Cameron faces election bloodbath as shock poll reveals Farage's 44% lead over Tories

[-:Undertaker:-]

http://www.dailymail.co.uk/news/article-2738787/Cameron-faces-Ukip-election-bloodbath-Party-set-win-Commons-seat-shock-poll-reveals-Farages-staggering-44-point-lead-Tories.html

Cameron faces Ukip by-election bloodbath: Party set to win first Commons seat after shock poll reveals Farage's staggering 44-point lead over the Tories

- Ukip set to win first Commons seat with a landslide 64 per cent of the vote
- Douglas Carswell is set to win Clacton by-election sparked by his defection
- Figures predict a record 48 point swing towards Ukip, humiliating the Tories


http://www.ezimba.com/work/140901C/ezimba13698118847000.png


Ukip are set to win their first Commons seat with a landslide 64 per cent of the vote following the biggest swing in modern political history.

Turncoat MP Douglas Carswell is set to humiliate David Cameron at the Clacton by-election sparked by his defection, a Survation poll for The Mail on Sunday has revealed.

The figures – the first test of public opinion since the politician rocked Westminster by defecting to Nigel Farage’s party – predict a record 48 point swing towards Ukip.


http://i.dailymail.co.uk/i/pix/2014/08/31/article-2738787-20ED5BAD00000578-85_634x891.jpg


And they put the anti-Brussels party a staggering 44 percentage points ahead of Mr Carswell’s former party in the Essex constituency.

If the results are repeated in the by-election, expected in October, the swing would exceed the current record 44 points achieved by Lib Dem MP Simon Hughes when he thrashed Labour’s Peter Tatchell, the gay rights campaigner, in a by-election in Bermondsey, South London, in 1983.

The biggest anti-Tory swing to date is the 35 points achieved by the Lib Dems in the Christchurch by-election in 1993.

This really IS a political earthquake, writes Damian Lyons Lowe, chief executive of poll firm Survation


http://www2.pictures.zimbio.com/gi/UKIP+Leader+Nigel+Farage+Douglas+Carswell+wXBQKxU_ KBVl.jpg


Political earthquake is an expression often used without justification, but not in the case of our Clacton poll result.

It shows that the by-election triggered by sitting Conservative MP Douglas Carswell’s defection to Ukip could go down in British political history as one of the most dramatic of all time.

Based on the poll, the 12,000 majority Mr Carswell won as a Conservative in Clacton in 2010 would turn into an even bigger one for Ukip of more than 15,000 on a similar turnout.

It has alarming implications for David Cameron and the Conservatives.

Mr Carswell has been accused of betraying his party, but that is not how the vast majority of his constituents see him.

Today’s survey lays to rest the notion that Mr Carswell defected for cynical reasons after becoming convinced that if he didn’t, he would lose to Ukip.

Such is his personal popularity that he would almost certainly have won the seat had he fought it with a blue rosette next May, rather than now in the yellow and purple colours of Ukip.

Most worrying of all for the Prime Minister is the reaction of other potential Tory MP defectors to the likely outcome of the by-election.

If Mr Carswell lost or won narrowly, other like-minded Conservatives toying with going over to Ukip would probably conclude it was far too risky.

But if he wins with a record-breaking swing, as our poll suggests he could, they may decide it is more risky to fight the next election as a Tory candidate than as a Ukip one.

And that could change the landscape of British politics.

Holy moly, that poll. If such a thing were to occur at the ballot box it'd represent the biggest by-election swing in modern history.

The Tories now face a huge dilemma, do they decide to really fight the by-election by sending ministers and the PM down to the seat, only to be humiliated - or do they simply make no effort, risking more defections to Ukip?

All reminds me increasingly of this by-election (http://en.wikipedia.org/wiki/Deborah_Grey#Political_career) in Canada, one that would change Canadian politics forever.

Thoughts?

[iBlueBox]

green party should win,

[Kardan]

I would be genuinely surprised if this ended up being the result.

Looking at the data, it seems the sample size is small as well, only 700 people in total. And some wards within the constituency have as little as 10 people being asked.

Also, out of those 700 people they asked - 55 of them voted for UKIP in the last election - but UKIP didn't run for Clacton last time around, so I'm guessing all of those voters have moved to the area in the last 4 years?

And then, out of those 700 people, 85 of them voted for Labour in 2010. That doesn't sound like a fair representation to me, I feel like either there's proportionally a lot of UKIP voters, or a lack of labour voters.

And for certain questions, the Lib Dems are getting various percentages on such opinions around ~30% etc. Which sounds like a lot, but only 4 people picked that answer. Also, if we take this poll into account, 0% of Lib Dem voters in Clacton like Nigel Farage (that's 0/12 people!). I'm pretty sure that's not true either.

- - - Updated - - -

Also, out of the 700 people they asked, the age distribution was as follows:

18-24 - 5.9%
25-34 - 12.1%
35-44 - 8.9%
45-54 - 18.6%
55-64 - 13.9%
65+ - 40.9%

Genuine question (because obviously the 65+ range covers more people), are polls always this way?

[-:Undertaker:-]

I would be genuinely surprised if this ended up being the result.

Looking at the data, it seems the sample size is small as well, only 700 people in total. And some wards within the constituency have as little as 10 people being asked.

Also, out of those 700 people they asked - 55 of them voted for UKIP in the last election - but UKIP didn't run for Clacton last time around, so I'm guessing all of those voters have moved to the area in the last 4 years?

And then, out of those 700 people, 85 of them voted for Labour in 2010. That doesn't sound like a fair representation to me, I feel like either there's proportionally a lot of UKIP voters, or a lack of labour voters.

And for certain questions, the Lib Dems are getting various percentages on such opinions around ~30% etc. Which sounds like a lot, but only 4 people picked that answer. Also, if we take this poll into account, 0% of Lib Dem voters in Clacton like Nigel Farage (that's 0/12 people!). I'm pretty sure that's not true either.

- - - Updated - - -

Also, out of the 700 people they asked, the age distribution was as follows:

18-24 - 5.9%
25-34 - 12.1%
35-44 - 8.9%
45-54 - 18.6%
55-64 - 13.9%
65+ - 40.9%

Genuine question (because obviously the 65+ range covers more people), are polls always this way?

700 is a fairly sound sample size, you can see smaller sample sizes have been used at previous by-elections such as at Eastleigh where Survation used samples of 500 early on the campaign and it pretty much mirrored the result: http://en.wikipedia.org/wiki/Eastleigh_by-election,_2013#Polling

In terms of all the rest, I am not sure whether you are looking at raw data or data after it has been weighted. For example, some may believe that older people are over-represented in polling as they are usually weighted upwards.. but there's a reason for that, and it historically has bumped the Tories higher in the polls than the raw data. It's because older people are more likely to turn out and vote compared with younger age brackets. Polling has to take into account the demographic balance in a seat (whether it has a older or younger population) but it will also take into account that fact: that older people turn out to vote in much larger numbers.

Damian Lyons Lowe, Mike Smithson of Political Betting and Betting companies are taking this poll seriously, it's pretty much standard.

[Chippiewill]

Methodology reads to be pretty sound. Only objection in my view is using telephone surveying which means the raw results will be heavily skewed towards UKIP. However the results are weighted so it's not too far off. Some of the questions were a bit iffy, and the survey as a whole is largely suspect due to over-testing, but considering the main aim of the poll was voting intention it's probably fairly accurate.

- - - Updated - - -


Also, out of the 700 people they asked, the age distribution was as follows:

18-24 - 5.9%
25-34 - 12.1%
35-44 - 8.9%
45-54 - 18.6%
55-64 - 13.9%
65+ - 40.9%

Genuine question (because obviously the 65+ range covers more people), are polls always this way?

Age distribution will be off because young people don't have landlines and aren't at home during the day.

[-:Undertaker:-]

Lord Ashcroft's Clacton poll just come out -


Lord Ashcroft @LordAshcroft · 29m

My Clacton by-election poll: UKIP 56%, CON 24%, LAB 16%, LDEM 2%, Others 2%. Full details on @ConHome, 2.30pm.


Mike Smithson @MSmithsonPB · 23m

Ukip lead in @LordAshcroft Clacton poll was 36% before reallocation of don't knows which favours Tories


Mike Smithson @MSmithsonPB · 14m

Main differences between the @LordAshcroft Clacton poll & the Survation one are methodological. Big picture same - great for Carswell/UKIP

[Chippiewill]

Just an FYI that Ashcroft's poll is much better in methodology and statistical analysis and backs up the previous poll.

Edit: Interesting tidbit from the poll is that the area are more bothered about the candidate than about nigel.

Edit2: Another fun one is that policy takes a smaller part in a decision to vote for libdems than other parties

Edit3: 56% of the UKIP vote is a protest vote

[FlyingJesus]

Good old telephone polling targeting a tiny number of old and unemployed people and claiming that it's representative of an entire population

[-:Undertaker:-]

Good old telephone polling targeting a tiny number of old and unemployed people and claiming that it's representative of an entire population

Hence... weighting!

[FlyingJesus]

Weighting ie: changing the results based on guesswork

[-:Undertaker:-]

Weighting ie: changing the results based on guesswork

And it 99% of the time turns out to be more or less accurate.

[Chippiewill]

Good old telephone polling targeting a tiny number of old and unemployed people and claiming that it's representative of an entire population

I'm not saying it's perfect, but once you have weighting into account it's accurate enough to say that a 30 point lead is statistically significant.

Also the polling occurred over the weekend so unemployed doesn't really come into it (Not that you'd generally vote for UKIP if you were unemployed anyway).

[FlyingJesus]

(Not that you'd generally vote for UKIP if you were unemployed anyway).

You say this as though the majority of people who vote UKIP actually know what their policies are other than GET THOSE DARKIES OUT :P Nationalism is rife among the unemployed because they (as prompted by UKIP posters talking about "stealing" jobs) blame immigrants rather than themselves for all their woes

But yeah polling is still stupidly unrepresentative, and weighting makes it even more so by literally making stuff up

[-:Undertaker:-]

You say this as though the majority of people who vote UKIP actually know what their policies are other than GET THOSE DARKIES OUT :P Nationalism is rife among the unemployed because they (as prompted by UKIP posters talking about "stealing" jobs) blame immigrants rather than themselves for all their woes

But yeah polling is still stupidly unrepresentative, and weighting makes it even more so by literally making stuff up

So just as I thought, you are disputing the poll purely because you don't like what the poll is saying. Happens all the time.

Interesting though that you mention darkies yet Ukip's poster campaign was aimed at white European immigration. Hmmmmm.

[FlyingJesus]

Well no but do continue making stuff up that I haven't said, you do it so well

[Chippiewill]

Interesting though that you mention darkies yet Ukip's poster campaign was aimed at white European immigration. Hmmmmm.

I'm pretty certain that was a generalisation and hardly a flaw in his argument.

[-:Undertaker:-]

Well no but do continue making stuff up that I haven't said, you do it so well

Well yes because you managed a diatribe on Ukip when we're talking about polling, not whether you like Ukip. The mask slipped there Tom. At the end of the day the pollsters keep getting it right time after time and make a lot of money for it, and you're just a guy with a keyboard without a clue.

- - - Updated - - -


I'm pretty certain that was a generalisation and hardly a flaw in his argument.

Oh no, that was a huge gaping hole. Just as if Tom would look at more polling he'd find attitudes towards mass immigration aren't based on skin colour, making his assumption completely wrong.

[FlyingJesus]

What is it that makes you so bad at reading Dan? The part about UKIP was quite clearly a response to what Jin said about the unemployed voting them, as you should have been able to work out quite easily by looking at what I'd quoted. You're really extremely bad at this

[-:Undertaker:-]

What is it that makes you so bad at reading Dan? The part about UKIP was quite clearly a response to what Jin said about the unemployed voting them, as you should have been able to work out quite easily by looking at what I'd quoted. You're really extremely bad at this

It's no wonder you dispute polling because it disproves virtually every assumption you've made in this thread.

I recommend a read of an academic study of Ukip's voters (http://www.amazon.co.uk/Revolt-Right-Explaining-Extremism-Democracy/dp/0415661501/ref=sr_1_1?s=books&ie=UTF8&qid=1409672298&sr=1-1&keywords=revolt+on+the+right), backed up by polling and profiling.


The UK Independence Party (UKIP) is the most significant new party in British politics for a generation. In recent years UKIP and their charismatic leader Nigel Farage have captivated British politics, media and voters. Yet both the party and the roots of its support remain poorly understood. Where has this political revolt come from? Who is supporting them, and why? How are UKIP attempting to win over voters? And how far can their insurgency against the main parties go? Drawing on a wealth of new data – from surveys of UKIP voters to extensive interviews with party insiders – in this book prominent political scientists Robert Ford and Matthew Goodwin put UKIP's revolt under the microscope and show how many conventional wisdoms about the party and the radical right are wrong. Along the way they provide unprecedented insight into this new revolt, and deliver some crucial messages for those with an interest in the state of British politics, the radical right in Europe and political behaviour more generally.

[FlyingJesus]

Well of course you'd recommend looking at something that's got nothing to do with what I said, that's your entire debate style

[-:Undertaker:-]

Well of course you'd recommend looking at something that's got nothing to do with what I said, that's your entire debate style

Oh no, it applies completely to you.


- in this book prominent political scientists Robert Ford and Matthew Goodwin put UKIP's revolt under the microscope and show how many conventional wisdoms about the party and the radical right are wrong.

You are wrong and the scientific polling (which is accurate, no matter what you say) says you are wrong.

[FlyingJesus]

Wrong about what? My claim was that unemployed people are more likely to have grievances against immigrant workers than those who are already securely in the workforce, that's got nothing to do with UKIP methods. Do at least ATTEMPT to keep up

[-:Undertaker:-]

Wrong about what? My claim was that unemployed people are more likely to have grievances against immigrant workers than those who are already securely in the workforce, that's got nothing to do with UKIP methods. Do at least ATTEMPT to keep up

"dey h8 darkies"
"its nationaliszm lyke 1930s"
"all unemployed chavs voting ukip innit"

All unfounded and lazy assumptions that aren't backed up by the studies and polling of the Ukip voter base.

Discussions of a voter demographic/polling can be interesting when the person you're talking to actually has a clue, but this is very boring.

[The Don]

"dey h8 darkies"
"its nationaliszm lyke 1930s"
"all unemployed chavs voting ukip innit"

All unfounded and lazy assumptions that aren't backed up by the studies and polling of the Ukip voter base.

Discussions of a voter demographic/polling can be interesting when the person you're talking to actually has a clue, but this is very boring.

Hardly unfounded, Dan.

[FlyingJesus]

Again with things I haven't said, well done. Did you use weighting to come up with that?


ps 700 people from the Clacton constituency amounts to a little over 1% of the electorate. It's like asking 2 people who have been on HxF today what their views on cats are (making sure to pick only the people whose answers are likely to fit what you want to see) and claiming that it's an accurate representation of the whole lot of us

[-:Undertaker:-]

Hardly unfounded, Dan.

Yes they are.

How do I know? Because i've looked into it *taps book full of polls, surveys and research* rather than making lofty Guardian-like assumptions.

[FlyingJesus]

Making assumptions is what this thread is entirely about lmao

[-:Undertaker:-]

Again with things I haven't said, well done. Did you use weighting to come up with that?


ps 700 people from the Clacton constituency amounts to a little over 1% of the electorate. It's like asking 2 people who have been on HxF today what their views on cats are (making sure to pick only the people whose answers are likely to fit what you want to see) and claiming that it's an accurate representation of the whole lot of us

The pollsters are rich because they get it right.

I would supply links to previous by-elections and elections where polling got it more or less right, but you're set in your ways so what's the point?

[The Don]

Yes they are.

How do I know? Because i've looked into it *taps book full of polls, surveys and research* rather than making lofty Guardian-like assumptions.

So it's unfounded to claim that the unemployed are more likely to have an issue with immigration than those in employment?

[-:Undertaker:-]

Making assumptions is what this thread is entirely about lmao

An assumption is made without proof, a poll of voter intentions is not something that is without any proof.

Hence why political parties pay very special attention to them and carry out their own secret polls throughout campaigning. They matter hugely.

[FlyingJesus]

Yeah what's the point in sourcing things and having a proper discussion about what's been written when you can just make stuff up about your opponent and derail

Proof of 1% of the electorate well done

[-:Undertaker:-]

Yeah what's the point in sourcing things and having a proper discussion about what's been written when you can just make stuff up about your opponent and derail

Proof of 1% of the electorate well done

We'll see who is right come election day: me, Chippiewill, the pollsters, politics wonks and the betting markets - or you.

[FlyingJesus]

It's not an either/or situation though since I haven't refuted that this may happen, just that the methodology is stupid

[Kardan]

So do you think UKIP will have a 44% lead over the Tories in the by-election?

[-:Undertaker:-]

It's not an either/or situation though since I haven't refuted that this may happen, just that the methodology is stupid

You'd be able to call it stupid if their methodology got it wrong all of the time, but the opposite is true. It's highly accurate.

I'm sure the very rich and experienced pollsters will sleep well though despite you calling their methods 'stupid'. :P


So do you think UKIP will have a 44% lead over the Tories in the by-election?

If it were being held today/tomorrow or within this week, I would put a large bet on that happening. Yes absolutely.

[FlyingJesus]

If you ever actually cared to discuss why it's a good system (despite asking 1% of an electorate what they think and targeting specific areas of that electorate through their methods) rather than just repeating IT WORKS I PROMISE THEY'RE RIGHT that would be superb

[Kardan]

You'd be able to call it stupid if their methodology got it wrong all of the time, but the opposite is true. It's highly accurate.

I'm sure the very rich and experienced pollsters will sleep well though despite you calling their methods 'stupid'. :P



If it were being held today/tomorrow or within this week, I would put a large bet on that happening. Yes absolutely.

But it's not :P So you agree with me, that you don't think they will get a 44% lead by the time the election happens? What do you reckon it will be? I'd hazard a guess at maybe 20/25.

[-:Undertaker:-]

If you ever actually cared to discuss why it's a good system (despite asking 1% of an electorate what they think and targeting specific areas of that electorate through their methods) rather than just repeating IT WORKS I PROMISE THEY'RE RIGHT that would be superb

It's a good system exactly because it does work. It's as simple as that, really.

- - - Updated - - -


But it's not :P So you agree with me, that you don't think they will get a 44% lead by the time the election happens? What do you reckon it will be? I'd hazard a guess at maybe 20/25.

I haven't a clue, polling will tell us nearer the time. Mike Smithson of Political Betting made a point the other day that the Ukip lead could potentially increase as the campaign goes on as a lot of loyal Tories once they see the polls that Ukip are ahead may decide to jump ship too.

The lead could also increase if loyal Tories/Labour voters see the polls and decide there's really no point in even bothering to vote.

But then again it could all go **** up for the purples, after all a week is a long time in politics. :P

[FlyingJesus]

It's not as simple as that, that's the entire point :P you're once again using a "just because" argument, even though you're quite openly admitting that these things are likely to change dramatically. I could release a guess of my own saying BUT IF IT HAPPENED NOW LIB DEMS WOULD GET 504% OF THE VOTES BECAUSE I DREAMT IT and then nearer the time say something completely different and claim that all of my statements have been scientifically proven to be true

[-:Undertaker:-]

It's not as simple as that, that's the entire point :P you're once again using a "just because" argument, even though you're quite openly admitting that these things are likely to change dramatically. I could release a guess of my own saying BUT IF IT HAPPENED NOW LIB DEMS WOULD GET 504% OF THE VOTES BECAUSE I DREAMT IT and then nearer the time say something completely different and claim that all of my statements have been scientifically proven to be true

Because opinion polls are a snap shot of public opinion at any given time, an election of course has so many scenarios and variables that there is huge scope change to happen in voter opinion. That said, in some elections the polls barely shift from the start of the campaign to the end (http://en.wikipedia.org/wiki/Nationwide_opinion_polling_for_the_United_States_p residential_election,_2012#Two-way_race).

[GommeInc]

That large percentage of people caring about immigration is weird. Usually the answer is "cost of living... because of..." or "low availability of jobs... because of..." Not "immigration is an issue", unless these people are a picnic hamper short of a full picnic. The poll(s) seem unreliable because of that :/

Clacton is a strange area. I wouldn't be surprised if he is elected in or a Conservative. It's extremely unlikely it will be a Labour, Green or Lib Dem candidate as this area is strictly Conservative with some UKIP values. If he wins it's not necessarily because he's a member of UKIP, but because he did a decent job (as far as I recall). The vote really is about whether or not people liked what he did, rather than party persuasion.

[Chippiewill]

ps 700 people from the Clacton constituency amounts to a little over 1% of the electorate. It's like asking 2 people who have been on HxF today what their views on cats are (making sure to pick only the people whose answers are likely to fit what you want to see) and claiming that it's an accurate representation of the whole lot of us

No. No it's not. That is not the way statistics work. Except when the population size is small the ratio between sample size and population size is largely irrelevant.

[FlyingJesus]

You're telling me 1% is not like 1%

[Chippiewill]

I'm telling you 1% is irrelevant.

[FlyingJesus]

Thank you for agreeing with me