politicalbetting.com » Blog Archive » Lord Ashcroft’s Newark poll: Tories set to win easily

Re. Lord A's national poll;

From what I can see 25% is the lowest vote share the Conservatives have had in a national opinion poll since December 2001;

http://ukpollingreport.co.uk/historical-polls/voting-intention-2001-2005

So Con effectively at a 13 year low with less than a year to go to a general election.

Not ideal, to say the least.

The swing in Newark in Ashcroft's poll is a mere 5% from Conservative to Labour. Labour will want to be doing better than that.

FPT
So far the Ashcroft poll have been all over the place from a lead of 2 for the tories to a lead of 9 for Labour in 15 days. I mean, really?

Unless they can settle down a bit they really are not going to be much use.

Chart now fixed.

OblitusSumMe said:

The swing in Newark in Ashcroft's poll is a mere 5% from Conservative to Labour. Labour will want to be doing better than that.

I reckon alot of Labour voters are heading off for a tactical UKIP vote in Newark - sounds like Labour is soft pedalling there.

Not a good strategy but might explain the disconnect a bit.

Didn't the Eastleigh polls underestimate UKIP support there? (per last week's thread?)

Life_ina_market_town said:

If Baron Ashcroft is underestimating the Tories nationally as seems to be the case, does it follow that he is underestimating the Tories in Newark?

No, because the underestimation in the National poll is likely to do with the specific sample of C2 voters, and the sample for Newark will of course be completely different.

UKIP's support seems to be spread quite evenly across the age range.

18-24: 36%
25-34: 22%
25-44: 28%
45-54: 31%
55-64: 29%
65+: 30%

(table3, p.4)
http://lordashcroftpolls.com/wp-content/uploads/2014/06/Newark-by-election-poll-Full-tables.pdf

Clearly the 9 point lead is an outlier - but then we said the same about the YG 7 point lead too.

I think the improvement for Labour is to be expected given:

1. The implosion of the Lib Dems in the last week - this may have pushed more switchers to Labour.

2. The UKIP publicity boost from the Euros is likely to have pushed Con to UKIP switchers.

3. The fading of the Euro boost for the Greens may be helping Labour.

The comforting thing for the Conservatives is that these are generally likely to be short term effects - so would expect to see the polls drift closer again particularly if Newark bursts UKIPs bubble.

If we can't trust Ashcroft's national poll, can we thrust his Newark poll?

OK, Con has a decent margin for error in Newark but if, say, Con is overstated by 5 and UKIP is understated by 5 then we have Con 37, UKIP 32 which could be still close on the day with some tactical voting.

However Con will be pleased that Lab figure is still good enough that there is unlikely to be massive tactical voting at the last minute - though there may well still be a fair bit.

MikeL said:

If we can't trust Ashcroft's national poll, can we thrust his Newark poll?

OK, Con has a decent margin for error in Newark but if, say, Con is overstated by 5 and UKIP is understated by 5 then we have Con 37, UKIP 32 which could be still close on the day with some tactical voting.

However Con will be pleased that Lab figure is still good enough that there is unlikely to be massive tactical voting at the last minute - though there may well still be a fair bit.

UKiP already have 22% of 2010 Labour voters. More might follow.

Oh wow, the internals of the Ashcroft poll are actually even worse for the Lib Dems than the 6% figure:

Will vote Lib Dem, and will definitely vote Lib Dem in a year's time: 16 respondents of 1000.

Of course they may pick up from other parties, and it could be an outlier, but that is a dreadful internal.

Mr. Blackburn, welcome to pb.com.

If UKIP got 20+ seats that would be staggeringly good for them, but I do think it's quite incredible.

Patrick said:

MikeL

I'm sure UKIP will STAND in a huge number of seats, possibly all of them. But they will TARGET their limited resources at the much smaller number they think they might win. Their other candidates will be running a DIY campaign locally.

Of course - I know that. The Greens will also stand in every seat (or almost every seat).

But that's not the point. If you are saying to OFCOM you are a "Major Party" that means you should have a reasonable expectation of winning a substantial number of seats.

OblitusSumMe said:

DavidL said:

FPT
So far the Ashcroft poll have been all over the place from a lead of 2 for the tories to a lead of 9 for Labour in 15 days. I mean, really?

Unless they can settle down a bit they really are not going to be much use.

I've been meaning to do a proper statistical analysis of opinion polling to estimate how much variability you would expect from poll-to-poll from sampling errors alone - ie assuming no underlying change in opinion.

It could be that the Ashcroft Poll is given you a genuine measure of the uncertainty, and the other polls are erroneously damping the variation via some adjustment.

Hopefully I will get around to this at some point, but I would caution that even with a perfect opinion poll you would expect a large amount of variation from poll to poll, and none of the opinion polls available are perfect.

That would be interesting.

I suspect that the established polling companies adjust and adjust after each election to reflect the discrepancies between their findings and the actual result. I think this is entirely sensible when your objective is to get as close as possible to an accurate prediction but I agree with you that it probably dampens their volatility considerably as well.

The change in the weighting when you can't find enough of a particular class of voter probably underplays the fact that some of that kind of voter is now so ticked off they won't admit who they voted for previously and by simply scaling up those who will admit it you are distorting the group.

I understand that Lord Ashcroft has got an established polling company doing his work. It is therefore curious that his results are not similarly tempered. No doubt next week will see a remarkable tory recovery. I am afraid that will not be very convincing either .

kieran said:

3. The fading of the Euro boost for the Greens may be helping Labour.

If I read the tables properly (never something to take for granted) the Greens are on 7% which is as high as I remember their GE voting intention being. I expect it to be back to 2% or so in the actual GE so that's perhaps a good potential boost for Labour (and Lib Dems in some areas) to come.

Life_ina_market_town said:

If Baron Ashcroft is underestimating the Tories nationally as seems to be the case, does it follow that he is underestimating the Tories in Newark?

We'll see on Thursday if that is indeed the case.

Labour have a bit of a dilemma in Newark though, do they try and GOTV or soft pedal

Is there anti-UKIP Lab -> Con switching going on also ?

UKIP are far far worse than the Conservatives according to some of my more left wing friends.

RodCrosby said:

Ashcroft National UNS

Lab 370
Con 220
LD 31
UKIP 0
Oths 11
NI 18

Lab maj 90

First person Ed get's on to dog to is...

Farage!

"Thanks for everything Nigel."

Pulpstar said:

Life_ina_market_town said:

If Baron Ashcroft is underestimating the Tories nationally as seems to be the case, does it follow that he is underestimating the Tories in Newark?

We'll see on Thursday if that is indeed the case.

Labour have a bit of a dilemma in Newark though, do they try and GOTV or soft pedal

Is there anti-UKIP Lab -> Con switching going on also ?

UKIP are far far worse than the Conservatives according to some of my more left wing friends.

There was a hint of that in the vox populi on R5 from Newark this morning where one of those spoken to was determinedly anti UKIP and uncertain where to place his vote between the 2 main parties.

There was also a baker selling buns with different colours of icing. I forget the exact figures but his tory figure and lead was pretty close to this poll!

Not sure of his source but:

@ TomMayerEuropa
!!! German Chancellor #Merkel says: if necessary vote for #juncker in EuCo against #cameron and british objections

OpenEurope have some possible blocking minorities:
https://pbs.twimg.com/media/BpIRhvDIYAAVdAq.jpg
Start with UK, Sweden, Holland and Hungary, then find another large country, the obvious ones being Italy (who are being a bit vague) and France (who want their former Socialist finance minister - not sure Cameron could sell that one as a win...)

The reason I still like Shadsy's 4/6 odds is that Juncker gets multiple shots at this. Not only is he clearly odds-on for the first vote, if the heads of government vote for somebody else the first time the parliament are likely to vote them down and tell them to try again.

GIN1138 said:

currystar said:

Ok who genuinely believes that the tories will poll 25% at next years GE?

The Ashcroft polls are a bit of an embarrassment to him.

The Tories won't poll 25% in the general election, but it other pollsters were to suddenly start finding a collapse in the Conservative vote it would be very worrying, so close to a general election.

Dangerous times for Con this. Newark MUST be held!

Cameron is wobbling. Labour sympathizers in Newark should tactically vote en masse for UKIP to further his instability.

Pulpstar said:

We'll see on Thursday if that is indeed the case.

Labour have a bit of a dilemma in Newark though, do they try and GOTV or soft pedal

Is there anti-UKIP Lab -> Con switching going on also ?

UKIP are far far worse than the Conservatives according to some of my more left wing friends.

Pro-Conservative tactical voting really would be a new phenomenon! Usually, the anti-Tory tactical voting is very efficient.

If the 'The party has the best candidate locally' respondents really were voting on that basis, it wouldn't be a winner from the top four - look at how many voting for one of the others have (correctly) said that they're better.

It also confirms what I said this morning - only the Tories have run a decent campaign in terms of voter contact: the LibDems have operated on a shoestring, Labour have been disappointingly poor at it, and UKIP has concentrated on leaflets rather than canvassing.

At least it makes it easier to know which way to do the postal vote.

GIN1138 said:

currystar said:

GIN1138 said:

currystar said:

Ok who genuinely believes that the tories will poll 25% at next years GE?

The Ashcroft polls are a bit of an embarrassment to him.

The Tories won't poll 25% in the general election, but it other pollsters were to suddenly start finding a collapse in the Conservative vote it would be very worrying, so close to a general election.

Dangerous times for Con this. Newark MUST be held!

Two weeks ago people on here were talking about how well the tories were doing in the polls and their was regular crossover. What has happened in the two weeks to make such dramatic changes? Maybe its just pollsters going for headlines. This poll reckons that others not including UKIP are on 16% , with UKIP they are on 35%. Its frankly daft. How can labour only be on 34% when the tories are on 25%?

I would guess what's changed is the publicity UKIP have achieved from their Euro election result.

Conservatives have held themselves together remarkably well, but if Newark is lost and more pollsters (especially YouGov or ICM) start taking them into the mid 20's all bet are off in terms of another bout of Con blood-letting.

Why would there be, the economy is booming, unemployment is falling, wages are rising, the UK is a good place to be. If people want to vote to allow a party who will ruin the economy again to govern its hardly the tory leadership's fault. People are just daft.

kieran said:

Having a look at the internals of the Ashcroft national poll it really is a stonking one for Labour.

They actually had a 10 point lead (35-25) before the spiral of silence adjustment.

As Neil points out the Greens are on 7% - which gives a sizeable vote still left to squeeze in the run up to the GE.

It also matches virtually every other poll this parliament showing very little switching between Lab and Con.

The main movement is being caused by 22% of 2010 Con voters going to UKIP and 34% of 2010 LD voters going to Labour.

For the LDs to be only keeping 22% of their own voters really is dreadful. They have to hope that they can at least recover to 10 - 12% once the negative leadership stuff dies down otherwise next year could turn into an extinction-level event - regardless of the incumbency factor.

Populus has Lib Dems retaining 41% of their 2010 voters .
We have two polls taken at the same time with differing results . pointless hypothesising with assumptions that the Ashcroft poll is the correct one .

GIN1138 said:

Nigel want's a referendum? Surely not. Imagine if NO won and we left the EU he'd be off the gravy train and UKIP's whole raison d'etre would be lost.

Cameron actually solves that for Farage with his "renegotiation" thing. Since the EU can't pass a treaty on that timescale, it's bound to involve some promises of things happening in a future treaty, so when that treaty ultimately falls apart, gets voted down by Britain or fails to live up to everything it was said to do in the campaign, Farage can say he was robbed, start demanding a new referendum and get the whole show back on the road again.

The odds haven't changed much with Betfair since the poll came out:

http://www.betfair.com/exchange/politics/market?id=1.113986228

GIN1138 said:

currystar said:

Ok who genuinely believes that the tories will poll 25% at next years GE?

The Ashcroft polls are a bit of an embarrassment to him.

The Tories won't poll 25% in the general election, but it other pollsters were to suddenly start finding a collapse in the Conservative vote it would be very worrying, so close to a general election.

Dangerous times for Con this. Newark MUST be held!

Newark to be Cameron's Stalingrad?

@currystar

Daft? Or shoved into a dead end job, with pay increases of circa 1% being the norm for most of the rest?
The rise in "average" remuneration is dependent mainly on which side of the line you fall, and is skewed towards those who were already comfortably well off, or in many cases, fabulously wealthy.

Pulpstar said:

I reckon alot of Labour voters are heading off for a tactical UKIP vote in Newark - sounds like Labour is soft pedalling there.

Yes. Labour is not running a proper campaign here. It's much better than the LibDems, but there is no visible Labour HQ in the centre of Newark, for example.

For a seat they won in 1997, it's rather embarrassing.

I understand from talking to two other parties that they think a chunk of 2010 Labour voters are going to vote UKIP with one set of fingers crossed (that they can beat the Tories) and a clothes peg on their nose (because they're really not UKIP voters).

I can see UKIP morphing into the WWC party and the metrosexual leftie luvvies joining the leftie LD's with the sensible ones migrating to Conservative so, effectively Labour would disappear. Hurrah.

Blue_rog said:

I can see UKIP morphing into the WWC party and the metrosexual leftie luvvies joining the leftie LD's with the sensible ones migrating to Conservative so, effectively Labour would disappear. Hurrah.

Labour won't disappear. Another party politically closish to them may do...

There are some interesting figures in the Newark poll for switchers between parties. Using the turnout-weighted figures (without the 'spiral of silence' adjustment), 2010 voters are splitting as follows:

2010 Con -> Con 69%, Lab 2%, LD 0%, UKIP 27%
2010 Lab -> Con 7%, Lab 63%, LD 2%, UKIP 22%
2010 LD -> Con 13%, Lab 16%, LD 22%, UKIP 28%

One needs to be a bit cautious about these figures, because the sample sizes are small, but three points stand out: (a) LibDems not in a good place, (b) UKIP attracting support rather evenly from the other three parties, (c) Little evidence of overwhelming LibDem->Lab switching here at least, indeed LD->Con and LD->Lab are fairly similar, but LD->UKIP is twice as big.

The fact that UKIP seem to be taking vote share from three contradictory directions suggests the NOTA protest component dominates.

http://lordashcroftpolls.com/wp-content/uploads/2014/06/Newark-by-election-poll-Full-tables.pdf
Table 3, Q2

I'm still sticking by my Newark prediction:

CON HOLD: 2,000-3,000 majority.

@OwenJones84: It'll be such an unlikely upset if the Tories lose Newark that I'll run through Leicester Square naked singing Lady Gaga's 'Born This Way'

Smarmeron said:

They didn't win that seat, they won one with the same name but different boundaries.

It is true that Retford went and was replaced by Bingham, but the bulk of the constituency is the same.

Even with a Labour candidate who had been found guilty of "corrupt practices" and disqualified (until that was overturned on appeal) with the local Party split for/against her, they still came within less than 4,100 votes of keeping the seat in 2001.

They should be doing a lot better than that in a by-election caused by the fallen Tory MP being worse.

Smarmeron said:

@currystar

Daft? Or shoved into a dead end job, with pay increases of circa 1% being the norm for most of the rest?
The rise in "average" remuneration is dependent mainly on which side of the line you fall, and is skewed towards those who were already comfortably well off, or in many cases, fabulously wealthy.

Total nonsense, in my business during 2006-2010 our business shrunk by 30%, since 2010 it has grown by 50%. During that period we have taken on 8 new apprentices and intend to take on another 4 this year. This is giving young people a chance at a trade for life and a successful career. There are tens of thousands of such jobs being created. These are proper jobs which will help the economy progress, not Labour pretend jobs in the Public Sector or Labour benefit bribes to shore up its vote.

You cannot expect wages to rise dramatically following the economic crisis this economy has gone through, but at least they are gradually increasing and jobs are being still being created.

There are a million less public sector workers now since 2010. What jobs were these people doing? My bins are still collected,the roads are still repaired, the hospitals are still open. If Labour get back in they will do the same. Create pretend jobs in the Public Sector which the country cant afford, it will destroy the work ethic that the coalition have developed in the country and they will start with their benefit handouts again to show how "fair" they are being. This policy simply does not work, people need an incentive to work, if you give people the opportunity to do less with the same money through benefits they will take it. This will again kill the economy and the tories will have to rebuild it again. It really confuses me why people want to go through the same old cycle.

Surely looking at what is happening in France would give a clue as to what will happen here if Labour get back in.

People hated Alex Ferguson, but it didn't mean that his management style was wrong, people hate the tories, but it does not mean their policies are wrong.

MoreCynicalThanThou said:

Even with a Labour candidate who had been found guilty of "corrupt practices" and disqualified (until that was overturned on appeal) with the local Party split for/against her, they still came within less than 4,100 votes of keeping the seat in 2001.

The boundary changes occurred after 2005. Labour would not have won the current constituency in 1997. Though it would have been close and it's strange to see a main opposition party not in the running in a by-election in a seat like this but we live in rather strange electoral times.

OblitusSumMe said:

DavidL said:

FPT
So far the Ashcroft poll have been all over the place from a lead of 2 for the tories to a lead of 9 for Labour in 15 days. I mean, really?

Unless they can settle down a bit they really are not going to be much use.

I've been meaning to do a proper statistical analysis of opinion polling to estimate how much variability you would expect from poll-to-poll from sampling errors alone - ie assuming no underlying change in opinion.

It could be that the Ashcroft Poll is given you a genuine measure of the uncertainty, and the other polls are erroneously damping the variation via some adjustment.

Hopefully I will get around to this at some point, but I would caution that even with a perfect opinion poll you would expect a large amount of variation from poll to poll, and none of the opinion polls available are perfect.

I've long been of the opinion that the weightings the online firms apply to UKIP are pure comedy gold. 19% in a phone poll is very significant, and Ashcroft/Populus' raw samples were statistically closish in terms of raw VI.

The NEV of 17.5 for UKIP may well be close to the mark as calculated by Rawling & Thrasher at the locals.... THey ain't on 11.

Similiarly Lib Dems benefit from generous upweighting alot of the time.

Note the lack of weighting in phone polls compared to online.

Sunil_Prasannan said:

Didn't the Eastleigh polls underestimate UKIP support there? (per last week's thread?)

Yes, I won't be ditching my UKIP (~ Nil) & Labour Green for silly prices on Betfair now...

anotherDave said:

Richard_Nabavi said:

There are some interesting figures in the Newark poll for switchers between parties. Using the turnout-weighted figures (without the 'spiral of silence' adjustment), 2010 voters are splitting as follows:

2010 Con -> Con 69%, Lab 2%, LD 0%, UKIP 27%
2010 Lab -> Con 7%, Lab 63%, LD 2%, UKIP 22%
2010 LD -> Con 13%, Lab 16%, LD 22%, UKIP 28%

One needs to be a bit cautious about these figures, because the sample sizes are small, but three points stand out: (a) LibDems not in a good place, (b) UKIP attracting support rather evenly from the other three parties, (c) Little evidence of overwhelming LibDem->Lab switching here at least, indeed LD->Con and LD->Lab are fairly similar, but LD->UKIP is twice as big.

The fact that UKIP seem to be taking vote share from three contradictory directions suggests the NOTA protest component dominates.

http://lordashcroftpolls.com/wp-content/uploads/2014/06/Newark-by-election-poll-Full-tables.pdf
Table 3, Q2

Survation's Newark poll was:

2010 Con -> Con 63%, Lab 1%, UKIP 31%
2010 Lab -> Con 2%, Lab 75%, UKIP 16%
2010 LD -> Con 18%, Lab 24%, UKIP 30%

(table 5, page 9)
http://survation.com/wp-content/uploads/2014/05/Newark-Poll-Sun-Tables.pdf

I'd say using table 4 would be better (i.e. not excluding don't knows etc)

Pulpstar said:

Neil said:

MoreCynicalThanThou said:

Even with a Labour candidate who had been found guilty of "corrupt practices" and disqualified (until that was overturned on appeal) with the local Party split for/against her, they still came within less than 4,100 votes of keeping the seat in 2001.

The boundary changes occurred after 2005. Labour would not have won the current constituency in 1997. Though it would have been close and it's strange to see a main opposition party not in the running in a by-election in a seat like this but we live in rather strange electoral times.

Neil said:

MoreCynicalThanThou said:

Even with a Labour candidate who had been found guilty of "corrupt practices" and disqualified (until that was overturned on appeal) with the local Party split for/against her, they still came within less than 4,100 votes of keeping the seat in 2001.

The boundary changes occurred after 2005. Labour would not have won the current constituency in 1997. Though it would have been close and it's strange to see a main opposition party not in the running in a by-election in a seat like this but we live in rather strange electoral times.

Who here remembers the Richmond 1989 By-Election ?

Was it anything like this ?

UKIP seem to be doing the SDP role in this one.

Richmond 1989 was a rather infamous one where the newly merged Alliance (or whatever name we were on at that time) candidate ran against a splinter continuity SDP candidate and split the vote to let the Conservatives in.

UKIP are doomed. Doomed I tell you based on the Newark poll.

The Tories are doomed. Doomed I tell you based on the National poll.

corporeal said:

Charles said:

kieran said:

Clearly the 9 point lead is an outlier - but then we said the same about the YG 7 point lead too.

I think the improvement for Labour is to be expected given:

1. The implosion of the Lib Dems in the last week - this may have pushed more switchers to Labour.

2. The UKIP publicity boost from the Euros is likely to have pushed Con to UKIP switchers.

3. The fading of the Euro boost for the Greens may be helping Labour.

The comforting thing for the Conservatives is that these are generally likely to be short term effects - so would expect to see the polls drift closer again particularly if Newark bursts UKIPs bubble.

I don't get the bubble burst argument.

Sure, in terms of media stories that may be so. And perhaps, at the margin, there are people who are just following a bandwagon in supporting UKIP / perhaps there will be some fair weather friends who get dispirited by the lack of progress.

But the fundamental reasons why people are supporting UKIP still exist. It will make it harder to grow, but why should the numbers deflate massively post a Newark loss?

People like supporting winning parties.

The LD tragedy suggests otherwise.

On the subject of "self employment" as pushed by those privatized job companies especially towards the long term unemployed, and those with circumstances that make normal employment almost impossible (those that Atos deem fit for work), there is a pretty big problem heading towards them, albeit slowly, and of course not till after the next election.
The much delayed flagship of "Universal Credit" will have massive impact on the future viability of "start ups" and small businesses unless there are reforms to it before full implementation.
But who cares really? As long as it remains hidden for the moment

MonikerDiCanio said:

corporeal said:

Charles said:

kieran said:

Clearly the 9 point lead is an outlier - but then we said the same about the YG 7 point lead too.

I think the improvement for Labour is to be expected given:

1. The implosion of the Lib Dems in the last week - this may have pushed more switchers to Labour.

2. The UKIP publicity boost from the Euros is likely to have pushed Con to UKIP switchers.

3. The fading of the Euro boost for the Greens may be helping Labour.

The comforting thing for the Conservatives is that these are generally likely to be short term effects - so would expect to see the polls drift closer again particularly if Newark bursts UKIPs bubble.

I don't get the bubble burst argument.

Sure, in terms of media stories that may be so. And perhaps, at the margin, there are people who are just following a bandwagon in supporting UKIP / perhaps there will be some fair weather friends who get dispirited by the lack of progress.

But the fundamental reasons why people are supporting UKIP still exist. It will make it harder to grow, but why should the numbers deflate massively post a Newark loss?

People like supporting winning parties.

The LD tragedy suggests otherwise.

The LDs for years is a key example of it. There's a reason "Winning Here" posters get stuck up everwhere.

Firstly there's a wasted vote element, where they won't vote for someone if they don't think they have a chance of winning.

Secondly, there's a kind of crowd enthusiasm thing where people feel more positively about a party because of the sense lots other people seem to feel positive about them.