats."
The obvious question is how they can determine which is more accurate - the figure with prompting, or the figure without prompting. The fact that the two figures for UKIP apparently converged doesn't, of course, give us any information about which was the more accurate before they converged. It's entirely possible that the figure without prompting was too low!
> That's a completely unscientific attitude. So what if YouGov did overestimate Brexit Party support? The way to correct that is to find out why and make appropriate adjustments, not to adopt a method that results in an underestimate of the Brexit Party vote for an entirely unrelated reason, and to hope that that error will cancel out the other error that they don't understand.
>
> No wonder opinion polling is in such a mess if that's the way the pollsters operate,.
--------
Something doesn't become unscientific just because you say it is. The whole reason why they have an 'other' category in the first place is because of the evidence that prompting for minor parties distorts the results.
What do you think the Lib Dem score would be if they weren’t prompted?
And why did they prompt for BXP at the Euros? They were a minor party there for all we knew
> @AlastairMeeks said: > > @AlastairMeeks said: > > > No evidence - apart from all the evidence we already have that if you prompt people with certain alternatives and exclude others, the alternatives that people are prompted with will benefit. > > > > > > > > > > > > Disappointing to see you being so naive - or disingenuous. > > > > > > You're missing the point completely. The Brexit party have had one outing. In that one outing, their poll ratings were generally overrecorded rather than underrecorded. This suggests that pollsters are finding too many Brexit party supporters. Anything that increases their poll rating runs the risk of increasing that error. > > > > > > Now it may well be that the correct action is to take a different action to compensate. But the idea that the Brexit party is being hard done to by pollsters (to what purpose, goodness only knows) is absurd. > > > > That's a completely unscientific attitude. So what if YouGov did overestimate Brexit Party support? The way to correct that is to find out why and make appropriate adjustments, not to adopt a method that results in an underestimate of the Brexit Party vote for an entirely unrelated reason, and to hope that that error will cancel out the other error that they don't understand. > > > > No wonder opinion polling is in such a mess if that's the way the pollsters operate,. > > This is not a new problem. Pollsters have had this problem with the Greens and with UKIP in the past too. Part of the problem relates to the lack of party infrastructure, as well as voter recognition, the ability to get out the party vote. The question of when to move the minor party from others to the big boys' list is not straightforward and isn't necessarily a question of how much they are polling.
Obviously you've nothing there to say that they should not prompt rather than prompting.
If you want to defend not prompting, you need some evidence that not prompting gives more accurate results.
> @Chris said: > > The obvious question is how they can determine which is more accurate - the figure with prompting, or the figure without prompting. ------
By comparing with actual election results, like the ones last week in which the polls that prompted for the Brexit Party were proved to have overstated their support.
Polling relies on a lot of social science and understanding of psychology. It's not a case of just giving a model of a ballot paper to a sample of people and reporting the raw results.
> @Chris said: > > @AlastairMeeks said: > > > @AlastairMeeks said: > > > > > No evidence - apart from all the evidence we already have that if you prompt people with certain alternatives and exclude others, the alternatives that people are prompted with will benefit. > > > > > > > > > > > > > > > > > > > > Disappointing to see you being so naive - or disingenuous. > > > > > > > > > > You're missing the point completely. The Brexit party have had one outing. In that one outing, their poll ratings were generally overrecorded rather than underrecorded. This suggests that pollsters are finding too many Brexit party supporters. Anything that increases their poll rating runs the risk of increasing that error. > > > > > > > > > > Now it may well be that the correct action is to take a different action to compensate. But the idea that the Brexit party is being hard done to by pollsters (to what purpose, goodness only knows) is absurd. > > > > > > > > That's a completely unscientific attitude. So what if YouGov did overestimate Brexit Party support? The way to correct that is to find out why and make appropriate adjustments, not to adopt a method that results in an underestimate of the Brexit Party vote for an entirely unrelated reason, and to hope that that error will cancel out the other error that they don't understand. > > > > > > > > No wonder opinion polling is in such a mess if that's the way the pollsters operate,. > > > > This is not a new problem. Pollsters have had this problem with the Greens and with UKIP in the past too. Part of the problem relates to the lack of party infrastructure, as well as voter recognition, the ability to get out the party vote. The question of when to move the minor party from others to the big boys' list is not straightforward and isn't necessarily a question of how much they are polling. > > Obviously you've nothing there to say that they should not prompt rather than prompting. > > If you want to defend not prompting, you need some evidence that not prompting gives more accurate results.
We don't know. What we do know is that yougov over estimated Brexit last Thursday using whatever method they used then.
And the purpose of a polling company is to provide accurate results not to ask questions that make people (not paying for said polling) happy.
The chance of any pollster getting it right at a particular election is pure luck I think for the most part (Over time provided they're not completely shit all should get closish at one election or another).
ICM "won" 2015 Survation worked very well for 2017 iirc Ipsos Mori for 2019 Euros.
> @williamglenn said: > > @Chris said: > > > > That's a completely unscientific attitude. So what if YouGov did overestimate Brexit Party support? The way to correct that is to find out why and make appropriate adjustments, not to adopt a method that results in an underestimate of the Brexit Party vote for an entirely unrelated reason, and to hope that that error will cancel out the other error that they don't understand. > > > > No wonder opinion polling is in such a mess if that's the way the pollsters operate,. > -------- > > Something doesn't become unscientific just because you say it is. The whole reason why they have an 'other' category in the first place is because of the evidence that prompting for minor parties distorts the results.
Sorry, but I don't think you're thinking about what you're saying.
You talk about prompting for minor parties "distorting" the results. Obviously it's true that prompting for minor parties benefits minor parties. But that doesn't mean it's a "distortion".
In a real election, there is prompting for minor parties (if TBP can still be called that), both during the campaign, and by reason of their presence on the ballot paper.
The question is whether pollsters' prompting or not prompting produces a more accurate result. That requires evidence, not just assertion. Remember that generally the Lib Dems' poll rating has tended to improve during election campaigns, because of their exposure. It's not exactly the same issue, but it's obviously closely related.
> The obvious question is how they can determine which is more accurate - the figure with prompting, or the figure without prompting.
------
By comparing with actual election results, like the ones last week in which the polls that prompted for the Brexit Party were proved to have overstated their support.
Polling relies on a lot of social science and understanding of psychology. It's not a case of just giving a model of a ballot paper to a sample of people and reporting the raw results.
> @eek said: > > @Chris said: > > > @AlastairMeeks said: > > > > @AlastairMeeks said: > > > > > > > No evidence - apart from all the evidence we already have that if you prompt people with certain alternatives and exclude others, the alternatives that people are prompted with will benefit. > > > > > > > > > > > > > > > > > > > > > > > > > > > > Disappointing to see you being so naive - or disingenuous. > > > > > > > > > > > > > > You're missing the point completely. The Brexit party have had one outing. In that one outing, their poll ratings were generally overrecorded rather than underrecorded. This suggests that pollsters are finding too many Brexit party supporters. Anything that increases their poll rating runs the risk of increasing that error. > > > > > > > > > > > > > > Now it may well be that the correct action is to take a different action to compensate. But the idea that the Brexit party is being hard done to by pollsters (to what purpose, goodness only knows) is absurd. > > > > > > > > > > > > That's a completely unscientific attitude. So what if YouGov did overestimate Brexit Party support? The way to correct that is to find out why and make appropriate adjustments, not to adopt a method that results in an underestimate of the Brexit Party vote for an entirely unrelated reason, and to hope that that error will cancel out the other error that they don't understand. > > > > > > > > > > > > No wonder opinion polling is in such a mess if that's the way the pollsters operate,. > > > > > > This is not a new problem. Pollsters have had this problem with the Greens and with UKIP in the past too. Part of the problem relates to the lack of party infrastructure, as well as voter recognition, the ability to get out the party vote. The question of when to move the minor party from others to the big boys' list is not straightforward and isn't necessarily a question of how much they are polling. > > > > Obviously you've nothing there to say that they should not prompt rather than prompting. > > > > If you want to defend not prompting, you need some evidence that not prompting gives more accurate results. > > We don't know. What we do know is that yougov over estimated Brexit last Thursday using whatever method they used then. > > And the purpose of a polling company is to provide accurate results not to ask questions that make people (not paying for said polling) happy.
But you have no evidence that they overestimated TBP because they were prompting. If the YouGov statement makes any sense at all, it implies that they don't prompt until they are satisfied it makes no difference!
So what sense does it make to try to correct the unknown error by making a different change in their polling method which they'd previously decided made no difference? That's what I mean by unscientific.
I thought you gov explained their approach very clearly up thread in NF doesn’t understand it it’s his problem
Nigel Farage understands it very well but he is following the Trump playbook (or is it the Farage/Trump/Farage playbook) so, as he says, #FakePolling
Maybe, but if you spend all day spouting lies you end up convincing yourself they are true. He may genuinely choose to believe its an establishment stitch up now. Trump for example may have seemed initially to be just cynically playing to the lockherup crowd, but now seems completely unhinged and paranoid.
> @Chris said: > > You talk about prompting for minor parties "distorting" the results. Obviously it's true that prompting for minor parties benefits minor parties. But that doesn't mean it's a "distortion". > > In a real election, there is prompting for minor parties (if TBP can still be called that), both during the campaign, and by reason of their presence on the ballot paper. > > The question is whether pollsters' prompting or not prompting produces a more accurate result. That requires evidence, not just assertion. Remember that generally the Lib Dems' poll rating has tended to improve during election campaigns, because of their exposure. It's not exactly the same issue, but it's obviously closely related. ----------------
Answering a poll has no consequences but voting in an election does. For very understandable reasons, if you prompt for minor parties in a poll, you will get higher results than you would in a real election, so it is a 'distortion'. This is borne out by substantial evidence and is why YouGov use the solution of putting minor parties behind "other", to find the people who really do want to vote for them.
> The obvious question is how they can determine which is more accurate - the figure with prompting, or the figure without prompting.
------
By comparing with actual election results, like the ones last week in which the polls that prompted for the Brexit Party were proved to have overstated their support.
Polling relies on a lot of social science and understanding of psychology. It's not a case of just giving a model of a ballot paper to a sample of people and reporting the raw results.
Did the pollster who had BXP on 27 prompt?
We’re missing the original point which was NF crying deliberate foul play on the part of the pollsters. Now most reasonable people would say that was b****cks only fellow conspiracy theory believers would agree with him
> @Endillion said: > > @Pulpstar said: > > > Final salary is an abomination in this day and age, 10+10 defined contribution superannuation along the lines of Oz is the way to go. > > > > Don't even get me started on public sector pensions. People talk about unfairness. How can in be fair that "The Few" can retire in their mid 50s while "The Many" have to carry on working until they are in their 70s. Why won't it change? Because the powers that be (including the Union Barons) like their own arrangements too much. This includes Nigel-man-o-the-people-Farridge > > Final salary pensions have not been available to Civil Servants since about 2005/6. The only ones still extant are those who worked before then. If you think jobs in the civil service are so easy and lucrative, the job site is/was jobs.civilservice.gov.uk. but you have to sit one test (two for the professions) an aptitude test, a panel interview, a security check, an enhanced background check for some jobs, and if you are successful the pay is crap and the annual payrises zero. I earn about £18k plus bonus more in the private sector than I would get in the public one. It is not as easy or as lucrative as you think. > > The real divide is between defined benefit (DB) and defined contribution (DC) schemes, rather than just final salary (which are a form of DB) and everything else. The reason being that on DB schemes, investment risk is on the sponsor while with DC it gets transferred to the employee. In addition, since civil service pension arrangements aren't funded, the only risk remaining with the employee is that the government refuses to pay up when the time comes, which is assumed to be vanishingly unlikely. > > As I understand it, most new civil servants/doctors/teachers etc get some form of career average pension, which has been set up to be slightly less generous than the previously available final salary, but is still vastly superior to anything available in the private sector. This is largely due to the hugely increased level of security that the eventual payments will cover the individual's lifestyle following retirement, but is also expected to be better in pure expected value terms as well (although obviously this is difficult to determine since it involves making assumptions about long term annuity rates/life expectancy improvements etc). > > I'd still bite your hand off for a government sponsored pension arrangement, but agree that I probably wouldn't want to accept the shorter term drop in salary that goes with it.
Credit to Royal Mail for starting to develop a hybrid DB/DC, with union support and government permission.
> @williamglenn said: > > @Chris said: > > > > The obvious question is how they can determine which is more accurate - the figure with prompting, or the figure without prompting. > ------ > > By comparing with actual election results, like the ones last week in which the polls that prompted for the Brexit Party were proved to have overstated their support.
The question is whether there is any evidence that YouGov's overestimate was owing to prompting rather than any number of other possible shortcomings in their methods.
You claimed before that "there is substantial historical evidence that prompting for minor/new parties results in overstating their support" and I asked you to quote the evidence for that. Please can you do so?
Is Peter Wilsman going to provide evidence for his claims about the Israeli embassy? He should at least be given a chance. One shouldn't forget the little incident with Robert Halfon's special adviser and the Israeli diplomat. If he can't provide any evidence then I have no sympathy.
> @nichomar said: > > @Chris said: > > > > > > The obvious question is how they can determine which is more accurate - the figure with prompting, or the figure without prompting. > > ------ > > > > By comparing with actual election results, like the ones last week in which the polls that prompted for the Brexit Party were proved to have overstated their support. > > > > Polling relies on a lot of social science and understanding of psychology. It's not a case of just giving a model of a ballot paper to a sample of people and reporting the raw results. > > Did the pollster who had BXP on 27 prompt? > > We’re missing the original point which was NF crying deliberate foul play on the part of the pollsters. Now most reasonable people would say that was b****cks only fellow conspiracy theory believers would agree with him
No - you need to make an argument why he was wrong. Not just invoking "most reasonable people". They aren't to be found here. They have better things to do.
> @Chris said: > > The question is whether there is any evidence that YouGov's overestimate was owing to prompting rather than any number of other possible shortcomings in their methods. > > You claimed before that "there is substantial historical evidence that prompting for minor/new parties results in overstating their support" and I asked you to quote the evidence for that. Please can you do so?
> @Chris said: > > @williamglenn said: > > > @Chris said: > > > > > > The obvious question is how they can determine which is more accurate - the figure with prompting, or the figure without prompting. > > ------ > > > > By comparing with actual election results, like the ones last week in which the polls that prompted for the Brexit Party were proved to have overstated their support. > > The question is whether there is any evidence that YouGov's overestimate was owing to prompting rather than any number of other possible shortcomings in their methods. > > You claimed before that "there is substantial historical evidence that prompting for minor/new parties results in overstating their support" and I asked you to quote the evidence for that. Please can you do so?
To try to get the discussion back on to a semi-sensible level, the obvious reason for potential error in the European elections was the uncertainty about which respondents would actually vote in an election with a small turnout. I think that uncertainty was clear to everyone throughout.
And indeed the reason people here expected TBP to do worse than had been indicated by most of the polls was the difference in turnout changes between Brexit-supporting and non-Brexit-supporting areas, which was very fully discussed here between Thursday and Sunday.
That suggests to me that what YouGov got wrong was their modelling of turnout, and nothing whatsoever to do with whether they prompted or not for the party that was fairly clearly going to top the poll in any case.
Is Peter Wilsman going to provide evidence for his claims about the Israeli embassy? He should at least be given a chance. One shouldn't forget the little incident with Robert Halfon's special adviser and the Israeli diplomat. If he can't provide any evidence then I have no sympathy.
Point of order: embassy staff member. Not an Israeli diplomat.
To your main point, he is presumably free to provide any evidence he has at any point via whatever channel he wishes. I'm not holding my breath.
> @HYUFD said: > > @williamglenn said: > > > @HYUFD said: > > > > > > Thus the LDs won Richmond Park, Islington, Kensington and Esher and Walton last week.but the Brexit Party won Bolsover, Merthyr Tydfil, Wigan and Stoke > > ------- > > > > And the Tories want to chase after Bolsover, Merthyr Tydfil, Wigan and Stoke by becoming a copy of the Brexit Party... It's a zero seat strategy. > > Rubbish. > > The Brexit Party won Epping Forest, Buckingham, Wellingborough, Tonbridge, Thurrock, Mansfield, Ryedale etc even a few Tory Remain seats like Tunbridge Wells and Maidenhead last week (Stoke already has a Tory seat, also lost to Farage). > > In fact the Tories lost more areas to the Brexit Party than the LDs last week and far more of their voters
“Won” in an FPTnP assessment isn’t necessarily the best measure, though, is it? If you look your at Tunbridge Wells, the LibDems, Greens and CUK got 50% of the vote between them, with BXP back on 33% (35% with UKIP). And there’s another 5% for Labour and some remain splinter parties. In a GE the Tories are far more likely to lose that seat to a Remain candidate, probably a LibDem, than they are to BXP.
BXP is flattered by not having any competition, and the remain parties the opposite. In a GE the Tories would surely be more competitive than just 10% in Tunbridge Wells, and we can only hope local voters realise who the leading remain party is in their seat, or better that they get their act together and start cooperating.
> In a real election, there is prompting for minor parties (if TBP can still be called that), both during the campaign, and by reason of their presence on the ballot paper. >
One issue with minor parties in general is that they're not on the ballot paper in every constituency. I think UKIP in 2015 were the first party out of the big three to do so.
I think that's the criterion I would use when deciding whether to prompt for a party. So I wouldn't prompt for the Green Party, and then I'd have to make a judgement on whether Farage's outfit would be likely to do so. On balance right now I would say yes - but that is a guess which might be wrong.
> @williamglenn said: > > @Chris said: > > > > The question is whether there is any evidence that YouGov's overestimate was owing to prompting rather than any number of other possible shortcomings in their methods. > > > > You claimed before that "there is substantial historical evidence that prompting for minor/new parties results in overstating their support" and I asked you to quote the evidence for that. Please can you do so? > > Here you go: > > https://hbr.org/2016/11/how-polls-overestimate-support-for-third-party-candidates
That's an article about opinion polls in the USA and support for non-Democratic/Republican candidates.
You actually claimed that YouGov had said this - and I presumed at the time you meant they had said it about British elections.
> @FrankBooth said: > When do the Tory declarations need to be in and when will the first vote be? > > Hopefully not all 12 will still be in the hunt once polling commences.
The contest hasn't even begun yet. TM isn't resigning formally until 7th June. So presumably the deadline for nominations will be sometime the following week.
Tory moderates are beginning to coalesce around candidates they believe could defeat Dominic Raab, rather than Boris Johnson, as several MPs told the Guardian they considered the former Brexit secretary the bigger threat.
The moves against Raab from moderates are likely to also benefit Johnson and the environment secretary, Michael Gove, who Tory backbenchers believe is gaining momentum over his cabinet rival Jeremy Hunt.
> @OblitusSumMe said: > > @Chris said: > > > In a real election, there is prompting for minor parties (if TBP can still be called that), both during the campaign, and by reason of their presence on the ballot paper. > > > > One issue with minor parties in general is that they're not on the ballot paper in every constituency. I think UKIP in 2015 were the first party out of the big three to do so. > > I think that's the criterion I would use when deciding whether to prompt for a party. So I wouldn't prompt for the Green Party, and then I'd have to make a judgement on whether Farage's outfit would be likely to do so. On balance right now I would say yes - but that is a guess which might be wrong.
I'd suggest if you wanted to try to approach that scientifically, you would prompt for every party and apply a correction factor according to how many candidates there were.
> > > The obvious question is how they can determine which is more accurate - the figure with prompting, or the figure without prompting.
> > ------
> >
> > By comparing with actual election results, like the ones last week in which the polls that prompted for the Brexit Party were proved to have overstated their support.
>
> The question is whether there is any evidence that YouGov's overestimate was owing to prompting rather than any number of other possible shortcomings in their methods.
>
> You claimed before that "there is substantial historical evidence that prompting for minor/new parties results in overstating their support" and I asked you to quote the evidence for that. Please can you do so?
To try to get the discussion back on to a semi-sensible level, the obvious reason for potential error in the European elections was the uncertainty about which respondents would actually vote in an election with a small turnout. I think that uncertainty was clear to everyone throughout.
And indeed the reason people here expected TBP to do worse than had been indicated by most of the polls was the difference in turnout changes between Brexit-supporting and non-Brexit-supporting areas, which was very fully discussed here between Thursday and Sunday.
That suggests to me that what YouGov got wrong was their modelling of turnout, and nothing whatsoever to do with whether they prompted or not for the party that was fairly clearly going to top the poll in any case.
But the two are tied up together, as I have already pointed out. A new party does not have the infrastructure to get out the vote. That means that their voters are going to have to be that bit more motivated.
This may also explain why both Labour and the Conservatives came at the bottom end of expectations this time: their get out the vote machinery was not working for them this time.
Polls are not trying to measure abstract party allegiance. They're trying to measure who will actually cast votes.
> > The question is whether there is any evidence that YouGov's overestimate was owing to prompting rather than any number of other possible shortcomings in their methods.
> >
> > You claimed before that "there is substantial historical evidence that prompting for minor/new parties results in overstating their support" and I asked you to quote the evidence for that. Please can you do so?
That's an article about opinion polls in the USA and support for non-Democratic/Republican candidates.
You actually claimed that YouGov had said this - and I presumed at the time you meant they had said it about British elections.
Has YouGov actually said this or not?
What exactly is your point here? That if YouGov haven't commissioned their own research on this point, they should be "forced" to continue doing something they think is suboptimal and publish results they don't agree with, purely to keep you and Nigel Farage happy on a point of methodological principle?
> @Chris said: > > That's an article about opinion polls in the USA and support for non-Democratic/Republican candidates. > > You actually claimed that YouGov had said this - and I presumed at the time you meant they had said it about British elections. > > Has YouGov actually said this or not? ---------
Yes, it's in their statement today. I gave you a US source to show that it's a generally accepted phenomenon, not a conspiracy against the Brexit Party.
"In the past, prompting for smaller parties has tended to overstate their support when compared to actual elections, and the two-stage approach to prompting has produced more accurate results."
On another forum someone was wondering why so many Tory MPs are standing for the leadership. I replied that I could understand them putting their names forward for the first round, in order to find out how much support they really have, as opposed to how many people "promise" to support them but don't really mean it. This is the only way to find out how many people genuinely support their candidacy and could be useful information for future contests. For instance, Sajid Javid might have about 30 people who've promised to support him over the last couple of years but if he only gets 10 votes in the first round he knows that most of them have been lying to his face all this time. A secret ballot is the only way to accurately assess the situation.
I am sure that YouGov, with a mere 19 years of experience and over 800 employees, who do nothing other than conduct polls and analyse them, would be foolish indeed to think they might know a tad more than @Chris about how to get the best results.
> > > The obvious question is how they can determine which is more accurate - the figure with prompting, or the figure without prompting.
> > ------
> >
> > By comparing with actual election results, like the ones last week in which the polls that prompted for the Brexit Party were proved to have overstated their support.
>
> The question is whether there is any evidence that YouGov's overestimate was owing to prompting rather than any number of other possible shortcomings in their methods.
>
> You claimed before that "there is substantial historical evidence that prompting for minor/new parties results in overstating their support" and I asked you to quote the evidence for that. Please can you do so?
To try to get the discussion back on to a semi-sensible level, the obvious reason for potential error in the European elections was the uncertainty about which respondents would actually vote in an election with a small turnout. I think that uncertainty was clear to everyone throughout.
And indeed the reason people here expected TBP to do worse than had been indicated by most of the polls was the difference in turnout changes between Brexit-supporting and non-Brexit-supporting areas, which was very fully discussed here between Thursday and Sunday.
That suggests to me that what YouGov got wrong was their modelling of turnout, and nothing whatsoever to do with whether they prompted or not for the party that was fairly clearly going to top the poll in any case.
But the two are tied up together, as I have already pointed out. A new party does not have the infrastructure to get out the vote. That means that their voters are going to have to be that bit more motivated.
This may also explain why both Labour and the Conservatives came at the bottom end of expectations this time: their get out the vote machinery was not working for them this time.
Polls are not trying to measure abstract party allegiance. They're trying to measure who will actually cast votes.
The Tory GOTV leafleting operation was weird in 2017, this house then that house 100 yards down the road...
> @Richard_Nabavi said: > Tory moderates are beginning to coalesce around candidates they believe could defeat Dominic Raab, rather than Boris Johnson, as several MPs told the Guardian they considered the former Brexit secretary the bigger threat. > > The moves against Raab from moderates are likely to also benefit Johnson and the environment secretary, Michael Gove, who Tory backbenchers believe is gaining momentum over his cabinet rival Jeremy Hunt. > > https://www.theguardian.com/politics/2019/may/31/stop-dominic-raab-tory-moderates-seek-to-block-path-to-pm > > Raab is a big help to the Boris campaign.
> Tory moderates are beginning to coalesce around candidates they believe could defeat Dominic Raab, rather than Boris Johnson, as several MPs told the Guardian they considered the former Brexit secretary the bigger threat.
> > The moves against Raab from moderates are likely to also benefit Johnson and the environment secretary, Michael Gove, who Tory backbenchers believe is gaining momentum over his cabinet rival Jeremy Hunt.
To an extent, but perhaps more importantly an illustration that you can't look at candidates' popularity in isolation, you have to look at them against each other, in the context of the calculations MPs will be making at each stage of the contest.
> @Carolus_Rex said: > > @Stark_Dawning said: > > > @HYUFD said: > > > > > https://twitter.com/Nigel_Farage/status/1134365557325799424 > > > > > > > > > > > > Another tin-foil hat lunatic airs his populist conspiracy theories. Ignore. > > > > Surely Nigel doesn't actually believe that the 'establishment' is controlling YouGov's polling methodology? I think he just wants to instil his followers with a lot of anger, a lot of hate and a lot of victimhood. Expect much more of this. > > Or he's just been spending too much time around the Donald.
Farage is probably the best politician we've got at the moment, but I really don't want to see him heading down the path of this Cadwalladr/Adonis/Esler style nonsense. He should ignore the polls and I'm sure if he focuses on and wins Peterborough they might even take the lead with yougov.
> @Endillion said: > Is Corbyn toast ? > > > > He was safe when polling was up - it covered up the weaknesses. > > > > Could he be gone this year or will he cling on for the GE defeat ? > > No chance. His supporters will remain convinced that since he made up a huge polling deficit once, he can do it again. > > In some ways, he's probably happier starting from way behind. Certainly he'll do better if people are more relaxed about his likelihood of winning.
Aren't Labour ahead in the polling averages? I mean, down but still ahead. The way things are going, he's the likeliest PM post-election.
> > The obvious question is how they can determine which is more accurate - the figure with prompting, or the figure without prompting.
>
> ------
>
>
>
> By comparing with actual election results, like the ones last week in which the polls that prompted for the Brexit Party were proved to have overstated their support.
>
>
>
> Polling relies on a lot of social science and understanding of psychology. It's not a case of just giving a model of a ballot paper to a sample of people and reporting the raw results.
>
> Did the pollster who had BXP on 27 prompt?
>
> We’re missing the original point which was NF crying deliberate foul play on the part of the pollsters. Now most reasonable people would say that was b****cks only fellow conspiracy theory believers would agree with him
No - you need to make an argument why he was wrong. Not just invoking "most reasonable people". They aren't to be found here. They have better things to do.
You gov clearly explained what they did and why. To claim it was done deliberately is more than twisting the truth.
> @williamglenn said: > > @Chris said: > > > > That's an article about opinion polls in the USA and support for non-Democratic/Republican candidates. > > > > You actually claimed that YouGov had said this - and I presumed at the time you meant they had said it about British elections. > > > > Has YouGov actually said this or not? > --------- > > Yes, it's in their statement today. I gave you a US source to show that it's a generally accepted phenomenon, not a conspiracy against the Brexit Party. > > https://yougov.co.uk/topics/politics/articles-reports/2019/05/31/heres-how-we-prompt-brexit-party-and-why-its-more- > > "In the past, prompting for smaller parties has tended to overstate their support when compared to actual elections, and the two-stage approach to prompting has produced more accurate results."
Thank you for finally clarifying that.
I note that YouGov say they don't know why they overestimated the TBP share of the vote, and that the list of possible reasons they are investigating doesn't even include prompting - far from your claim that "it [prompting] resulted in them significantly overestimating their support."
So certainly YouGov is not trying to justify the lack of prompting in its Westminster polls on the basis of their got the European polls wrong.
> @Richard_Nabavi said: > I am sure that YouGov, with a mere 19 years of experience and over 800 employees, who do nothing other than conduct polls and analyse them, would be foolish indeed to think they might know a tad more than @Chris about how to get the best results.
Of course you're right. Large commercial companies always know best. Even if they're talking complete bollocks.
> @ah009 said: > > @Endillion said: > > Is Corbyn toast ? > > > > > > > > He was safe when polling was up - it covered up the weaknesses. > > > > > > > > Could he be gone this year or will he cling on for the GE defeat ? > > > > No chance. His supporters will remain convinced that since he made up a huge polling deficit once, he can do it again. > > > > In some ways, he's probably happier starting from way behind. Certainly he'll do better if people are more relaxed about his likelihood of winning. > > Aren't Labour ahead in the polling averages? I mean, down but still ahead. The way things are going, he's the likeliest PM post-election.
If there were a market for PM after GE, he should be a very weak favourite right now because we know
i) He is running a party probably competing in all 650 seats (Exclude all NI, SNP and Plaid) ii) He is currently a party leader (Tories, Lib Dems are changing theirs) iii) He will run for a seat (Farage might not) iv) Seats below 200 are stickier for Labour on UNS than other parties on split vote shares. v) He is more confidence and supply friendly to the Lib Dems & Greens for the right deal (Revocation/2nd referendum); SNP (2nd indy ref) ? than the Tories. The Tories could be c&s friendly to the BXP or the other way round mind...
Against this we have:
i) His next PM ratings are absolubtely appalling. These are a good long term indicator that say he won't be PM. ii) He's clearly uncomfortable leading a party with such a remainy membership when in his heart he is a leaver. iii) Labour's anti-semitism issues. iv) Labour's polling is heading south right now. v) He could be challenged for the leadership and he is getting older so might pass it over voluntarily.
> @Pulpstar said: > i) His next PM ratings are absolubtely appalling. These are a good long term indicator that say he won't be PM. > ii) He's clearly uncomfortable leading a party with such a remainy membership when in his heart he is a leaver. > iii) Labour's anti-semitism issues. > iv) Labour's polling is heading south right now. > v) He could be challenged for the leadership and he is getting older so might pass it over voluntarily.
i) Personal polling is terrible. In ordinary circumstances, it's fatal but these aren't ordinary circumstances. ii) Doesn't matter that he's uncomfortable, he's still leader iii) Already factored into the polling. It ought to put more people off, I agree, but it isn't. iv) Previous trend has been down. I'm really not sure whether past trends can ever be used to predict future trends. Could be like the stockmarket, pseudorandom. v) Very good point. I heard rumours recently he was on the brink of leaving.
There is no evidence that the equivalent public sector position (if it is possible to draw comparisons) pays less than private, and most studies that I have seen point to the opposite. What is a clear fact is that the public, while not so much "job for life" (unless you are a doctor) as it used to be, is massively safer than private and still carries much better pension arrangements. And that is fundamentally unfair on most workers
Thanks for the job link, but no I have no need of a job in the civil service, and the suggestion was a bit silly. I was not suggesting it is easy to be in the civil service, or any other public sector job.
I just offered you evidence. Me. Over the last year I have changed jobs (pay rise, thank you God) and some of the jobs I applied for were Service jobs. I know what my skills will get in the market but the Service is good at training and that is valuable. However when the offers were in the larger salary and smaller training in the private sector outweighed the public sector opportunities by a considerable amount: not in the sense of smaller holidays, but not being able to actually afford the commute or maintain my health (the South East is very expensive). So from actual empirical evidence (albeit with a small sample size) it contradicts your point.
As for the studies you mention (but do not cite) haven't we gone beyond that now? There is a tendency in modern society to quote studies at each other. Whilst a good approach it is not enough and one must examine the studies, extract the data and do one's own research. Cases where studies are cited far beyond their plausibility include the Laffer curve, devaluation stimulating the economy, and betting odds being a good predictor of elections.
Comments
> > @HYUFD said:
>
> > https://twitter.com/Nigel_Farage/status/1134365557325799424
>
>
>
>
>
> Another tin-foil hat lunatic airs his populist conspiracy theories. Ignore.
>
> Surely Nigel doesn't actually believe that the 'establishment' is controlling YouGov's polling methodology? I think he just wants to instil his followers with a lot of anger, a lot of hate and a lot of victimhood. Expect much more of this.
Or he's just been spending too much time around the Donald.
> https://twitter.com/davidallengreen/status/1134431944798801920
>
>
>
> And John Redwood.
No one likes to talk about John Redwood, understandably.
Actually from being the go-to Tory loon nonpareil, he's been somewhat eclipsed by the current rich array. He must feel a bit miffed.
Their statement today seems entirely reasonable.
Fakhar would have gotten more, but he wasn't prompted for
As I understand it, most new civil servants/doctors/teachers etc get some form of career average pension, which has been set up to be slightly less generous than the previously available final salary, but is still vastly superior to anything available in the private sector. This is largely due to the hugely increased level of security that the eventual payments will cover the individual's lifestyle following retirement, but is also expected to be better in pure expected value terms as well (although obviously this is difficult to determine since it involves making assumptions about long term annuity rates/life expectancy improvements etc).
I'd still bite your hand off for a government sponsored pension arrangement, but agree that I probably wouldn't want to accept the shorter term drop in salary that goes with it.
But really I don't think "we're still testing whether this makes any difference" is sufficient justification for concluding it doesn't make any difference. When a party has been registering in the 20s both in opinion polls and actual elections, it makes no sense to exclude it from the list of named parties.
>
> I hadn't realised YouGov had been quoted on this, but thank you for letting me know about it.
>
> But really I don't think "we're still testing whether this makes any difference" is sufficient justification for concluding it doesn't make any difference. When a party has been registering in the 20s both in opinion polls and actual elections, it makes no sense to exclude it from the list of named parties.
---------------
YouGov explained in their statement. For the European Election polling, they did name them, and it resulted in them significantly overestimating their support. YouGov had them at 37% in those polls.
> > @anothernick said:
> > > @kamski said:
> > > Anyone know why sterling has fallen again against the euro the last 2 days? I would have thought with Hammond's clear anti-no-deal statement the pound might have perked up slightly?
> >
> > Market beginning to realise that Brexit uncertainty is not going to be resolved anytime soon.
> >
> > City types have been extraordinarily optimistic about Brexit - I have lost count of the number of presentations I have been to (as part of my job) at which managers have opined that sterling is set for a rise as soon as the markets have clarity. Up until recently the comforting assumption has been that clarity was just around the corner, it is only now that they are beginning to realise that it may be delayed for a decade or two.
>
> Interesting, tho I assume it would depend on the type of "clarity" - wouldn't No-Deal clarity lead to sterling falling further? And conversely No-Deal being ruled out would lead to a rise from current levels?
> Or was the assumption up to now that clarity would definitely involve No-Deal being ruled out?
>
> Maybe the calculation is that a competitive Conservative leadership election will lead to contenders all promising No-Deal (if Unicorn doesn't appear), and it now looks more likely to be competitive than a few days ago?
Another tenet of city thinking has been that the chances of no deal were minimal and in that they have been proved correct. But although we have avoided no deal we are no closer to defining our long-term relationship with the EU.
> > @eek said:
>
> > > @Chris said:
>
> > > > @ah009 said:
>
> > > > > @HYUFD said:
>
> > > > > https://twitter.com/Nigel_Farage/status/1134365557325799424
>
>
>
> > > >
>
> > > > Another tin-foil hat lunatic airs his populist conspiracy theories. Ignore.
>
> > >
>
> > > Actually, it's a fair point, isn't it?
>
> > >
>
> > > Much as I hate and loathe the Brexit Party, and would like to see its leader soaked in milk shakes for the rest of his natural life, It does seem a bit peculiar to categorise them under "Others" when they have just topped the poll in a national election.
>
> >
>
> > Hardly, given that Brexit polled higher in the EU polls than they actually got in the EU election itself.
>
>
>
> Why does that make it fair to class them among Others?
>
> The aim of an opinion poll is to get an accurate poll reading, not to massage the ego of the party's leader. YouGov are entitled to use whatever methodology they believe works best. There's no evidence as yet that their poll rating is being underrecorded.
>
> Their statement today seems entirely reasonable.
No evidence - apart from all the evidence we already have that if you prompt people with certain alternatives and exclude others, the alternatives that people are prompted with will benefit.
Disappointing to see you being so naive - or disingenuous.
> > @Chris said:
> >
> > I hadn't realised YouGov had been quoted on this, but thank you for letting me know about it.
> >
> > But really I don't think "we're still testing whether this makes any difference" is sufficient justification for concluding it doesn't make any difference. When a party has been registering in the 20s both in opinion polls and actual elections, it makes no sense to exclude it from the list of named parties.
> ---------------
>
> YouGov explained in their statement. For the European Election polling, they did name them, and it resulted in them significantly overestimating their support. YouGov had them at 37% in those polls.
You say "it resulted in them significantly overestimating their support". But as YouGov are saying there's no evidence it makes any difference, I suspect you're making that up.
> Surely Nigel doesn't actually believe that the 'establishment' is controlling YouGov's polling methodology? I think he just wants to instil his followers with a lot of anger, a lot of hate and a lot of victimhood. Expect much more of this.
Exactly. It's the old "poor me" control drama.
Covers up for the fact that they have only one policy, keeps the spotlight on him, fuels that nasty populist conflagration the Kremlin has been nursing.
> > @nichomar said:
> > > @brokenwheel said:
> >
> > > Back in the 2010 GE campaign when the LDs were polling at a similar level their support was 25% of ABC1 and 23% of C2DE. It’s now 29% ABC1 and 16% C2DE.
> >
> > The Brexit Party meanwhile are 29% C2DE and 17% ABC1.
> >
> > We thought class politics was dying in 2017 with the Tories and Labour with the Tories having just a 2% higher lead with ABC1s over Labour than with C2DEs, if it is Brexit Party v LDs then class politics is back with a vengeance
> >
> > I don’t really understand why the group who stand to suffer most by no deal brexit are the most keen on it.
>
> Immigration and sovereignty and they have less to lose anyway
________________________
Some of 'the other half' considered that there was nothing to lose, so they might as well vote that way as a gigantic protest at all manner of things which they thought were less good than before. That's partly what 40 years of Thatcher and Sons does to a country.
I'm not saying that they all voted that way for that reason. But enough did to give the unexpected referendum result. Douglas Alexander ex-MP got at least part of the message via the usual drastic route, i.e. he lost his seat! Still, he makes a reasonable radio presenter, just like Rory Stewart made a good TV series https://www.bbc.co.uk/programmes/m0001d9y
Now it may well be that the correct action is to take a different action to compensate. But the idea that the Brexit party is being hard done to by pollsters (to what purpose, goodness only knows) is absurd.
>
> The Brexit Party won Epping Forest, Buckingham, Wellingborough, Tonbridge, Thurrock, Mansfield, Ryedale etc even a few Tory Remain seats like Tunbridge Wells and Maidenhead last week (Stoke already has a Tory seat, also lost to Farage).
>
> In fact the Tories lost more areas to the Brexit Party than the LDs last week and far more of their voters
----------
The attempt to win back votes from the Brexit Party comes at the cost of losing the votes of people like Richard Nabavi. In addition, by splitting the combined Brexit vote, you help Remain parties to win.
>
> You say "it resulted in them significantly overestimating their support". But as YouGov are saying there's no evidence it makes any difference, I suspect you're making that up.
------
No, YouGov are saying that there is substantial historical evidence that prompting for minor/new parties results in overstating their support, and on the face of it that's what happened last week.
> No evidence - apart from all the evidence we already have that if you prompt people with certain alternatives and exclude others, the alternatives that people are prompted with will benefit.
>
>
>
> Disappointing to see you being so naive - or disingenuous.
>
> You're missing the point completely. The Brexit party have had one outing. In that one outing, their poll ratings were generally overrecorded rather than underrecorded. This suggests that pollsters are finding too many Brexit party supporters. Anything that increases their poll rating runs the risk of increasing that error.
>
> Now it may well be that the correct action is to take a different action to compensate. But the idea that the Brexit party is being hard done to by pollsters (to what purpose, goodness only knows) is absurd.
Finding too many, or overestimating the chance that they will actually get off their arse and vote?
> > @noneoftheabove said:
> > > @HYUFD said:
> > > > @Nigel_Foremain said:
> > > > > @HYUFD said:
> > > > > > @brokenwheel said:
> > > > > > Back in the 2010 GE campaign when the LDs were polling at a similar level their support was 25% of ABC1 and 23% of C2DE. It’s now 29% ABC1 and 16% C2DE.
> > > > >
> > > > > The Brexit Party meanwhile are 29% C2DE and 17% ABC1.
> > > > >
> > > > > We thought class politics was dying in 2017 with the Tories and Labour with the Tories having just a 2% higher lead with ABC1s over Labour than with C2DEs, if it is Brexit Party v LDs then class politics is back with a vengeance
> > > >
> > > > Only if you define class by intellect, then you might be correct. There are a lot of bright "working class" people I have met who think Brexit is for nutters. Now I have lit the touch paper, I had better go to lunch!!
> > >
> > > There are also some rich Brexit Party voters, I suspect the divide is more based on the class divide in terms of education than wealth and income though the divide is still there even with the latter (wealthier Leave voters are more likely to be soft Brexiteers and thus less likely to support the No Deal Brexit Party)
> >
> > Given the age profiles it would be interesting to see the net worth profiles of Brexit v LD, I would not be surprised if the mean Brexit net worth > mean LD net worth despite the class differences above (even excluding Putin).
>
> No way.
>
> LDs will be far wealthier on average than Brexit Party voters if they lead with ABC1s including lawyers, doctors and managers but the Brexit Party lead with the working class.
>
> Thus the LDs won Richmond Park, Islington, Kensington and Esher and Walton last week but the Brexit Party won Bolsover, Merthyr Tydfil, Wigan and Stoke.
>
> The middle class battle at the moment is LD v Tory, the working class battle is Brexit Party v Labour with the former winning currently in each case
Again you underestimate the impact of the last twenty years on those under 45 which is half the LD vote but less than a third of the BP vote. Most of the LD ABC1 <45s will have negative net worth due to cost of going to university and high housing costs. Only 5% of under 35s are in the top 40% of net worth and 44% are in the bottom 20%. Net worth is so skewed by age that it seems unlikely that the age effect is offset by the class impact on voting.
https://twitter.com/PM4EastRen/status/1134185540134277127
https://twitter.com/PM4EastRen/status/1134187748313718785
> No evidence - apart from all the evidence we already have that if you prompt people with certain alternatives and exclude others, the alternatives that people are prompted with will benefit.
>
>
>
> Disappointing to see you being so naive - or disingenuous.
>
> You're missing the point completely. The Brexit party have had one outing. In that one outing, their poll ratings were generally overrecorded rather than underrecorded. This suggests that pollsters are finding too many Brexit party supporters. Anything that increases their poll rating runs the risk of increasing that error.
>
> Now it may well be that the correct action is to take a different action to compensate. But the idea that the Brexit party is being hard done to by pollsters (to what purpose, goodness only knows) is absurd.
That's a completely unscientific attitude. So what if YouGov did overestimate Brexit Party support? The way to correct that is to find out why and make appropriate adjustments, not to adopt a method that results in an underestimate of the Brexit Party vote for an entirely unrelated reason, and to hope that that error will cancel out the other error that they don't understand.
No wonder opinion polling is in such a mess if that's the way the pollsters operate,.
>
> That's a completely unscientific attitude. So what if YouGov did overestimate Brexit Party support? The way to correct that is to find out why and make appropriate adjustments, not to adopt a method that results in an underestimate of the Brexit Party vote for an entirely unrelated reason, and to hope that that error will cancel out the other error that they don't understand.
>
> No wonder opinion polling is in such a mess if that's the way the pollsters operate,.
--------
Something doesn't become unscientific just because you say it is. The whole reason why they have an 'other' category in the first place is because of the evidence that prompting for minor parties distorts the results.
> > @Dura_Ace said:
>
> > https://twitter.com/thebeakbrewery/status/1134002379765010432
>
>
>
> >
>
> >
>
> >
>
> > We are 12-18 months away from Sarajevo '92.
>
>
>
> That is ludicrous
>
> There were plenty of people in SFRY expressing similar sentiments right up until the day they were scavenging for snow dusted firewood in sniper alley.
Go and have a nice sit down in a quiet room
> > @Chris said:
> >
> > You say "it resulted in them significantly overestimating their support". But as YouGov are saying there's no evidence it makes any difference, I suspect you're making that up.
> ------
>
> No, YouGov are saying that there is substantial historical evidence that prompting for minor/new parties results in overstating their support, and on the face of it that's what happened last week.
If they are saying that, please can you quote them?
All I see in what they were quoted as seeing earlier is that they are still trying to determine whether the Brexit Party's support was affected by prompting, and this on their previous policy about UKIP:
"We regularly tested the effect of prompting on UKIP support, and, once it seemed it was no longer giving them an artifical boost, we started including UKIP in the main prompt alongside Labour, the Conservatives and Liberal Democrats."
The obvious question is how they can determine which is more accurate - the figure with prompting, or the figure without prompting. The fact that the two figures for UKIP apparently converged doesn't, of course, give us any information about which was the more accurate before they converged. It's entirely possible that the figure without prompting was too low!
And why did they prompt for BXP at the Euros? They were a minor party there for all we knew
> > @AlastairMeeks said:
>
> > No evidence - apart from all the evidence we already have that if you prompt people with certain alternatives and exclude others, the alternatives that people are prompted with will benefit.
>
> >
>
> >
>
> >
>
> > Disappointing to see you being so naive - or disingenuous.
>
> >
>
> > You're missing the point completely. The Brexit party have had one outing. In that one outing, their poll ratings were generally overrecorded rather than underrecorded. This suggests that pollsters are finding too many Brexit party supporters. Anything that increases their poll rating runs the risk of increasing that error.
>
> >
>
> > Now it may well be that the correct action is to take a different action to compensate. But the idea that the Brexit party is being hard done to by pollsters (to what purpose, goodness only knows) is absurd.
>
>
>
> That's a completely unscientific attitude. So what if YouGov did overestimate Brexit Party support? The way to correct that is to find out why and make appropriate adjustments, not to adopt a method that results in an underestimate of the Brexit Party vote for an entirely unrelated reason, and to hope that that error will cancel out the other error that they don't understand.
>
>
>
> No wonder opinion polling is in such a mess if that's the way the pollsters operate,.
>
> This is not a new problem. Pollsters have had this problem with the Greens and with UKIP in the past too. Part of the problem relates to the lack of party infrastructure, as well as voter recognition, the ability to get out the party vote. The question of when to move the minor party from others to the big boys' list is not straightforward and isn't necessarily a question of how much they are polling.
Obviously you've nothing there to say that they should not prompt rather than prompting.
If you want to defend not prompting, you need some evidence that not prompting gives more accurate results.
>
> The obvious question is how they can determine which is more accurate - the figure with prompting, or the figure without prompting.
------
By comparing with actual election results, like the ones last week in which the polls that prompted for the Brexit Party were proved to have overstated their support.
Polling relies on a lot of social science and understanding of psychology. It's not a case of just giving a model of a ballot paper to a sample of people and reporting the raw results.
> > @AlastairMeeks said:
> > > @AlastairMeeks said:
> >
> > > No evidence - apart from all the evidence we already have that if you prompt people with certain alternatives and exclude others, the alternatives that people are prompted with will benefit.
> >
> > >
> >
> > >
> >
> > >
> >
> > > Disappointing to see you being so naive - or disingenuous.
> >
> > >
> >
> > > You're missing the point completely. The Brexit party have had one outing. In that one outing, their poll ratings were generally overrecorded rather than underrecorded. This suggests that pollsters are finding too many Brexit party supporters. Anything that increases their poll rating runs the risk of increasing that error.
> >
> > >
> >
> > > Now it may well be that the correct action is to take a different action to compensate. But the idea that the Brexit party is being hard done to by pollsters (to what purpose, goodness only knows) is absurd.
> >
> >
> >
> > That's a completely unscientific attitude. So what if YouGov did overestimate Brexit Party support? The way to correct that is to find out why and make appropriate adjustments, not to adopt a method that results in an underestimate of the Brexit Party vote for an entirely unrelated reason, and to hope that that error will cancel out the other error that they don't understand.
> >
> >
> >
> > No wonder opinion polling is in such a mess if that's the way the pollsters operate,.
> >
> > This is not a new problem. Pollsters have had this problem with the Greens and with UKIP in the past too. Part of the problem relates to the lack of party infrastructure, as well as voter recognition, the ability to get out the party vote. The question of when to move the minor party from others to the big boys' list is not straightforward and isn't necessarily a question of how much they are polling.
>
> Obviously you've nothing there to say that they should not prompt rather than prompting.
>
> If you want to defend not prompting, you need some evidence that not prompting gives more accurate results.
We don't know. What we do know is that yougov over estimated Brexit last Thursday using whatever method they used then.
And the purpose of a polling company is to provide accurate results not to ask questions that make people (not paying for said polling) happy.
ICM "won" 2015
Survation worked very well for 2017 iirc
Ipsos Mori for 2019 Euros.
No guarantee Yougov will be right for 2022 or w/e
> What do you think the Lib Dem score would be if they weren’t prompted?
The answer is: probably less accurate.
That's all that matters. Accuracy.
> > @Chris said:
> >
> > That's a completely unscientific attitude. So what if YouGov did overestimate Brexit Party support? The way to correct that is to find out why and make appropriate adjustments, not to adopt a method that results in an underestimate of the Brexit Party vote for an entirely unrelated reason, and to hope that that error will cancel out the other error that they don't understand.
> >
> > No wonder opinion polling is in such a mess if that's the way the pollsters operate,.
> --------
>
> Something doesn't become unscientific just because you say it is. The whole reason why they have an 'other' category in the first place is because of the evidence that prompting for minor parties distorts the results.
Sorry, but I don't think you're thinking about what you're saying.
You talk about prompting for minor parties "distorting" the results. Obviously it's true that prompting for minor parties benefits minor parties. But that doesn't mean it's a "distortion".
In a real election, there is prompting for minor parties (if TBP can still be called that), both during the campaign, and by reason of their presence on the ballot paper.
The question is whether pollsters' prompting or not prompting produces a more accurate result. That requires evidence, not just assertion. Remember that generally the Lib Dems' poll rating has tended to improve during election campaigns, because of their exposure. It's not exactly the same issue, but it's obviously closely related.
> > @Chris said:
> > > @AlastairMeeks said:
> > > > @AlastairMeeks said:
> > >
> > > > No evidence - apart from all the evidence we already have that if you prompt people with certain alternatives and exclude others, the alternatives that people are prompted with will benefit.
> > >
> > > >
> > >
> > > >
> > >
> > > >
> > >
> > > > Disappointing to see you being so naive - or disingenuous.
> > >
> > > >
> > >
> > > > You're missing the point completely. The Brexit party have had one outing. In that one outing, their poll ratings were generally overrecorded rather than underrecorded. This suggests that pollsters are finding too many Brexit party supporters. Anything that increases their poll rating runs the risk of increasing that error.
> > >
> > > >
> > >
> > > > Now it may well be that the correct action is to take a different action to compensate. But the idea that the Brexit party is being hard done to by pollsters (to what purpose, goodness only knows) is absurd.
> > >
> > >
> > >
> > > That's a completely unscientific attitude. So what if YouGov did overestimate Brexit Party support? The way to correct that is to find out why and make appropriate adjustments, not to adopt a method that results in an underestimate of the Brexit Party vote for an entirely unrelated reason, and to hope that that error will cancel out the other error that they don't understand.
> > >
> > >
> > >
> > > No wonder opinion polling is in such a mess if that's the way the pollsters operate,.
> > >
> > > This is not a new problem. Pollsters have had this problem with the Greens and with UKIP in the past too. Part of the problem relates to the lack of party infrastructure, as well as voter recognition, the ability to get out the party vote. The question of when to move the minor party from others to the big boys' list is not straightforward and isn't necessarily a question of how much they are polling.
> >
> > Obviously you've nothing there to say that they should not prompt rather than prompting.
> >
> > If you want to defend not prompting, you need some evidence that not prompting gives more accurate results.
>
> We don't know. What we do know is that yougov over estimated Brexit last Thursday using whatever method they used then.
>
> And the purpose of a polling company is to provide accurate results not to ask questions that make people (not paying for said polling) happy.
But you have no evidence that they overestimated TBP because they were prompting. If the YouGov statement makes any sense at all, it implies that they don't prompt until they are satisfied it makes no difference!
So what sense does it make to try to correct the unknown error by making a different change in their polling method which they'd previously decided made no difference? That's what I mean by unscientific.
People always end up believing their own hype.
He was safe when polling was up - it covered up the weaknesses.
Could he be gone this year or will he cling on for the GE defeat ?
>
> You talk about prompting for minor parties "distorting" the results. Obviously it's true that prompting for minor parties benefits minor parties. But that doesn't mean it's a "distortion".
>
> In a real election, there is prompting for minor parties (if TBP can still be called that), both during the campaign, and by reason of their presence on the ballot paper.
>
> The question is whether pollsters' prompting or not prompting produces a more accurate result. That requires evidence, not just assertion. Remember that generally the Lib Dems' poll rating has tended to improve during election campaigns, because of their exposure. It's not exactly the same issue, but it's obviously closely related.
----------------
Answering a poll has no consequences but voting in an election does. For very understandable reasons, if you prompt for minor parties in a poll, you will get higher results than you would in a real election, so it is a 'distortion'. This is borne out by substantial evidence and is why YouGov use the solution of putting minor parties behind "other", to find the people who really do want to vote for them.
> > @Pulpstar said:
>
> > Final salary is an abomination in this day and age, 10+10 defined contribution superannuation along the lines of Oz is the way to go.
>
>
>
> Don't even get me started on public sector pensions. People talk about unfairness. How can in be fair that "The Few" can retire in their mid 50s while "The Many" have to carry on working until they are in their 70s. Why won't it change? Because the powers that be (including the Union Barons) like their own arrangements too much. This includes Nigel-man-o-the-people-Farridge
>
> Final salary pensions have not been available to Civil Servants since about 2005/6. The only ones still extant are those who worked before then. If you think jobs in the civil service are so easy and lucrative, the job site is/was jobs.civilservice.gov.uk. but you have to sit one test (two for the professions) an aptitude test, a panel interview, a security check, an enhanced background check for some jobs, and if you are successful the pay is crap and the annual payrises zero. I earn about £18k plus bonus more in the private sector than I would get in the public one. It is not as easy or as lucrative as you think.
>
> The real divide is between defined benefit (DB) and defined contribution (DC) schemes, rather than just final salary (which are a form of DB) and everything else. The reason being that on DB schemes, investment risk is on the sponsor while with DC it gets transferred to the employee. In addition, since civil service pension arrangements aren't funded, the only risk remaining with the employee is that the government refuses to pay up when the time comes, which is assumed to be vanishingly unlikely.
>
> As I understand it, most new civil servants/doctors/teachers etc get some form of career average pension, which has been set up to be slightly less generous than the previously available final salary, but is still vastly superior to anything available in the private sector. This is largely due to the hugely increased level of security that the eventual payments will cover the individual's lifestyle following retirement, but is also expected to be better in pure expected value terms as well (although obviously this is difficult to determine since it involves making assumptions about long term annuity rates/life expectancy improvements etc).
>
> I'd still bite your hand off for a government sponsored pension arrangement, but agree that I probably wouldn't want to accept the shorter term drop in salary that goes with it.
Credit to Royal Mail for starting to develop a hybrid DB/DC, with union support and government permission.
> > @Chris said:
> >
> > The obvious question is how they can determine which is more accurate - the figure with prompting, or the figure without prompting.
> ------
>
> By comparing with actual election results, like the ones last week in which the polls that prompted for the Brexit Party were proved to have overstated their support.
The question is whether there is any evidence that YouGov's overestimate was owing to prompting rather than any number of other possible shortcomings in their methods.
You claimed before that "there is substantial historical evidence that prompting for minor/new parties results in overstating their support" and I asked you to quote the evidence for that. Please can you do so?
Plaid 0.5%
UKIP 3%
SNP 3.5%
CHUK 4%
Green 8%
Tories 11%
Lib Dem 15.5%
Labour 19%
Brexit 32.5%
Bounds:
Plaid 0 - 1%
UKIP 2 - 4%
SNP 3 - 4%
CHUK 3 - 5%
Green 4 - 12%
Tories 7 - 14%
Lib Dem 12 - 19%
Labour 13 - 25%
BXP 27 - 38 %
Actual results for GB :
Plaid 1.0% (+.5%)
UKIP 3.3% (+.3%)
SNP 3.6% (+.1%)
CHUK 3.4% (-.6%)
Green 12.09% (+4.09%)
Tories 9.09% (-1.81%)
Lib Dem 20.3% (+4.8%)
Labour 14.1% (-4.9%)
BRX 31.6% (-0.9%)
Absolubte error of 18% over 9 parties.
In some ways, he's probably happier starting from way behind. Certainly he'll do better if people are more relaxed about his likelihood of winning.
> > @Chris said:
>
> >
>
> > The obvious question is how they can determine which is more accurate - the figure with prompting, or the figure without prompting.
>
> ------
>
>
>
> By comparing with actual election results, like the ones last week in which the polls that prompted for the Brexit Party were proved to have overstated their support.
>
>
>
> Polling relies on a lot of social science and understanding of psychology. It's not a case of just giving a model of a ballot paper to a sample of people and reporting the raw results.
>
> Did the pollster who had BXP on 27 prompt?
>
> We’re missing the original point which was NF crying deliberate foul play on the part of the pollsters. Now most reasonable people would say that was b****cks only fellow conspiracy theory believers would agree with him
No - you need to make an argument why he was wrong. Not just invoking "most reasonable people". They aren't to be found here. They have better things to do.
> I noted this from simple averages at the time (May 22) for the euro polling from the minimum and maximum range bounds
>
> Plaid 0.5%
> UKIP 3%
> SNP 3.5%
> CHUK 4%
> Green 8%
> Tories 11%
> Lib Dem 15.5%
> Labour 19%
> Brexit 32.5%
>
> Bounds:
>
> Plaid 0 - 1%
> UKIP 2 - 4%
> SNP 3 - 4%
> CHUK 3 - 5%
> Green 4 - 12%
> Tories 7 - 14%
> Lib Dem 12 - 19%
> Labour 13 - 25%
> BXP 27 - 38 %
>
> Actual results for GB :
>
> Plaid 1.0% (+.5%)
> UKIP 3.3% (+.3%)
> SNP 3.6% (+.1%)
> CHUK 3.4% (-.6%)
> Green 12.09% (+4.09%)
> Tories 9.09% (-1.81%)
> Lib Dem 20.3% (+4.8%)
> Labour 14.1% (-4.9%)
> BRX 31.6% (-0.9%)
>
> Absolubte error of 18% over 9 parties.
>
>
So pollsters were actually underestimating BXP?
>
> Did the pollster who had BXP on 27 prompt?
That was Kantar, who had Labour on 24% so had major sampling issues. Yes, they did prompt for the Brexit Party.
https://uk.kantar.com/public-opinion/politics/2019/the-brexit-party-has-four-point-lead-over-labour-ahead-of-european-parliament-elections/
>
> The question is whether there is any evidence that YouGov's overestimate was owing to prompting rather than any number of other possible shortcomings in their methods.
>
> You claimed before that "there is substantial historical evidence that prompting for minor/new parties results in overstating their support" and I asked you to quote the evidence for that. Please can you do so?
Here you go:
https://hbr.org/2016/11/how-polls-overestimate-support-for-third-party-candidates
> > @Pulpstar said:
>
> > I noted this from simple averages at the time (May 22) for the euro polling from the minimum and maximum range bounds
>
> >
>
> > Plaid 0.5%
>
> > UKIP 3%
>
> > SNP 3.5%
>
> > CHUK 4%
>
> > Green 8%
>
> > Tories 11%
>
> > Lib Dem 15.5%
>
> > Labour 19%
>
> > Brexit 32.5%
>
> >
>
> > Bounds:
>
> >
>
> > Plaid 0 - 1%
>
> > UKIP 2 - 4%
>
> > SNP 3 - 4%
>
> > CHUK 3 - 5%
>
> > Green 4 - 12%
>
> > Tories 7 - 14%
>
> > Lib Dem 12 - 19%
>
> > Labour 13 - 25%
>
> > BXP 27 - 38 %
>
> >
>
> > Actual results for GB :
>
> >
>
> > Plaid 1.0% (+.5%)
>
> > UKIP 3.3% (+.3%)
>
> > SNP 3.6% (+.1%)
>
> > CHUK 3.4% (-.6%)
>
> > Green 12.09% (+4.09%)
>
> > Tories 9.09% (-1.81%)
>
> > Lib Dem 20.3% (+4.8%)
>
> > Labour 14.1% (-4.9%)
>
> > BRX 31.6% (-0.9%)
>
> >
>
> > Absolubte error of 18% over 9 parties.
>
> >
>
> >
>
>
>
> So pollsters were actually underestimating BXP?
>
> No, according to the (very rough and ready) methodology I laid out a priori to the Euros, BXP were overestimated by 0.9%.
Hah. Blame my lack of coffee for that one.
Hopefully not all 12 will still be in the hunt once polling commences.
> > @williamglenn said:
> > > @Chris said:
> > >
> > > The obvious question is how they can determine which is more accurate - the figure with prompting, or the figure without prompting.
> > ------
> >
> > By comparing with actual election results, like the ones last week in which the polls that prompted for the Brexit Party were proved to have overstated their support.
>
> The question is whether there is any evidence that YouGov's overestimate was owing to prompting rather than any number of other possible shortcomings in their methods.
>
> You claimed before that "there is substantial historical evidence that prompting for minor/new parties results in overstating their support" and I asked you to quote the evidence for that. Please can you do so?
To try to get the discussion back on to a semi-sensible level, the obvious reason for potential error in the European elections was the uncertainty about which respondents would actually vote in an election with a small turnout. I think that uncertainty was clear to everyone throughout.
And indeed the reason people here expected TBP to do worse than had been indicated by most of the polls was the difference in turnout changes between Brexit-supporting and non-Brexit-supporting areas, which was very fully discussed here between Thursday and Sunday.
That suggests to me that what YouGov got wrong was their modelling of turnout, and nothing whatsoever to do with whether they prompted or not for the party that was fairly clearly going to top the poll in any case.
To your main point, he is presumably free to provide any evidence he has at any point via whatever channel he wishes. I'm not holding my breath.
> > @williamglenn said:
> > > @HYUFD said:
> > >
> > > Thus the LDs won Richmond Park, Islington, Kensington and Esher and Walton last week.but the Brexit Party won Bolsover, Merthyr Tydfil, Wigan and Stoke
> > -------
> >
> > And the Tories want to chase after Bolsover, Merthyr Tydfil, Wigan and Stoke by becoming a copy of the Brexit Party... It's a zero seat strategy.
>
> Rubbish.
>
> The Brexit Party won Epping Forest, Buckingham, Wellingborough, Tonbridge, Thurrock, Mansfield, Ryedale etc even a few Tory Remain seats like Tunbridge Wells and Maidenhead last week (Stoke already has a Tory seat, also lost to Farage).
>
> In fact the Tories lost more areas to the Brexit Party than the LDs last week and far more of their voters
“Won” in an FPTnP assessment isn’t necessarily the best measure, though, is it? If you look your at Tunbridge Wells, the LibDems, Greens and CUK got 50% of the vote between them, with BXP back on 33% (35% with UKIP). And there’s another 5% for Labour and some remain splinter parties. In a GE the Tories are far more likely to lose that seat to a Remain candidate, probably a LibDem, than they are to BXP.
BXP is flattered by not having any competition, and the remain parties the opposite. In a GE the Tories would surely be more competitive than just 10% in Tunbridge Wells, and we can only hope local voters realise who the leading remain party is in their seat, or better that they get their act together and start cooperating.
> In a real election, there is prompting for minor parties (if TBP can still be called that), both during the campaign, and by reason of their presence on the ballot paper.
>
One issue with minor parties in general is that they're not on the ballot paper in every constituency. I think UKIP in 2015 were the first party out of the big three to do so.
I think that's the criterion I would use when deciding whether to prompt for a party. So I wouldn't prompt for the Green Party, and then I'd have to make a judgement on whether Farage's outfit would be likely to do so. On balance right now I would say yes - but that is a guess which might be wrong.
> > @Chris said:
> >
> > The question is whether there is any evidence that YouGov's overestimate was owing to prompting rather than any number of other possible shortcomings in their methods.
> >
> > You claimed before that "there is substantial historical evidence that prompting for minor/new parties results in overstating their support" and I asked you to quote the evidence for that. Please can you do so?
>
> Here you go:
>
> https://hbr.org/2016/11/how-polls-overestimate-support-for-third-party-candidates
That's an article about opinion polls in the USA and support for non-Democratic/Republican candidates.
You actually claimed that YouGov had said this - and I presumed at the time you meant they had said it about British elections.
Has YouGov actually said this or not?
> When do the Tory declarations need to be in and when will the first vote be?
>
> Hopefully not all 12 will still be in the hunt once polling commences.
The contest hasn't even begun yet. TM isn't resigning formally until 7th June. So presumably the deadline for nominations will be sometime the following week.
The moves against Raab from moderates are likely to also benefit Johnson and the environment secretary, Michael Gove, who Tory backbenchers believe is gaining momentum over his cabinet rival Jeremy Hunt.
https://www.theguardian.com/politics/2019/may/31/stop-dominic-raab-tory-moderates-seek-to-block-path-to-pm
Raab is a big help to the Boris campaign.
> > @Chris said:
>
> > In a real election, there is prompting for minor parties (if TBP can still be called that), both during the campaign, and by reason of their presence on the ballot paper.
> >
>
> One issue with minor parties in general is that they're not on the ballot paper in every constituency. I think UKIP in 2015 were the first party out of the big three to do so.
>
> I think that's the criterion I would use when deciding whether to prompt for a party. So I wouldn't prompt for the Green Party, and then I'd have to make a judgement on whether Farage's outfit would be likely to do so. On balance right now I would say yes - but that is a guess which might be wrong.
I'd suggest if you wanted to try to approach that scientifically, you would prompt for every party and apply a correction factor according to how many candidates there were.
This may also explain why both Labour and the Conservatives came at the bottom end of expectations this time: their get out the vote machinery was not working for them this time.
Polls are not trying to measure abstract party allegiance. They're trying to measure who will actually cast votes.
I'd stick them (And Brexit party) back into the cut off for first page parties, Plaid and SNP showing if the pollster is Welsh or Scottish.
Change UK, UKIP keep in the others.
>
> That's an article about opinion polls in the USA and support for non-Democratic/Republican candidates.
>
> You actually claimed that YouGov had said this - and I presumed at the time you meant they had said it about British elections.
>
> Has YouGov actually said this or not?
---------
Yes, it's in their statement today. I gave you a US source to show that it's a generally accepted phenomenon, not a conspiracy against the Brexit Party.
https://yougov.co.uk/topics/politics/articles-reports/2019/05/31/heres-how-we-prompt-brexit-party-and-why-its-more-
"In the past, prompting for smaller parties has tended to overstate their support when compared to actual elections, and the two-stage approach to prompting has produced more accurate results."
Suspect he put that in just to get another D.
> Tory moderates are beginning to coalesce around candidates they believe could defeat Dominic Raab, rather than Boris Johnson, as several MPs told the Guardian they considered the former Brexit secretary the bigger threat.
>
> The moves against Raab from moderates are likely to also benefit Johnson and the environment secretary, Michael Gove, who Tory backbenchers believe is gaining momentum over his cabinet rival Jeremy Hunt.
>
> https://www.theguardian.com/politics/2019/may/31/stop-dominic-raab-tory-moderates-seek-to-block-path-to-pm
>
> Raab is a big help to the Boris campaign.
Overton window in action.
> > @Stark_Dawning said:
> > > @HYUFD said:
> >
> > > https://twitter.com/Nigel_Farage/status/1134365557325799424
> >
> >
> >
> >
> >
> > Another tin-foil hat lunatic airs his populist conspiracy theories. Ignore.
> >
> > Surely Nigel doesn't actually believe that the 'establishment' is controlling YouGov's polling methodology? I think he just wants to instil his followers with a lot of anger, a lot of hate and a lot of victimhood. Expect much more of this.
>
> Or he's just been spending too much time around the Donald.
Farage is probably the best politician we've got at the moment, but I really don't want to see him heading down the path of this Cadwalladr/Adonis/Esler style nonsense. He should ignore the polls and I'm sure if he focuses on and wins Peterborough they might even take the lead with yougov.
> Is Corbyn toast ?
>
>
>
> He was safe when polling was up - it covered up the weaknesses.
>
>
>
> Could he be gone this year or will he cling on for the GE defeat ?
>
> No chance. His supporters will remain convinced that since he made up a huge polling deficit once, he can do it again.
>
> In some ways, he's probably happier starting from way behind. Certainly he'll do better if people are more relaxed about his likelihood of winning.
Aren't Labour ahead in the polling averages? I mean, down but still ahead. The way things are going, he's the likeliest PM post-election.
https://twitter.com/YouGov/status/1134396993802002433
https://yougov.co.uk/topics/politics/articles-reports/2019/05/31/heres-how-we-prompt-brexit-party-and-why-its-more-
> YouGov statement
>
> https://twitter.com/YouGov/status/1134396993802002433
>
> https://yougov.co.uk/topics/politics/articles-reports/2019/05/31/heres-how-we-prompt-brexit-party-and-why-its-more-
If they got Con, Lab, LD and Greens right but significantly overstated Brexit, who got understated ?
> > @Chris said:
> >
> > That's an article about opinion polls in the USA and support for non-Democratic/Republican candidates.
> >
> > You actually claimed that YouGov had said this - and I presumed at the time you meant they had said it about British elections.
> >
> > Has YouGov actually said this or not?
> ---------
>
> Yes, it's in their statement today. I gave you a US source to show that it's a generally accepted phenomenon, not a conspiracy against the Brexit Party.
>
> https://yougov.co.uk/topics/politics/articles-reports/2019/05/31/heres-how-we-prompt-brexit-party-and-why-its-more-
>
> "In the past, prompting for smaller parties has tended to overstate their support when compared to actual elections, and the two-stage approach to prompting has produced more accurate results."
Thank you for finally clarifying that.
I note that YouGov say they don't know why they overestimated the TBP share of the vote, and that the list of possible reasons they are investigating doesn't even include prompting - far from your claim that "it [prompting] resulted in them significantly overestimating their support."
So certainly YouGov is not trying to justify the lack of prompting in its Westminster polls on the basis of their got the European polls wrong.
> I am sure that YouGov, with a mere 19 years of experience and over 800 employees, who do nothing other than conduct polls and analyse them, would be foolish indeed to think they might know a tad more than @Chris about how to get the best results.
Of course you're right. Large commercial companies always know best. Even if they're talking complete bollocks.
> > @Endillion said:
> > Is Corbyn toast ?
> >
> >
> >
> > He was safe when polling was up - it covered up the weaknesses.
> >
> >
> >
> > Could he be gone this year or will he cling on for the GE defeat ?
> >
> > No chance. His supporters will remain convinced that since he made up a huge polling deficit once, he can do it again.
> >
> > In some ways, he's probably happier starting from way behind. Certainly he'll do better if people are more relaxed about his likelihood of winning.
>
> Aren't Labour ahead in the polling averages? I mean, down but still ahead. The way things are going, he's the likeliest PM post-election.
If there were a market for PM after GE, he should be a very weak favourite right now because we know
i) He is running a party probably competing in all 650 seats (Exclude all NI, SNP and Plaid)
ii) He is currently a party leader (Tories, Lib Dems are changing theirs)
iii) He will run for a seat (Farage might not)
iv) Seats below 200 are stickier for Labour on UNS than other parties on split vote shares.
v) He is more confidence and supply friendly to the Lib Dems & Greens for the right deal (Revocation/2nd referendum); SNP (2nd indy ref) ? than the Tories.
The Tories could be c&s friendly to the BXP or the other way round mind...
Against this we have:
i) His next PM ratings are absolubtely appalling. These are a good long term indicator that say he won't be PM.
ii) He's clearly uncomfortable leading a party with such a remainy membership when in his heart he is a leaver.
iii) Labour's anti-semitism issues.
iv) Labour's polling is heading south right now.
v) He could be challenged for the leadership and he is getting older so might pass it over voluntarily.
> i) His next PM ratings are absolubtely appalling. These are a good long term indicator that say he won't be PM.
> ii) He's clearly uncomfortable leading a party with such a remainy membership when in his heart he is a leaver.
> iii) Labour's anti-semitism issues.
> iv) Labour's polling is heading south right now.
> v) He could be challenged for the leadership and he is getting older so might pass it over voluntarily.
i) Personal polling is terrible. In ordinary circumstances, it's fatal but these aren't ordinary circumstances.
ii) Doesn't matter that he's uncomfortable, he's still leader
iii) Already factored into the polling. It ought to put more people off, I agree, but it isn't.
iv) Previous trend has been down. I'm really not sure whether past trends can ever be used to predict future trends. Could be like the stockmarket, pseudorandom.
v) Very good point. I heard rumours recently he was on the brink of leaving.
> opaltruce Great match for Indian bookies I reckon.
>
> "Muslim-Administered India"?
Well I hope not. Seems quite exciting. The lad who took 4 wickets took 5 against England didn't he?
Perhaps we'll see some proper cricket in this world cup rather than the smash around the all rounders rubbish.
As for the studies you mention (but do not cite) haven't we gone beyond that now? There is a tendency in modern society to quote studies at each other. Whilst a good approach it is not enough and one must examine the studies, extract the data and do one's own research. Cases where studies are cited far beyond their plausibility include the Laffer curve, devaluation stimulating the economy, and betting odds being a good predictor of elections.