politicalbetting.com » Blog Archive » Why I think that UKIP is being understated in many polls

System

politicalbetting.com » Blog Archive » Why I think that UKIP is being understated in many polls

The publication last week of the latest round of the Bown-funded constituency polling set off a debate about methodology with efforts to attack what Survation had done. The main objections were that the firm wasn’t using what have become standard approaches to ensure politically balanced samples.

Read the full story here

FluffyThoughts

Gabble....

dr_spyn

cue for Downfall spoof.

ShneurOdze

Very true, but only set to get worse as UKIP attracts new types of voters which will make it harder for pollsters to identify. Complicated with that UKIP always had late surges in the days/weeks preceding elections, making it almost impossible to have a reliable 'final poll'.

ojcorbs

It's the biggest issue in modern polling, and having prominent Ukippers on twitter accusing half the polling companies of intentional biases does not help. The reality is that UKIP has seen a 300-600% increase in polling vote share, an unheard of phenomenon in modern politics.
Pollsters aren't being biased, they are simply sticking to the methodology that has worked well in the past, and as Mike says in the post, a solution is not readily available.

The reality is we will probably have to wait until after the next election to get a firm idea of where UKIP's new base is. Once that correction has taken place, the current methodology will probably be pretty accurate.

saddened

This will be funny.

Sean_F

ShneurOdze said:

Very true, but only set to get worse as UKIP attracts new types of voters which will make it harder for pollsters to identify. Complicated with that UKIP always had late surges in the days/weeks preceding elections, making it almost impossible to have a reliable 'final poll'.

There was a bandwagon effect in the run-up to this year's local elections, and I expect we'll see one in 2014?

I don't think that weighting UKIP down from 208 to 21 can be correct.

isam

Mist be difficult for all pollsters as firstly, as Mike says their formulas are based on a three party system, and secondly, it seems many ukip voters are previous non voters.... The fact we have a coalition complicates matters further...

My prediction is there will be more rickets from bookmakers than ever before, but whether we can identify which are the rickets and who is ahead of the game is another matter...

SMukesh

I bet the right-wing rags didn`t expect this UKIP surge threatening the Tories when they artificially boosted them through the Rotherham adoption controversy.

tpfkar

FTP: No less off-topic here than it was there

Happy Christmas from a disappointingly snow-less Bavaria. By the looks of reports today we were lucky to get out of the UK without problems yesterday, must be worrying for those trying to leave today.

On-topic, this 'row' looks like it's all scripted and agreed in advance. It's no secret to anyone that the Lib Dems have a different perspective to the Tories on immigration and have no time for the Tories playing at who can talk nastiest. Both parties gain by asserting where they stand and particularly as Vince is one of the few people the 28% or so LD->LAB switchers will still listen to. He reaches the parts that Nick Clegg can't.

For me personally, this is probably one of the reasons I've ended up in the Lib Dems, as no-one else seems to be prepared to listen to the evidence of the net benefits of immigration. I find it almost comical how everyone else tries to talk as tough as they can at home, and then as soon as a minister goes to India, or China, or Brazil, it's always a pitch to come to Britain and relaxing visa processes. Sooner or later those inconsistencies will be highlighted but it seems to be an unwritten rule that they should be kept as quiet as possible at the moment. One of the clearest cases I've seen of where the political and economic angles collide head on.

I wonder if it's a generational thing - my generation aren't as angry about immigration as some that have come before. Sure, I've an internationalist viewpoint in particular as I've lived in Poland and my wife is German, but like stodge there are people who are turned off by all the anti-immigration talk and they should be easy pickings for the Lib Dems. They won't touch the Tories or UKIP in their current modes, so I imagine David Cameron has more pressing matters to worry about than Vince Cable.

isam

Just in from shopping/job interview in London and can honestly say its the most horrible weather I have can remember being out in... Three nice incidents on the tube/at railway stations of young people helping carry prams onto trains / old people's luggage up the stairs at Barking and Upminster

richardDodd

Having just watched a spokesman for the Muslim Council on Sky, justifying asking customers to go to another till in M&S....I think UKIP will get a few more members

AveryLP

FluffyThoughts said:

Gabble....

Gobbled.

RichardNabavi

You have to make a distinction between on-line polls, and telephone polls. The former are very prone to 'enthusiasm' bias: as we see in these august pages, the Kippers are terribly excited; that makes them much more likely to respond to a poll if they are members of a panel which is sent an invitation to participate in a political poll. Therefore it is almost certainly correct to weight them down in such polls, possibly by quite a lot. The unknown, of course, is by how much.

Considering just the on-line polls for the moment, we can see a huge disagreement between Survation, ComRes and Opinium on the one hand, and Populus on the other. The former routinely report UKIP figures in the 16-19% range, the latter in the 7-8% range. YouGov seems to be somewhere in the middle.

Normally we might be able to form a view on which is more likely to be correct by looking at the telephone polls, where the typical UKIP figure is around 10% or 11%, but they in turn might be subject to error because of the past-vote weighting.

I suspect that the telephone polls are a better guide, but we shall only really know the answer when the election comes and then only if, in the polls leading up to the election, the discrepancy persists.

Quincel

ojcorbs said:

The reality is we will probably have to wait until after the next election to get a firm idea of where UKIP's new base is. Once that correction has taken place, the current methodology will probably be pretty accurate.

Potentially some pollsters could try out different techniques for weighting and prompting in Euro 2014 and see which work best. Obviously it isn't a perfect test for a GE, but it should be reasonably close and allow some good data for them.

AveryLP

RichardNabavi said:

You have to make a distinction between on-line polls, and telephone polls. The former are very prone to 'enthusiasm' bias: as we see in these august pages, the Kippers are terribly excited; that makes them much more likely to respond to a poll if they are members of a panel which is sent an invitation to participate in a political poll. Therefore it is almost certainly correct to weight them down in such polls, possibly by quite a lot. The unknown, of course, is by how much.

Considering just the on-line polls for the moment, we can see a huge disagreement between Survation, ComRes and Opinium on the one hand, and Populus on the other. The former routinely report UKIP figures in the 16-19% range, the latter in the 7-8% range. YouGov seems to be somewhere in the middle.

Normally we might be able to form a view on which is more likely to be correct by looking at the telephone polls, where the typical UKIP figure is around 10% or 11%, but they in turn might be subject to error because of the past-vote weighting.

I suspect that the telephone polls are a better guide, but we shall only really know the answer when the election comes and then only if, in the polls leading up to the election, the discrepancy persists.

Richard

UKIP supporters can't remember where they placed their spectacles five minutes ago.

Can pollsters really expect them to remember how they voted in May 2010?

OGH's big red marker pen should be drawing an arrow which points at the 208/10% figures not the 21/1% ones.

Sean_F

SMukesh said:

I bet the right-wing rags didn`t expect this UKIP surge threatening the Tories when they artificially boosted them through the Rotherham adoption controversy.

UKIP's impact is paradoxical. They take the biggest single element of their support from ex-conservatives. But, they pull in a fair amount of support from ex-Lib Dems, Labour and non-voters. And, they stop swing voters from going Labour in mid-term.

saddened

SMukesh said:

I bet the right-wing rags didn`t expect this UKIP surge threatening the Tories when they artificially boosted them through the Rotherham adoption controversy.

I'm in no way a supporter of the kippers, but what happened in Rotherham was a disgrace. The only thing that could have made it worse would have been for the press to ignore it. It's sad that you seem to view it from a purely political advantage perspective.

RichardNabavi

AveryLP said:

Richard

UKIP supporters can't remember where they placed their spectacles five minutes ago.

Can pollsters really expect them to remember how they voted in May 2010?

It's important to realise that YouGov, at least, doesn't depend on them being able to remember how they voted: they have the party identification and past-vote data on file already (although AFAIK they use only the former). The fact that YouGov's figures are broadly comparable with ICM and ComRes (telephone) polls is one reason why I tend to think that the 10% to 12% figure is the most trustworthy. In other words, Populus are probably being over-enthusiastic in their weighting, whereas ComRes (online), Survation and Opinium are not weighting down by enough. But it's all largely informed guesswork.

JohnO

Sean_F said:

SMukesh said:

I bet the right-wing rags didn`t expect this UKIP surge threatening the Tories when they artificially boosted them through the Rotherham adoption controversy.

UKIP's impact is paradoxical. They take the biggest single element of their support from ex-conservatives. But, they pull in a fair amount of support from ex-Lib Dems, Labour and non-voters. And, they stop swing voters from going Labour in mid-term.

I'm sure that's correct. By the way, are you still planning to vote Conservative in May 2015 if you are still living in a Con-Lab marginal?

anotherDave

ojcorbs said:

The reality is we will probably have to wait until after the next election to get a firm idea of where UKIP's new base is. Once that correction has taken place, the current methodology will probably be pretty accurate.

The margin of error in a four-way vote split will be too high to describe as "pretty accurate".

corporeal

Sean_F said:

SMukesh said:

I bet the right-wing rags didn`t expect this UKIP surge threatening the Tories when they artificially boosted them through the Rotherham adoption controversy.

UKIP's impact is paradoxical. They take the biggest single element of their support from ex-conservatives. But, they pull in a fair amount of support from ex-Lib Dems, Labour and non-voters. And, they stop swing voters from going Labour in mid-term.

Do they actually? Iirc last ICM poll had UKIP pulling in pretty low numbers from those places, particularly Labour.

richardDodd said:

Having just watched a spokesman for the Muslim Council on Sky, justifying asking customers to go to another till in M&S....I think UKIP will get a few more members

*shrugs* M&S employ a member of staff. If M&S and staff member have come to an arrangement where they don't have to sell alcohol then that's up to M&S.

It's certainly unheard of, though usually relating to staff member being under 18 and hence unable to sell alcohol.

CarlottaVance

SMukesh said:

I bet the right-wing rags didn`t expect this UKIP surge threatening the Tories when they artificially boosted them through the Rotherham adoption controversy.

I think there was more to Rotherham than 'adoption'....and it was shocking the way the right left wing rags "artificially boosted" it (sic):

Police probe at least 54 more evil child sex grooming gangs

http://www.mirror.co.uk/news/uk-news/police-probe-least-54-more-1896991#ixzz2oJuYlXpX

Garethofthevale

Mike, as a market researcher (although in healthcare), I believe there is some possibility that the next election could turn out like 92 did for the polling industry. As you rightly point out, the problem is measuring the UKIP share. UKIP seems to be picking up votes from other parties and also non-voters. The key question is how 'sticky' these votes are going to be in 2015? We don't know.

Another problem is that we have a Coalition government for the first time since the start of most company's polling archives. Now in the past it has sometimes happened that voters move back to the Government closer to the election. Could this happen with LD-Lab switchers moving back to LD? Again we don't know.

Coming back to the UKIP problem. The issue for researchers is how we make sure our sample is representative. If we were doing a survey about say a new car brand, then we could weight by region, age, gender and social class and that should be sufficient. The problem is you can have a sample of 40-50 year old C1 men in the W Midlands who vote Labour and an identical sample who demographically look the same but vote Tory. So you then end up using past vote weighting. The problem here is as you identify is that people misremember - and that becomes more likely as we are now 4 years down the line and there have been several other elections in-between. Another problem you can get is deliberate dishonesty - for example 2% voted for the BNP at the last election but not all may admit it.

One of the things I would like to see used in political polling is consideration. Something we use in other types of polling is the Juster scale, which has been shown to have good results in understanding how likely someone is to do something. For example I would ask this:

Thinking about the next General Election in 2015, how likely are you to vote for [insert party]?

10 Certain, practically certain (99 in 100)
9 Almost sure (9 in 10)
8 Very probable (8 in 10)
7 Probable (7 in 10)
6 Good possibility (6 in 10)
5 Fairly good possibility (5 in 10)
4 Fair possibility (4 in 10)
3 Some possibility (3 in 10)
2 Slight possibility (2 in 10)
1 Very slight possibility (1 in 10)
0 No chance, almost no chance (1 in 100)

You would need to put some validation in e.g. to make sure that someone couldn't say they were certain to vote for Cons and certain to vote for Lab.

What you would expect to find is that there is a large group who are certain or very likely to vote for a certain party. There is another group who are true floating voters and will consider anyone. And then there will be groups who will only consider 2 or 3 parties e.g. Lab/LD.

Garethofthevale

(continued) What I would then follow up with if someone is considering say Lab/LD is:

What could the Labour party do to make it more likely you would vote for them in 2015?
What could the Liberal Democrat party do to make it more likely you would vote for them in 2015?

I think this would be fascinating. We have the issues data from Ipsos but this includes all the voters who are 'nailed on'. It would be great to find out what the key swing voter groups really care about.

Used in conjunction with standard polling data, consideration data would hopefully allow us to find out how 'sticky' UKIP and Labour support really is.

corporeal

ojcorbs said:

It's the biggest issue in modern polling, and having prominent Ukippers on twitter accusing half the polling companies of intentional biases does not help. The reality is that UKIP has seen a 300-600% increase in polling vote share, an unheard of phenomenon in modern politics.
Pollsters aren't being biased, they are simply sticking to the methodology that has worked well in the past, and as Mike says in the post, a solution is not readily available.

The reality is we will probably have to wait until after the next election to get a firm idea of where UKIP's new base is. Once that correction has taken place, the current methodology will probably be pretty accurate.

On a pedantic note, not quite unheard of I think. The alliance vote in the 80s famously polled over 50% at one point (insert your own caveats), and the referendum party in the run up to '97 experienced a pretty meteoric rise (0 to ~3% iirc)

AlastairMeeks

Empty vessels make most noise. It doesn't mean that there are more empty vessels.

Grandiose

These are people who mostly haven't voted UKIP in Westminster, don't have experience of the full UKIP manifesto, mostly switched to UKIP (at Westminster) two years or so ago, without a European election since; and have the much easier protest opinion than putting an "X" in the box. I'm not surprised their vote is down weighted. The fact is that all pollsters do a "how people said they were going to vote" to "how we think they'll actually vote" adjustment, which might seem odd, but is entirely valid.

isam

Another thread of smug, cheap (unfunny?) smart arse shots at ukip and suspension of belief that people who disagree with Lib Dems, Labour and Conservatives really exist

@antifrank you can have another £100@4/6 LD /UKIP 2010 vote percentage if you so desire

david_herdson

ojcorbs said:

...
Pollsters aren't being biased, they are simply sticking to the methodology that has worked well in the past, and as Mike says in the post, a solution is not readily available.

I'm sure they're not being intentionally biased in the sense of trying to advantage one side or another, but I've no doubt personally that polls that shift raw figures to weighted ones in the way the one Mike quotes does are biased against large-scale change. Local and by-election results constantly point to UKIP doing *at least* as well as the most optimistic polls suggest; the only caveat being that those elections have smaller turnouts than a general election and UKIP's support seems to be disproportionately enthusiastic (i.e. an increase in turnout from 35% to 60% would see UKIP's share drop).

The reality is we will probably have to wait until after the next election to get a firm idea of where UKIP's new base is. Once that correction has taken place, the current methodology will probably be pretty accurate.

The problem is that the current parliament may well be a unique occurrence, or at least very significantly different from the next one. If there's no coalition, the UKIP's place as the NOTA party becomes undermined. If there's no coalition, all parties can stay closer to their base. 2015 will be an election held in unusual circumstances and there's no reason to assume that the results will provide any more reliable baseline for 2015-20 than 2010 did for 2010-5.

AlastairMeeks

@isam Yes, I'll happily have that. Thank you.

EDIT I'll even let you do it on 2015 vote percentages rather than 2010.

TOPPING

With the exception of very few, and I include many of the Kippers on PB, a tiny tiny amount of UKIP polling support is much more than NOTA/hell-in-a-handbasket/please someone make it all ok.

It's impossible to categorise or count UKIP voters because as has been pointed out, UKIP is such a recent phenomenon, and has occurred in such untypical circumstances (country really, really f&cked, NOTA party now in Govt...) that it is without precedent. So these are not polls of UKIP voters, they are polls of protest. of single issues, of NOTA.

Come GE2015 most UKIP "support" will dissolve into one of the three grown-up parties.

Sean_F

JohnO said:

Sean_F said:

SMukesh said:

I bet the right-wing rags didn`t expect this UKIP surge threatening the Tories when they artificially boosted them through the Rotherham adoption controversy.

UKIP's impact is paradoxical. They take the biggest single element of their support from ex-conservatives. But, they pull in a fair amount of support from ex-Lib Dems, Labour and non-voters. And, they stop swing voters from going Labour in mid-term.

I'm sure that's correct. By the way, are you still planning to vote Conservative in May 2015 if you are still living in a Con-Lab marginal?

Yes, I would, unless the Conservative MP was someone I detested, or if it was a seat where UKIP was surging.

Sean_F

corporeal said:

Sean_F said:

SMukesh said:

I bet the right-wing rags didn`t expect this UKIP surge threatening the Tories when they artificially boosted them through the Rotherham adoption controversy.

UKIP's impact is paradoxical. They take the biggest single element of their support from ex-conservatives. But, they pull in a fair amount of support from ex-Lib Dems, Labour and non-voters. And, they stop swing voters from going Labour in mid-term.

Do they actually? Iirc last ICM poll had UKIP pulling in pretty low numbers from those places, particularly Labour.

richardDodd said:

Having just watched a spokesman for the Muslim Council on Sky, justifying asking customers to go to another till in M&S....I think UKIP will get a few more members

*shrugs* M&S employ a member of staff. If M&S and staff member have come to an arrangement where they don't have to sell alcohol then that's up to M&S.

It's certainly unheard of, though usually relating to staff member being under 18 and hence unable to sell alcohol.

Yougov regularly show 10%+ of Lib Dems from 2010 going UKIP. Down in the West Country, there are loads of Lib Dem - UKIP switchers.

MarkSenior

Whilst Mike is making a valid point re the harsh weightings used by Populus in particular he does not cover the main paradox in current polling , that even before weighting telephone pollsters find many fewer UKIP supporters than the online pollsters . This is true even for Comres whose online and telephone polls diverge by 6 to 8% for UKIP support .
If you can find the reason for this discrepancy it may give some clue as to what the true level of UKIP support is .

nigel4england

TOPPING said:

With the exception of very few, and I include many of the Kippers on PB, a tiny tiny amount of UKIP polling support is much more than NOTA/hell-in-a-handbasket/please someone make it all ok.

It's impossible to categorise or count UKIP voters because as has been pointed out, UKIP is such a recent phenomenon, and has occurred in such untypical circumstances (country really, really f&cked, NOTA party now in Govt...) that it is without precedent. So these are not polls of UKIP voters, they are polls of protest. of single issues, of NOTA.

Come GE2015 most UKIP "support" will dissolve into one of the three grown-up parties.

Wishful thinking, if the main parties and Crosby in particular share your view then you are in for a big surprise. It's the sort of arrogance that has helped to put UKIP where they are.

MikeK

TOPPING said:

With the exception of very few, and I include many of the Kippers on PB, a tiny tiny amount of UKIP polling support is much more than NOTA/hell-in-a-handbasket/please someone make it all ok.

It's impossible to categorise or count UKIP voters because as has been pointed out, UKIP is such a recent phenomenon, and has occurred in such untypical circumstances (country really, really f&cked, NOTA party now in Govt...) that it is without precedent. So these are not polls of UKIP voters, they are polls of protest. of single issues, of NOTA.

Come GE2015 most UKIP "support" will dissolve into one of the three grown-up parties.

You keep saying that UKIP will dissolve/disappear Topping and repeatedly declaiming it won't make it happen however much you want it. You should send your little prayer to Santa, Topping, that will make his Christmas- and mine.

anotherDave

corporeal said:

Do they actually? Iirc last ICM poll had UKIP pulling in pretty low numbers from those places, particularly Labour.

Labour only won 29% of the vote in 2010. So its unlikely that many 2010 Labour voters would be swing voters.

Mark Pack had a good piece showing how UKIP had reduced Labour's 2013 numbers by attracting LD/Con/Other swing voters that might have chosen Labour.

http://www.markpack.org.uk/47012/how-ukip-is-damaging-labour-reprised/

Labour's local election results:
2010: 27%
2011: 37%
2012: 38%
2013: 29% (UKIP: 23%)

stodge

Evening all

As far as UKIP is concerned, I've never had an issue with the party or its supporters. They are sincere in their views and it's understandable given what's happened in recent years to hold the political process in contempt.

That said, it's the process we have. As far as policy is concerned, I've not yet found an issue (apart from electoral reform) where I agree with UKIP at all. The experience of the LDs should be informative - IF UKIP is seen as any kind of threat, they will be placed under an intensive and instrusive media onslaught during or just before the campaign.

Inconsistencies and contradictory statements will be picked up and embellished so the 2015 UKIP manifesto will need to be bombproof in terms of commitments and policies. Inevitably, the party will be asked, if it found itself in a position of influence in the next Parliament, what it would do and more importantly what kind of Government it would support.

MikeK

@AveryLP
"UKIP supporters can't remember where they placed their spectacles five minutes ago."
-----------------
Come, come, AveryLP. I assume that you're a bit more intelligent than that. The fact that you mock UKIP supporters in such a way, shows how s**t scared you, and a few others on PB, are of UKIPs 2013 surge and advance.

MarkSenior

nigel4england said:

TOPPING said:

With the exception of very few, and I include many of the Kippers on PB, a tiny tiny amount of UKIP polling support is much more than NOTA/hell-in-a-handbasket/please someone make it all ok.

It's impossible to categorise or count UKIP voters because as has been pointed out, UKIP is such a recent phenomenon, and has occurred in such untypical circumstances (country really, really f&cked, NOTA party now in Govt...) that it is without precedent. So these are not polls of UKIP voters, they are polls of protest. of single issues, of NOTA.

Come GE2015 most UKIP "support" will dissolve into one of the three grown-up parties.

Wishful thinking, if the main parties and Crosby in particular share your view then you are in for a big surprise. It's the sort of arrogance that has helped to put UKIP where they are.

And where are UKIP today ? Unable to win council by elections when turnouts are typically 15 to 25% and UKIP supporters are the keenest to turn out and cast a protest vote of any party . 25% vote share when turnout is just 20% would reflect around 6 to 8% on a GE turnout .

Charles

This just repeats the standard question though - and it is misleading just to think of it as UKIP being understated.

The two big shifts in the chart above are:

UKIP: 208 (10%) ---> 21 (1%)
DNV: 262 (13%) ---> 452 (22%)

The implication - as we all know - is that UKIP has attracted a lot of support from DNV.

Will they vote in 2015? I am sceptical, but may be they will.

The key question - which @NickPalmer tries to address from time to time, tbf - is whether there is any reason whether UKIP will manage to persuade DNV to change their voting habits.

Flockers_pb

I am not convinced it matters greatly if Ukip are being understated in the polls at the moment, or whether their true level of current support is 10% or 16%. The key question is whether the polls today are a good guide for the outcome of the election in 2015. We each have our own answer to that and they will be borne largely out of a combination of prejudice, hope and selective references to an insignificant number of precedents and our dealings with a limited, again statistically insignificant number of people.

I believe that the polls are currently a poor guide to Ukip's 2015 outturn, which I expect to be below the bottom of their present polling range and worse than their current local election performances. I base that belief on Ukip's manifest lack of preparation and personnel to govern, which I trust the British people to identify, the dissipation of the protest vote as the election nears, the strengthening economy bringing some ex-Tories back to the fold, a strengthening of Liberal Democrat support as they break free from coalition and the likely return to loyalty of the Tory papers, amongst other factors.

If I wanted to support the alternative proposition I would point to the leader ratings demonstrating broad contempt for the leaders of the three main parties, Cameron's inability to connect to the Tory base and the appaling weakness of the Labour front bench providing fuel for "fourth party" support and the Euro boost we all expect Ukip to get next year providing momentum into the election campaign, as reasons to do so.

All are valid points, but the weight one attaches to them will vary according to one's perception and prejudices. The central point though, is even if the polls uniformly agreed that if the election was held tomorrow 16% of the electorate would vote UKIP, and even if that was a correct reflection of the electorate's current intention, it tells us very little about what will happen in 2015, in the same way that Tory scores in the mid-40s in 2009 or Labour scores in the high 50s/low 60s at various points from 1994 onwards told us little about what would happen at the ballot box. About as far as one can go is to note that the direction of travel is favourable to Ukip and they are likely to poll better in 2015 than they did in 2010.

All of this assumes, of course, no black swans. Dangerous assumption.

anotherDave

MarkSenior said:

25% vote share when turnout is just 20% would reflect around 6 to 8% on a GE turnout .

So you're expecting to see LDs 2013 14% reduced to 4% in the GE?

nigel4england

MarkSenior said:

nigel4england said:

TOPPING said:

With the exception of very few, and I include many of the Kippers on PB, a tiny tiny amount of UKIP polling support is much more than NOTA/hell-in-a-handbasket/please someone make it all ok.

It's impossible to categorise or count UKIP voters because as has been pointed out, UKIP is such a recent phenomenon, and has occurred in such untypical circumstances (country really, really f&cked, NOTA party now in Govt...) that it is without precedent. So these are not polls of UKIP voters, they are polls of protest. of single issues, of NOTA.

Come GE2015 most UKIP "support" will dissolve into one of the three grown-up parties.

Wishful thinking, if the main parties and Crosby in particular share your view then you are in for a big surprise. It's the sort of arrogance that has helped to put UKIP where they are.

And where are UKIP today ? Unable to win council by elections when turnouts are typically 15 to 25% and UKIP supporters are the keenest to turn out and cast a protest vote of any party . 25% vote share when turnout is just 20% would reflect around 6 to 8% on a GE turnout .

6% behind Labour according to the figures below.

Sean_F

MikeK said:

TOPPING said:

With the exception of very few, and I include many of the Kippers on PB, a tiny tiny amount of UKIP polling support is much more than NOTA/hell-in-a-handbasket/please someone make it all ok.

It's impossible to categorise or count UKIP voters because as has been pointed out, UKIP is such a recent phenomenon, and has occurred in such untypical circumstances (country really, really f&cked, NOTA party now in Govt...) that it is without precedent. So these are not polls of UKIP voters, they are polls of protest. of single issues, of NOTA.

Come GE2015 most UKIP "support" will dissolve into one of the three grown-up parties.

You keep saying that UKIP will dissolve/disappear Topping and repeatedly declaiming it won't make it happen however much you want it. You should send your little prayer to Santa, Topping, that will make his Christmas- and mine.

I daresay that the Liberals once thought that irritating little Labour Representation Committee would disappear.

MarkSenior

anotherDave said:

MarkSenior said:

25% vote share when turnout is just 20% would reflect around 6 to 8% on a GE turnout .

So you're expecting to see LDs 2013 14% reduced to 4% in the GE?

Nope , because the Lib Dems are in government and therefore subject tp protest voting against them in mid term . I still expect Lib Dems to poll 16 to 18% in 2015 .

AlastairMeeks

@Flockers_pb That's pretty much my own view (though I expect the Lib Dems to see the coalition out). I would also add that none of the three main parties have yet made their retail offers to the public for their votes. We can expect some eye-catching initiatives from all three, many of which will be calculated to appeal to the battlers who are currently UKIP-identifiers.

The current UKIP identifiers are going to be critical for the next election. Will they remain loyal to UKIP? Will they melt away? If so, how will they break between the other parties and non-voters?

MikeSmithson

Garethofthevale said:

Mike, as a market researcher (although in healthcare), I believe there is some possibility that the next election could turn out like 92 did for the polling industry. As you rightly point out, the problem is measuring the UKIP share. UKIP seems to be picking up votes from other parties and also non-voters. The key question is how 'sticky' these votes are going to be in 2015? We don't know.

Another problem is that we have a Coalition government for the first time since the start of most company's polling archives. Now in the past it has sometimes happened that voters move back to the Government closer to the election. Could this happen with LD-Lab switchers moving back to LD? Again we don't know.

Coming back to the UKIP problem. The issue for researchers is how we make sure our sample is representative. If we were doing a survey about say a new car brand, then we could weight by region, age, gender and social class and that should be sufficient. The problem is you can have a sample of 40-50 year old C1 men in the W Midlands who vote Labour and an identical sample who demographically look the same but vote Tory. So you then end up using past vote weighting. The problem here is as you identify is that people misremember - and that becomes more likely as we are now 4 years down the line and there have been several other elections in-between. Another problem you can get is deliberate dishonesty - for example 2% voted for the BNP at the last election but not all may admit it.

One of the things I would like to see used in political polling is consideration. Something we use in other types of polling is the Juster scale, which has been shown to have good results in understanding how likely someone is to do something. For example I would ask this:

Thinking about the next General Election in 2015, how likely are you to vote for [insert party]?

10 Certain, practically certain (99 in 100)
9 Almost sure (9 in 10)
8 Very probable (8 in 10)
7 Probable (7 in 10)
6 Good possibility (6 in 10)
5 Fairly good possibility (5 in 10)
4 Fair possibility (4 in 10)
3 Some possibility (3 in 10)
2 Slight possibility (2 in 10)
1 Very slight possibility (1 in 10)
0 No chance, almost no chance (1 in 100)

You would need to put some validation in e.g. to make sure that someone couldn't say they were certain to vote for Cons and certain to vote for Lab.

What you would expect to find is that there is a large group who are certain or very likely to vote for a certain party. There is another group who are true floating voters and will consider anyone. And then there will be groups who will only consider 2 or 3 parties e.g. Lab/LD.

Thanks. That is an interesting idea that I will pass on.

At the moment, I don't have Lord Ashcroft's wealth and don't have the resources to fund 20k sample phone polls.

nigel4england

AveryLP said:

RichardNabavi said:

You have to make a distinction between on-line polls, and telephone polls. The former are very prone to 'enthusiasm' bias: as we see in these august pages, the Kippers are terribly excited; that makes them much more likely to respond to a poll if they are members of a panel which is sent an invitation to participate in a political poll. Therefore it is almost certainly correct to weight them down in such polls, possibly by quite a lot. The unknown, of course, is by how much.

Considering just the on-line polls for the moment, we can see a huge disagreement between Survation, ComRes and Opinium on the one hand, and Populus on the other. The former routinely report UKIP figures in the 16-19% range, the latter in the 7-8% range. YouGov seems to be somewhere in the middle.

Normally we might be able to form a view on which is more likely to be correct by looking at the telephone polls, where the typical UKIP figure is around 10% or 11%, but they in turn might be subject to error because of the past-vote weighting.

I suspect that the telephone polls are a better guide, but we shall only really know the answer when the election comes and then only if, in the polls leading up to the election, the discrepancy persists.

Richard

UKIP supporters can't remember where they placed their spectacles five minutes ago.

Can pollsters really expect them to remember how they voted in May 2010?

OGH's big red marker pen should be drawing an arrow which points at the 208/10% figures not the 21/1% ones.

Wonderful stuff Avery, I will remember these cheap insults from you and your shambles of a party when you are whinging that it was UKIP that stopped you getting a majority in 2015.

Sean_F

Charles said:

This just repeats the standard question though - and it is misleading just to think of it as UKIP being understated.

The two big shifts in the chart above are:

UKIP: 208 (10%) ---> 21 (1%)
DNV: 262 (13%) ---> 452 (22%)

The implication - as we all know - is that UKIP has attracted a lot of support from DNV.

Will they vote in 2015? I am sceptical, but may be they will.

The key question - which @NickPalmer tries to address from time to time, tbf - is whether there is any reason whether UKIP will manage to persuade DNV to change their voting habits.

There are two groups. Won't Vote. And Given Up Voting. 70-80% voted from 1979 to 1997. That dropped to 57% in 2001, and has only recovered to 65%.

UKIP won't get much support among the former. They will, among the latter.

MarkSenior

nigel4england said:

MarkSenior said:

nigel4england said:

TOPPING said:

With the exception of very few, and I include many of the Kippers on PB, a tiny tiny amount of UKIP polling support is much more than NOTA/hell-in-a-handbasket/please someone make it all ok.

It's impossible to categorise or count UKIP voters because as has been pointed out, UKIP is such a recent phenomenon, and has occurred in such untypical circumstances (country really, really f&cked, NOTA party now in Govt...) that it is without precedent. So these are not polls of UKIP voters, they are polls of protest. of single issues, of NOTA.

Come GE2015 most UKIP "support" will dissolve into one of the three grown-up parties.

Wishful thinking, if the main parties and Crosby in particular share your view then you are in for a big surprise. It's the sort of arrogance that has helped to put UKIP where they are.

And where are UKIP today ? Unable to win council by elections when turnouts are typically 15 to 25% and UKIP supporters are the keenest to turn out and cast a protest vote of any party . 25% vote share when turnout is just 20% would reflect around 6 to 8% on a GE turnout .

6% behind Labour according to the figures below.

In mid term when UKIP supporters are keenest to turn out . As I have posted before look at the performance of the SDP and Liberals in 1985 the equivalent year in the 1983 to 1987 parliament . They won more council by elections than either Labour or the Conservatives but it did not presage any kind of breakthrough in the 1987 GE .

nigel4england

MarkSenior said:

nigel4england said:

MarkSenior said:

nigel4england said:

TOPPING said:

With the exception of very few, and I include many of the Kippers on PB, a tiny tiny amount of UKIP polling support is much more than NOTA/hell-in-a-handbasket/please someone make it all ok.

It's impossible to categorise or count UKIP voters because as has been pointed out, UKIP is such a recent phenomenon, and has occurred in such untypical circumstances (country really, really f&cked, NOTA party now in Govt...) that it is without precedent. So these are not polls of UKIP voters, they are polls of protest. of single issues, of NOTA.

Come GE2015 most UKIP "support" will dissolve into one of the three grown-up parties.

Wishful thinking, if the main parties and Crosby in particular share your view then you are in for a big surprise. It's the sort of arrogance that has helped to put UKIP where they are.

And where are UKIP today ? Unable to win council by elections when turnouts are typically 15 to 25% and UKIP supporters are the keenest to turn out and cast a protest vote of any party . 25% vote share when turnout is just 20% would reflect around 6 to 8% on a GE turnout .

6% behind Labour according to the figures below.

In mid term when UKIP supporters are keenest to turn out . As I have posted before look at the performance of the SDP and Liberals in 1985 the equivalent year in the 1983 to 1987 parliament . They won more council by elections than either Labour or the Conservatives but it did not presage any kind of breakthrough in the 1987 GE .

Have you taken isam up on his bet?

AndyJS

Rowena Holland stands down as Conservative candidate for Nottingham South:

http://www.youtube.com/profile_redirector/111099468426127108930

MarkSenior

MikeSmithson said:

Garethofthevale said:

Mike, as a market researcher (although in healthcare), I believe there is some possibility that the next election could turn out like 92 did for the polling industry. As you rightly point out, the problem is measuring the UKIP share. UKIP seems to be picking up votes from other parties and also non-voters. The key question is how 'sticky' these votes are going to be in 2015? We don't know.


One of the things I would like to see used in political polling is consideration. Something we use in other types of polling is the Juster scale, which has been shown to have good results in understanding how likely someone is to do something. For example I would ask this:

Thinking about the next General Election in 2015, how likely are you to vote for [insert party]?

10 Certain, practically certain (99 in 100)
9 Almost sure (9 in 10)
8 Very probable (8 in 10)
7 Probable (7 in 10)
6 Good possibility (6 in 10)
5 Fairly good possibility (5 in 10)
4 Fair possibility (4 in 10)
3 Some possibility (3 in 10)
2 Slight possibility (2 in 10)
1 Very slight possibility (1 in 10)
0 No chance, almost no chance (1 in 100)

You would need to put some validation in e.g. to make sure that someone couldn't say they were certain to vote for Cons and certain to vote for Lab.

What you would expect to find is that there is a large group who are certain or very likely to vote for a certain party. There is another group who are true floating voters and will consider anyone. And then there will be groups who will only consider 2 or 3 parties e.g. Lab/LD.

Thanks. That is an interesting idea that I will pass on.

At the moment, I don't have Lord Ashcroft's wealth and don't have the resources to fund 20k sample phone polls.

The Survation poll they conducted for the Lib Dems published in September on their website covers some of these points .

Flockers_pb

antifrank said:

@Flockers_pb That's pretty much my own view (though I expect the Lib Dems to see the coalition out). I would also add that none of the three main parties have yet made their retail offers to the public for their votes. We can expect some eye-catching initiatives from all three, many of which will be calculated to appeal to the battlers who are currently UKIP-identifiers.

The current UKIP identifiers are going to be critical for the next election. Will they remain loyal to UKIP? Will they melt away? If so, how will they break between the other parties and non-voters?

My guess would be about 25-33% of previous non-voters and previous Conservative or Liberal Democrat voters who are currently saying they will vote Ukip will do so, and about 33-50% of previous Labour voters who are currently saying they will vote Ukip will do so. There's no scientific basis for that, just a guess based on observation over the years of the fragility of minority party support.

The problem that the main parties have in trying to appeal to Ukip supporters is that it is quite hard to identify who they are and what they support. They are a supposedly libertarian party with considerable support amongst social conservatives. Their anti-immigration message appeals to the Tory blue-rinser brigade and Labour's D/E heartlands (neither of which identifies with the other). They can't be out-Ukipped on Europe (don't even try Dave!). Polling data indicates them to be reliably negative and cynical, so it is unlikely good news will do the trick. I think that leaves the Tories flogging welfare reform and jobs, with a side order of Euro-bashing and Labour flogging cost of living, each to no real effect.

surbiton

MarkSenior said:

nigel4england said:

TOPPING said:

With the exception of very few, and I include many of the Kippers on PB, a tiny tiny amount of UKIP polling support is much more than NOTA/hell-in-a-handbasket/please someone make it all ok.

It's impossible to categorise or count UKIP voters because as has been pointed out, UKIP is such a recent phenomenon, and has occurred in such untypical circumstances (country really, really f&cked, NOTA party now in Govt...) that it is without precedent. So these are not polls of UKIP voters, they are polls of protest. of single issues, of NOTA.

Come GE2015 most UKIP "support" will dissolve into one of the three grown-up parties.

Wishful thinking, if the main parties and Crosby in particular share your view then you are in for a big surprise. It's the sort of arrogance that has helped to put UKIP where they are.

And where are UKIP today ? Unable to win council by elections when turnouts are typically 15 to 25% and UKIP supporters are the keenest to turn out and cast a protest vote of any party . 25% vote share when turnout is just 20% would reflect around 6 to 8% on a GE turnout .

Even 6% - 8% would do the Tories serious damage. Let's face it, even the 3.1% in GE2010 damaged the Tories badly. E.g. Balls' seat.

david_herdson

MarkSenior said:

nigel4england said:

TOPPING said:

With the exception of very few, and I include many of the Kippers on PB, a tiny tiny amount of UKIP polling support is much more than NOTA/hell-in-a-handbasket/please someone make it all ok.

It's impossible to categorise or count UKIP voters because as has been pointed out, UKIP is such a recent phenomenon, and has occurred in such untypical circumstances (country really, really f&cked, NOTA party now in Govt...) that it is without precedent. So these are not polls of UKIP voters, they are polls of protest. of single issues, of NOTA.

Come GE2015 most UKIP "support" will dissolve into one of the three grown-up parties.

Wishful thinking, if the main parties and Crosby in particular share your view then you are in for a big surprise. It's the sort of arrogance that has helped to put UKIP where they are.

And where are UKIP today ? Unable to win council by elections when turnouts are typically 15 to 25% and UKIP supporters are the keenest to turn out and cast a protest vote of any party . 25% vote share when turnout is just 20% would reflect around 6 to 8% on a GE turnout .

Council by-elections aren't the be all and end all of politics. For a party whose primary purpose is specifically at a national / international level, it's surprising that UKIP do as well as they do at local level.

That said, it'll be interesting to see how they do in May, when we might get some idea of their split-vote levels and can use that against pre-UKIP split vote levels for the other parties at local/general elections to give some idea of how they'd do at a general election.

I think one thing we can all agree on is that UKIP will out-poll the Lib Dems at the Euros (again). The question is whether they'll outpoll everyone else as well (and also, whether the Lib Dems will finish in the top four or not).

nigel4england

I've bought the match performance tonight at 104, a nice night for mistakes and badly timed tackles

Charles

Sean_F said:

Charles said:

This just repeats the standard question though - and it is misleading just to think of it as UKIP being understated.

The two big shifts in the chart above are:

UKIP: 208 (10%) ---> 21 (1%)
DNV: 262 (13%) ---> 452 (22%)

The implication - as we all know - is that UKIP has attracted a lot of support from DNV.

Will they vote in 2015? I am sceptical, but may be they will.

The key question - which @NickPalmer tries to address from time to time, tbf - is whether there is any reason whether UKIP will manage to persuade DNV to change their voting habits.

There are two groups. Won't Vote. And Given Up Voting. 70-80% voted from 1979 to 1997. That dropped to 57% in 2001, and has only recovered to 65%.

UKIP won't get much support among the former. They will, among the latter.

That seems plausible, although I would argue that once someone has Given Up Voting for more than 1 election in a row they begin to morph into Won't Vote.

More interestingly, if the bulk of UKIP support at the election is from Given Up Voting then perhaps the impact on the other parties is overstated. i.e. the presumed loss of Tory seats comes from Tory voters shifting to UKIP and Labour coming out with a plurality. If UKIP is from GUV then - inter alia - the relative position of the Tories and Labour is unchanged and hence the FPTP result remains the same

david_herdson

One precedent we do have for the UKIP/WNV question is the Clegg bounce in the 2010GE, which according to the polls did pull in a lot of very soft support for two-three weeks but didn't materialise to anything like the same extent come polling day. Even so, the LD shares *did* pick up across the country in seats they didn't target which almost certainly was the legacy of that effect. Turnout was also up on the previous two elections although the extent to which that was due to Clegg and how much was due to the apparent closeness we don't know.

Cyclefree

Thanks to Morris Dancer and Floater for their very kind words from the previous thread.

Sean_F

nigel4england said:

AveryLP said:

RichardNabavi said:

You have to make a distinction between on-line polls, and telephone polls. The former are very prone to 'enthusiasm' bias: as we see in these august pages, the Kippers are terribly excited; that makes them much more likely to respond to a poll if they are members of a panel which is sent an invitation to participate in a political poll. Therefore it is almost certainly correct to weight them down in such polls, possibly by quite a lot. The unknown, of course, is by how much.

Considering just the on-line polls for the moment, we can see a huge disagreement between Survation, ComRes and Opinium on the one hand, and Populus on the other. The former routinely report UKIP figures in the 16-19% range, the latter in the 7-8% range. YouGov seems to be somewhere in the middle.

Normally we might be able to form a view on which is more likely to be correct by looking at the telephone polls, where the typical UKIP figure is around 10% or 11%, but they in turn might be subject to error because of the past-vote weighting.

I suspect that the telephone polls are a better guide, but we shall only really know the answer when the election comes and then only if, in the polls leading up to the election, the discrepancy persists.

Richard

UKIP supporters can't remember where they placed their spectacles five minutes ago.

Can pollsters really expect them to remember how they voted in May 2010?

OGH's big red marker pen should be drawing an arrow which points at the 208/10% figures not the 21/1% ones.

Wonderful stuff Avery, I will remember these cheap insults from you and your shambles of a party when you are whinging that it was UKIP that stopped you getting a majority in 2015.

Being rude about people whose support you want is a strange strategy.

TOPPING

Flockers_pb said:

I am not convinced it matters greatly if Ukip are being understated in the polls at the moment, or whether their true level of current support is 10% or 16%. The key question is whether the polls today are a good guide for the outcome of the election in 2015.

I believe that the polls are currently a poor guide to Ukip's 2015 outturn, which I expect to be below the bottom of their present polling range and worse than their current local election performances. I base that belief on Ukip's manifest lack of preparation and personnel to govern, which I trust the British people to identify, the dissipation of the protest vote as the election nears, the strengthening economy bringing some ex-Tories back to the fold, a strengthening of Liberal Democrat support as they break free from coalition and the likely return to loyalty of the Tory papers, amongst other factors.

If I wanted to support the alternative proposition I would point to the leader ratings demonstrating broad contempt for the leaders of the three main parties, Cameron's inability to connect to the Tory base and the appaling weakness of the Labour front bench providing fuel for "fourth party" support and the Euro boost we all expect Ukip to get next year providing momentum into the election campaign, as reasons to do so.

All are valid points, but the weight one attaches to them will vary according to one's perception and prejudices. The central point though, is even if the polls uniformly agreed that if the election was held tomorrow 16% of the electorate would vote UKIP, and even if that was a correct reflection of the electorate's current intention, it tells us very little about what will happen in 2015, in the same way that Tory scores in the mid-40s in 2009 or Labour scores in the high 50s/low 60s at various points from 1994 onwards told us little about what would happen at the ballot box. About as far as one can go is to note that the direction of travel is favourable to Ukip and they are likely to poll better in 2015 than they did in 2010.

All of this assumes, of course, no black swans. Dangerous assumption.

Yes.

In fact the only place I take issue is that even suppose all three grown-up parties were embroiled, for example as a vaguely possible black swan event, in huge fraud and criminality or somesuch, I would expect that the "good politics" remnants of those parties to come together to offer the electorate a credible choice over UKIP.

Even if today UKIP were handed the reins of power there are no policies with which to govern. I don't see them having the patience to create a comprehensive set of policies before 2015 which means forever protest, albeit a useful one like many protest parties, to keep their hot button issues on the broader agenda.

Edit: apologies snipped small part of one of your paragraphs so the answer would fit.

Pulpstar

There is a value in the raw figures.... Not the whole picture but it can be worthwhile to see if a movement is due to downscaling or a lack of raw numbers.

I think Nabavi is closest to the mark - I will expect the phone pollsters to be closest to the UKIP total on the night.

Right now that is ~ 11%.

TOPPING

Flockers_pb said:

antifrank said:

@Flockers_pb That's pretty much my own view (though I expect the Lib Dems to see the coalition out). I would also add that none of the three main parties have yet made their retail offers to the public for their votes. We can expect some eye-catching initiatives from all three, many of which will be calculated to appeal to the battlers who are currently UKIP-identifiers.

The current UKIP identifiers are going to be critical for the next election. Will they remain loyal to UKIP? Will they melt away? If so, how will they break between the other parties and non-voters?

My guess would be about 25-33% of previous non-voters and previous Conservative or Liberal Democrat voters who are currently saying they will vote Ukip will do so, and about 33-50% of previous Labour voters who are currently saying they will vote Ukip will do so. There's no scientific basis for that, just a guess based on observation over the years of the fragility of minority party support.

The problem that the main parties have in trying to appeal to Ukip supporters is that it is quite hard to identify who they are and what they support. They are a supposedly libertarian party with considerable support amongst social conservatives. Their anti-immigration message appeals to the Tory blue-rinser brigade and Labour's D/E heartlands (neither of which identifies with the other). They can't be out-Ukipped on Europe (don't even try Dave!). Polling data indicates them to be reliably negative and cynical, so it is unlikely good news will do the trick. I think that leaves the Tories flogging welfare reform and jobs, with a side order of Euro-bashing and Labour flogging cost of living, each to no real effect.

Again, a super-acute post ('cos it agrees with me, ahem).

There are so many contradictions not least wrt gay marriage which you would expect strong Kipper libertarian support for but which, um, they condemn.

Go figure.

AndyJS

The polling/elections event of the year for me was when, a few days before the local elections, some of the most experienced experts in the country predicted UKIP would register a projected national share of 12% and then those same experts calculated after the elections that the figure was actually 22%.

Pulpstar

There is still significant up, and downside potential to UKIP though. A big surge in the phone polls could be coming off the back of the immigration...

david_herdson

TOPPING said:

Yes.

In fact the only place I take issue is that even suppose all three grown-up parties were embroiled, for example as a vaguely possible black swan event, in huge fraud and criminality or somesuch, I would expect that the "good politics" remnants of those parties to come together to offer the electorate a credible choice over UKIP.

Even if today UKIP were handed the reins of power there are no policies with which to govern. I don't see them having the patience to create a comprehensive set of policies before 2015 which means forever protest, albeit a useful one like many protest parties, to keep their hot button issues on the broader agenda.

Edit: apologies snipped small part of one of your paragraphs so the answer would fit.

Protest parties can win elections. Governing afterwards is a different matter but that won't change the fact they the made it in the first place. Syriza may win the next Greek election. It'd be a disaster for Greece but that won't necessarily prevent it.

Now, I don't expect UKIP to win in 2015. Indeed, I don't expect them to win many seats at all, wouldn't be at all surprised if they won none and could see them back at under 5% nationally if the circumstances fall badly for them. On the other hand, not having too many policies and keeping the focus on those they're protesting against goes a long way with those not very engaged in politics (which is the majority - not far off half the GE electorate will only vote once every four or five years and skip the minor interim elections). If things go right for them, 25%+ is not unachievable.

TOPPING

david_herdson said:

TOPPING said:

Yes.

In fact the only place I take issue is that even suppose all three grown-up parties were embroiled, for example as a vaguely possible black swan event, in huge fraud and criminality or somesuch, I would expect that the "good politics" remnants of those parties to come together to offer the electorate a credible choice over UKIP.

Even if today UKIP were handed the reins of power there are no policies with which to govern. I don't see them having the patience to create a comprehensive set of policies before 2015 which means forever protest, albeit a useful one like many protest parties, to keep their hot button issues on the broader agenda.

Edit: apologies snipped small part of one of your paragraphs so the answer would fit.

Protest parties can win elections. Governing afterwards is a different matter but that won't change the fact they the made it in the first place. Syriza may win the next Greek election. It'd be a disaster for Greece but that won't necessarily prevent it.

Now, I don't expect UKIP to win in 2015. Indeed, I don't expect them to win many seats at all, wouldn't be at all surprised if they won none and could see them back at under 5% nationally if the circumstances fall badly for them. On the other hand, not having too many policies and keeping the focus on those they're protesting against goes a long way with those not very engaged in politics (which is the majority - not far off half the GE electorate will only vote once every four or five years and skip the minor interim elections). If things go right for them, 25%+ is not unachievable.

I think that would indicate a national collective moment of madness. Protest parties can win elections when the nation is if not in peril but in a situation where one or two key issues can drive voting preferences as those issues dominate all others.

@Flockers_pb's tail events notwithstanding I think your base case <5% is spot on.

corporeal

Charles said:

Sean_F said:

Charles said:

This just repeats the standard question though - and it is misleading just to think of it as UKIP being understated.

The two big shifts in the chart above are:

UKIP: 208 (10%) ---> 21 (1%)
DNV: 262 (13%) ---> 452 (22%)

The implication - as we all know - is that UKIP has attracted a lot of support from DNV.

Will they vote in 2015? I am sceptical, but may be they will.

The key question - which @NickPalmer tries to address from time to time, tbf - is whether there is any reason whether UKIP will manage to persuade DNV to change their voting habits.

There are two groups. Won't Vote. And Given Up Voting. 70-80% voted from 1979 to 1997. That dropped to 57% in 2001, and has only recovered to 65%.

UKIP won't get much support among the former. They will, among the latter.

That seems plausible, although I would argue that once someone has Given Up Voting for more than 1 election in a row they begin to morph into Won't Vote.

More interestingly, if the bulk of UKIP support at the election is from Given Up Voting then perhaps the impact on the other parties is overstated. i.e. the presumed loss of Tory seats comes from Tory voters shifting to UKIP and Labour coming out with a plurality. If UKIP is from GUV then - inter alia - the relative position of the Tories and Labour is unchanged and hence the FPTP result remains the same

70-80% is the standard range going a long way back in elections iirc (through various forms of enfranchisement). A couple of elections go outside this from time to time (eg 1950 and 1951) but as a rule it has returned to that range.

Of course history is no guarantee of the future, but it waggles its eyebrows suggestively in that direction.

Jonathan

Interesting thread Mike. Often thought the past weighting model wasn't good for the smaller parties.

Happy Xmas.

nigel4england

Man City at 1.8 at home to Liverpool is a great price, as is the 2.3 for City to win the league

AndyJS

Betting on turnout being >70% seems quite attractive when you consider that so many UKIP supporters are apparently people who haven't bothered voting for ages.

isam

antifrank said:

@isam Yes, I'll happily have that. Thank you.

EDIT I'll even let you do it on 2015 vote percentages rather than 2010.

oops! haha youre a sport!

AndyJS

London homicides so far this year are 96. Last year the figure was 99:

http://www.murdermap.co.uk/investigate.asp

SquareRoot

Jonathan said:

Interesting thread Mike. Often thought the past weighting model wasn't good for the smaller parties.

Happy Xmas.

That's why its overstating Labour.

Wulfrun_Phil

"A by-product of the UKIP surge is that many more people are “remembering” that they voted for the purples in 2010 than the 3.1% that actually did so. These responses are then scaled back according to formulas linked to the 2010 result."

That's quite plausible.

As is this: A by-product of the Lib Dem implosion is that many fewer people are “remembering” that they voted for the yellows in 2010 than the 23.6% that actually did so. These responses are then scaled up according to formulas linked to the 2010 result.

Is Lib Dem VI being systematically overstated in opinion polls?

MarkSenior

Wulfrun_Phil said:

"A by-product of the UKIP surge is that many more people are “remembering” that they voted for the purples in 2010 than the 3.1% that actually did so. These responses are then scaled back according to formulas linked to the 2010 result."

That's quite plausible.

As is this: A by-product of the Lib Dem implosion is that many fewer people are “remembering” that they voted for the yellows in 2010 than the 23.6% that actually did so. These responses are then scaled up according to formulas linked to the 2010 result.

Is Lib Dem VI being systematically overstated in opinion polls?

Whilst true of the online posters , that is not true of the telephone pollsters who are finding that around the correct number of people say they voted UKIP in 2010 as did so . Your comment re the Lib Dems is also incorrect as a number of pollsters particularly again the online ones find more people saying they voted Lib Dem in 2010 than actually did so . I can recall us discussing this phenomenum on here before .

Wulfrun_Phil

TOPPING said:

david_herdson said:

TOPPING said:

Yes.

In fact the only place I take issue is that even suppose all three grown-up parties were embroiled, for example as a vaguely possible black swan event, in huge fraud and criminality or somesuch, I would expect that the "good politics" remnants of those parties to come together to offer the electorate a credible choice over UKIP.

Even if today UKIP were handed the reins of power there are no policies with which to govern. I don't see them having the patience to create a comprehensive set of policies before 2015 which means forever protest, albeit a useful one like many protest parties, to keep their hot button issues on the broader agenda.

Edit: apologies snipped small part of one of your paragraphs so the answer would fit.

Protest parties can win elections. Governing afterwards is a different matter but that won't change the fact they the made it in the first place. Syriza may win the next Greek election. It'd be a disaster for Greece but that won't necessarily prevent it.

Now, I don't expect UKIP to win in 2015. Indeed, I don't expect them to win many seats at all, wouldn't be at all surprised if they won none and could see them back at under 5% nationally if the circumstances fall badly for them. On the other hand, not having too many policies and keeping the focus on those they're protesting against goes a long way with those not very engaged in politics (which is the majority - not far off half the GE electorate will only vote once every four or five years and skip the minor interim elections). If things go right for them, 25%+ is not unachievable.

I think that would indicate a national collective moment of madness. Protest parties can win elections when the nation is if not in peril but in a situation where one or two key issues can drive voting preferences as those issues dominate all others.

@Flockers_pb's tail events notwithstanding I think your base case <5% is spot on.</p>

I disagree. In GB, between them UKIP (3.1%) and the BNP (1.9%) got 5% in 2010. Given that the BNP have self-destructed, it seems inconceivable that UKIP won't by themselves get to 5% in 2015.

Stuart_Dickson

Scotland has had a 4 party system for decades. London based polling firms have never accounted for that. YouGov still hilariously downweights SNP respondents.

corporeal

Stuart_Dickson said:

Scotland has had a 4 party system for decades. London based polling firms have never accounted for that. YouGov still hilariously downweights SNP respondents.

How did the YouGov polling compare to the results at GE2010?

Pulpstar

Price on Sherwood has collapsed to odds on ? Info leaked ?

Pulpstar

Chelsea parking the bus

another_richard

FPT RN:

"That is completely the wrong way round. Those of us who think Osborne has judged this correctly do so precisely because he has correctly judged the speed of getting the public finances back towards some semblance of sanity at the fastest rate which is commensurate with NOT impeding recovery of growth. In other words, he has steered the optimal course between the Scylla of ever-worsening public finances combined with a collapse in market confidence in government debt, and the Charybdis of growth collapsing completely, with the attendant very high unemployment."

Scylla
Government debt May 2010 - £846bn
Government debt Nov 2013 - £1232bn

Charybdis
Industrial production May 2010 - 100.4
Industrial production Oct 2013 - 97.1

Osborne is no Odysseus

TOPPING

Wulfrun_Phil said:

TOPPING said:

david_herdson said:

TOPPING said:

Yes.

In fact the only place I take issue is that even suppose all three grown-up parties were embroiled, for example as a vaguely possible black swan event, in huge fraud and criminality or somesuch, I would expect that the "good politics" remnants of those parties to come together to offer the electorate a credible choice over UKIP.

Even if today UKIP were handed the reins of power there are no policies with which to govern. I don't see them having the patience to create a comprehensive set of policies before 2015 which means forever protest, albeit a useful one like many protest parties, to keep their hot button issues on the broader agenda.

Edit: apologies snipped small part of one of your paragraphs so the answer would fit.

Protest parties can win elections. Governing afterwards is a different matter but that won't change the fact they the made it in the first place. Syriza may win the next Greek election. It'd be a disaster for Greece but that won't necessarily prevent it.

Now, I don't expect UKIP to win in 2015. Indeed, I don't expect them to win many seats at all, wouldn't be at all surprised if they won none and could see them back at under 5% nationally if the circumstances fall badly for them. On the other hand, not having too many policies and keeping the focus on those they're protesting against goes a long way with those not very engaged in politics (which is the majority - not far off half the GE electorate will only vote once every four or five years and skip the minor interim elections). If things go right for them, 25%+ is not unachievable.

I think that would indicate a national collective moment of madness. Protest parties can win elections when the nation is if not in peril but in a situation where one or two key issues can drive voting preferences as those issues dominate all others.

@Flockers_pb's tail events notwithstanding I think your base case <5% is spot on.</p>

I disagree. In GB, between them UKIP (3.1%) and the BNP (1.9%) got 5% in 2010. Given that the BNP have self-destructed, it seems inconceivable that UKIP won't by themselves get to 5% in 2015.

Pretty damning view of bedfellows.

another_richard

Sean_F said:

nigel4england said:

AveryLP said:

RichardNabavi said:

You have to make a distinction between on-line polls, and telephone polls. The former are very prone to 'enthusiasm' bias: as we see in these august pages, the Kippers are terribly excited; that makes them much more likely to respond to a poll if they are members of a panel which is sent an invitation to participate in a political poll. Therefore it is almost certainly correct to weight them down in such polls, possibly by quite a lot. The unknown, of course, is by how much.

Considering just the on-line polls for the moment, we can see a huge disagreement between Survation, ComRes and Opinium on the one hand, and Populus on the other. The former routinely report UKIP figures in the 16-19% range, the latter in the 7-8% range. YouGov seems to be somewhere in the middle.

Normally we might be able to form a view on which is more likely to be correct by looking at the telephone polls, where the typical UKIP figure is around 10% or 11%, but they in turn might be subject to error because of the past-vote weighting.

I suspect that the telephone polls are a better guide, but we shall only really know the answer when the election comes and then only if, in the polls leading up to the election, the discrepancy persists.

Richard

UKIP supporters can't remember where they placed their spectacles five minutes ago.

Can pollsters really expect them to remember how they voted in May 2010?

OGH's big red marker pen should be drawing an arrow which points at the 208/10% figures not the 21/1% ones.

Wonderful stuff Avery, I will remember these cheap insults from you and your shambles of a party when you are whinging that it was UKIP that stopped you getting a majority in 2015.

Being rude about people whose support you want is a strange strategy.

I said back in 2011 that the Cameroons will lash out in all directions as their ship sinks.

The one group which they wont lay any blame against are themselves.

The sense of entitlement really does play badly.

corporeal

another_richard said:

FPT RN:

"That is completely the wrong way round. Those of us who think Osborne has judged this correctly do so precisely because he has correctly judged the speed of getting the public finances back towards some semblance of sanity at the fastest rate which is commensurate with NOT impeding recovery of growth. In other words, he has steered the optimal course between the Scylla of ever-worsening public finances combined with a collapse in market confidence in government debt, and the Charybdis of growth collapsing completely, with the attendant very high unemployment."

Scylla
Government debt May 2010 - £846bn
Government debt Nov 2013 - £1232bn

Charybdis
Industrial production May 2010 - 100.4
Industrial production Oct 2013 - 97.1

Osborne is no Odysseus

As is the way of PB, things went off after the metaphor. Odysseus first aimed at Scylla on the grounds that only some of his men would die (6 of them did) then later he had to go through the straits again he went closer to Charybdis and survived by clinging to a fig tree.

Not exactly an unqualified victory.

another_richard

TOPPING said:

Wulfrun_Phil said:

TOPPING said:

david_herdson said:

TOPPING said:

Yes.

In fact the only place I take issue is that even suppose all three grown-up parties were embroiled, for example as a vaguely possible black swan event, in huge fraud and criminality or somesuch, I would expect that the "good politics" remnants of those parties to come together to offer the electorate a credible choice over UKIP.

Even if today UKIP were handed the reins of power there are no policies with which to govern. I don't see them having the patience to create a comprehensive set of policies before 2015 which means forever protest, albeit a useful one like many protest parties, to keep their hot button issues on the broader agenda.

Edit: apologies snipped small part of one of your paragraphs so the answer would fit.

Protest parties can win elections. Governing afterwards is a different matter but that won't change the fact they the made it in the first place. Syriza may win the next Greek election. It'd be a disaster for Greece but that won't necessarily prevent it.

Now, I don't expect UKIP to win in 2015. Indeed, I don't expect them to win many seats at all, wouldn't be at all surprised if they won none and could see them back at under 5% nationally if the circumstances fall badly for them. On the other hand, not having too many policies and keeping the focus on those they're protesting against goes a long way with those not very engaged in politics (which is the majority - not far off half the GE electorate will only vote once every four or five years and skip the minor interim elections). If things go right for them, 25%+ is not unachievable.

I think that would indicate a national collective moment of madness. Protest parties can win elections when the nation is if not in peril but in a situation where one or two key issues can drive voting preferences as those issues dominate all others.

@Flockers_pb's tail events notwithstanding I think your base case <5% is spot on.</p>

I disagree. In GB, between them UKIP (3.1%) and the BNP (1.9%) got 5% in 2010. Given that the BNP have self-destructed, it seems inconceivable that UKIP won't by themselves get to 5% in 2015.

Pretty damning view of bedfellows.

But its okay if those same people vote Conservative ?

another_richard

corporeal said:

another_richard said:

FPT RN:

"That is completely the wrong way round. Those of us who think Osborne has judged this correctly do so precisely because he has correctly judged the speed of getting the public finances back towards some semblance of sanity at the fastest rate which is commensurate with NOT impeding recovery of growth. In other words, he has steered the optimal course between the Scylla of ever-worsening public finances combined with a collapse in market confidence in government debt, and the Charybdis of growth collapsing completely, with the attendant very high unemployment."

Scylla
Government debt May 2010 - £846bn
Government debt Nov 2013 - £1232bn

Charybdis
Industrial production May 2010 - 100.4
Industrial production Oct 2013 - 97.1

Osborne is no Odysseus

As is the way of PB, things went off after the metaphor. Odysseus first aimed at Scylla on the grounds that only some of his men would die (6 of them did) then later he had to go through the straits again he went closer to Charybdis and survived by clinging to a fig tree.

Not exactly an unqualified victory.

That's right I remember now.

It must have been just before Odysseus ended up on Calypso's island.

The modern equivalent would be getting a big job at the UN.

My favourite part of the Odyssey was the build up to big fight at the end where the suitors get killed.

SandyRentool

Rentool formula for predicting UKIP GE vote share (to be applied next May):
Apply the percentage decline in UKIP vote share per annum from previous Euro and GE vote shares, and apply this to the 2014 Euro share. Out pops the 2015 GE vote share...

anotherDave

AndyJS said:

Betting on turnout being >70% seems quite attractive when you consider that so many UKIP supporters are apparently people who haven't bothered voting for ages.

There's been a lot of immigration/emigration since the last >70% turnout.

1997-2011 Non-British Net Migration: 3.7 million
1997-2011 British Net Migration: -1 million

http://www.migrationwatchuk.org.uk/pdfs/BP12_3.pdf

http://www.migrationwatchuk.org.uk/briefing-papers/category/9

Quincel

Pulpstar said:

Price on Sherwood has collapsed to odds on ? Info leaked ?

Much appreciated post Pulpstar, got on him at almost evens just in time thanks to you.

TheScreamingEagles

Times have it that Sherwood has been offered the Spurs Manager job on an interim basis until the end of the season.

Edit Guardian have it as well

Tottenham offer Tim Sherwood interim manager role until summer
• Spurs hand over the reins after impressive win at Southampton
• Van Gaal and De Boer seen as potential long-term candidates

http://www.theguardian.com/football/2013/dec/23/tim-sherwood-tottenham-hotspur-manager

TheScreamingEagles

Apologies if this has been already posted

A previously unpublished independent report has found evidence that veteran Portsmouth MP Mike Hancock sexually assaulted and harassed a constituent and made "unwelcome sexual approaches", the Guardian can reveal.

Nigel Pascoe QC, a leading barrister in sexual crime cases, interviewed Hancock's accuser at length at the request of Portsmouth city council and concluded in August that there was evidence of sexual advances made by Hancock and that the MP was fully aware of his alleged victim's mental health problems.

http://www.theguardian.com/politics/2013/dec/23/mike-hancock-sexual-harassment-allegations

Morris_Dancer

Good evening, everyone.

Miss Cyclefree, it's your own fault. If your posts had been more ill-considered, bad-tempered or poorly thought out I wouldn't've said anything

Sean_F

another_richard said:

Sean_F said:

nigel4england said:

AveryLP said:

RichardNabavi said:

You have to make a distinction between on-line polls, and telephone polls. The former are very prone to 'enthusiasm' bias: as we see in these august pages, the Kippers are terribly excited; that makes them much more likely to respond to a poll if they are members of a panel which is sent an invitation to participate in a political poll. Therefore it is almost certainly correct to weight them down in such polls, possibly by quite a lot. The unknown, of course, is by how much.

Considering just the on-line polls for the moment, we can see a huge disagreement between Survation, ComRes and Opinium on the one hand, and Populus on the other. The former routinely report UKIP figures in the 16-19% range, the latter in the 7-8% range. YouGov seems to be somewhere in the middle.

Normally we might be able to form a view on which is more likely to be correct by looking at the telephone polls, where the typical UKIP figure is around 10% or 11%, but they in turn might be subject to error because of the past-vote weighting.

I suspect that the telephone polls are a better guide, but we shall only really know the answer when the election comes and then only if, in the polls leading up to the election, the discrepancy persists.

Richard

UKIP supporters can't remember where they placed their spectacles five minutes ago.

Can pollsters really expect them to remember how they voted in May 2010?

OGH's big red marker pen should be drawing an arrow which points at the 208/10% figures not the 21/1% ones.

Wonderful stuff Avery, I will remember these cheap insults from you and your shambles of a party when you are whinging that it was UKIP that stopped you getting a majority in 2015.

Being rude about people whose support you want is a strange strategy.

I said back in 2011 that the Cameroons will lash out in all directions as their ship sinks.

The one group which they wont lay any blame against are themselves.

The sense of entitlement really does play badly.

Between them, Avery and Topping exemplify that attitude perfectly. They hate the people whose support they demand.

TheScreamingEagles

George Osborne as Odysseus?

I've had him down as the Artabanus to David Cameron's Xerxes I.

TheScreamingEagles

SPURS LATEST

Phil McNulty ‏@philmcnulty 48s

Tim Sherwood has been appointed Tottenham Head Coach with a contract to the end of the 2014/15 season.

Sky Sports News have it as well

Pulpstar

Spurs manager markets pulled/suspended !

Quincel

And the bookies have suspended the market. BBC have it too.