Political polling

[butchersapron]

polling from two more top Tory target seats - and they're nowhere.
http://www.theguardian.com/politics/2013/dec/03/tories-behind-labour-level-with-ukip-key-marginals

Note, not a single one of the lib-dem voters from 2010 in Dudley North (they got around 4000 votes/10.5%) said in this poll that they intend to vote lib-dem this time around - not one. The poll is relatively small - 526 people, but 10% of that is still 50 people. And not one are now lib-dem as things stand today. They are all labour or DKs.

(Spotted by Mike Smithson @MSmithsonPB)

edit: how annoying, on checking the tables, it seems there were only 11 2010 lib-dem voters in the sample.

 

[where to]

still an amazing stat.

 

[Quartz]

2015 election will be a psephologist's dream, if not the electorate's.

The one to look at will be next year's EU elections.

not looking even unlikely imo - look at that poll above, he's in touching distance. a high profile campaign (all Ukip's chunk of the media share will be on one candidate and one seat) against a no-hoper and i think he'll walk it. cant see them dropping back in polls between now and then either.

Could he defeat the Speaker? AFAIAA he's not held in very high regard around the country, but I don't know about within his constituency.

 

[where to]

he contested Bercow's seat in 2010. got a bloody nose.

 

[Spanky Longhorn]

The one to look at will be next year's EU elections.



Could he defeat the Speaker? AFAIAA he's not held in very high regard around the country, but I don't know about within his constituency.

Is there no limit to what you don't know?

 

[butchersapron]

Huge 12% labour lead with yg today, 41/29. Work done before the autumn statement, but early polling on that indicates balls in winning there too.

 

[brogdale]

Sunday's....

There were three voting intention polls in this morning’s papers, topline figures are below:

Opinium/Observer :- CON 30%(+2), LAB 37%(+2), LDEM 8%(nc), UKIP 16%(-3)

ComRes/Indy on Sunday :– CON 29%(nc), LAB 36%(+1), LDEM 8%(-2), UKIP 18%(+1)

YouGov/Sunday Times :– CON 32%, LAB 38%, LDEM 9%, UKIP 13%

Tabs for ComRes are here, tabs for YouGov are here.

Tories still polling around and about 'core' support level.

 

[butchersapron]

4 more of these Brown funded UKIP/marginal polls to be published tonight.

 

[butchersapron]

Here we go.

Tories in serious trouble.

Folkestone & Hythe, previously considered a safe Conservative seat with a 19 point lead over the Liberal Democrats at the last election, now shows a significant drop in the vote shares for both coalition partners, with Labour and particularly UKIP as the main beneficiaries.

Bognor Regis & Littlehampton has traditionally been an even more secure seat for the Conservatives, with a 27 point majority in 2010. Whilst the swings shown here are not quite as large as those seen in Folkestone & Hythe, they still show a significant drop in Conservative and Liberal Democrats support here and a large boost for UKIP who have nearly quadrupled their already significant vote share of 7% in 2010.

Great Yarmouth has been a key swing seat since the 1980s, which the Conservatives regained from Labour in 2010. Our polling shows that not only is Labour forecast to regain this seat in 2015, but the Conservatives are in danger of slipping from first to third place, behind UKIP who have surged from 5% to 30%

If only to further illustrate the point that UKIP’s impact on the Conservative – Labour race is not as significant as many have claimed, one need only compare the results in Great Yarmouth above with the poll results in another key marginal seat, Crewe & Nantwich. The 2010 results here were remarkably similar to Great Yarmouth (only 7 points difference across all political parties), but by contrast this is a seat in which UKIP is performing significantly less well, on only 11 points. But the lower UKIP vote share does not seem to have helped the Conservatives at all – in fact the Conservative to Labour swing shown here is three points higher than in Great Yarmouth.

 

[brogdale]

4 more of these Brown funded UKIP/marginal polls to be published tonight.

Whereabouts are they published?

This graphic, based upon polling, has been causing some sort of twatter stir apparently...

First 4 columns read like one of those make your own Mail headline things.

e2a : thanks butchers...you've answered my q

 

[killer b]

Whereabouts are they published?

This graphic, based upon polling, has been causing some sort of twatter stir apparently...

First 4 columns read like one of those make your own Mail headline things.

e2a : thanks butchers...you've answered my q

there's something I don't like about those graphics. they're basically saying 'look how stupid & racist you/they are' - actually, i'd probably have some up with similar numbers if asked with no prompting too.

 

[brogdale]

Remarkable stuff in those marginals, and yes the tories are in trouble.

Here we go.

Tories in serious trouble.

Yep, and significant evidence that neither UKIP's success or failure holds such electoral fears for Lab...

We have now seen the publication of seven constituency polls. In five of these seven we included the question:


“Which of the following statements is closest to your opinion?”

- “I would rather vote UKIP than Conservative, even if that means Ed Miliband becomes Prime Minister”

- “I would rather stop Ed Miliband from becoming Prime Minister, even if that means I had to vote Conservative rather than UKIP”

This question was asked to UKIP voters, and the aggregate results from the five polls where it was asked are shown in the chart to the left. Most UKIP voters would prefer to stick with their UKIP vote even if this made Ed Miliband Prime Minister. Because this chart includes two ‘safe’ conservative seats (Bognor & Littlehampton and Folkestone & Hythe) as well as three ‘marginal’ seats (South Thanet, Great Yarmouth and Crewe & Nantwich), it understates the proportion of people who would prefer voting UKIP to stopping Miliband across just the marginals. In Great Yarmouth, for instance, the figures were 68% who would rather vote UKIP to 23% who would rather stop Ed Miliband.

In the most recent four polls (South Thanet, Bognor Regis & Littlehampton, Crewe & Nantwich and Folkestone & Hythe), we also asked the question to UKIP voters:

“Imagine for a moment that there was no UKIP candidate standing in your [Folkestone & Hythe] constituency.


Which of the other parties would you then be most likely to vote for in the next election, or would you not vote?”

The results can be seen in the chart to the right. The net gain to the Conservatives over Labour if UKIP were removed from the equation can be seen to be 13% of the UKIP vote (or no more than 2 points nationally if UKIP is taken to be around 18%).

In marginals again the picture is tighter, with the results in South Thanet for instance showing a net gain to the Conservatives of only 2% of the UKIP vote if UKIP were not standing – for a total gain of only 0.6% over Labour. The fact that the most popular response to this question was “would not vote” suggests that UKIP voters have developed their own sense of identity with the party and are no longer “in the market” for other political parties to try and win them over.

Interesting stuff; they'll be shitting themselves at tory hq

 

[brogdale]

there's something I don't like about those graphics. they're basically saying 'look how stupid & racist you/they are' - actually, i'd probably have some up with similar numbers if asked with no prompting too.

Maybe, but that begs the question about how it is that large numbers of the population are so mistaken.

 

[Santino]

Maybe, but that begs the question about how it is that large numbers of the population are so mistaken.

People are incredibly bad at estimating quantities. Guesses can be thrown off by tiny, apparently unrelated facts about how the question is asked. For example, you can influence the value someone will give by generating a random number in front of them and asking them to think about how much higher or lower than that number their answer is.

 

[killer b]

Maybe, but that begs the question about how it is that large numbers of the population are so mistaken.

well, they aren't mistaken are they? they're being asked to take a stab at a figure that most of them have no real idea what it actually is. so they'll mainly think 'well, it's less than 50%. dunno, about 30?' hence why it's hovering at around 30% for all of them.

I'm of the opinion they have no purpose except to make people who read the guardian feel a bit more right. but I bet they'd all guess at around 30% too.

 

[killer b]

also what santino said.

 

[taffboy gwyrdd]

That's not just Daily Mail headlines, it has to be raw stupidity to seriously think that nearly 1/4 of the entire population is Muslim. A person might live in an area where they see quite a few (apparent) muslims round the place, but what happened to the ability to consider that other significantly less mixed places exist in large amount? It's baffling.[/quote]

 

[killer b]

my point, illustrated. raw stupidity.

 

[Santino]

That's not just Daily Mail headlines, it has to be raw stupidity to seriously think that nearly 1/4 of the entire population is Muslim. A person might live in an area where they see quite a few (apparent) muslims round the place, but what happened to the ability to consider that other significantly less mixed places exist in large amount? It's baffling.

[/quote]
Most people atrocious at dealing with large numbers.

 

[taffboy gwyrdd]

Interesting stuff; they'll be shitting themselves at tory hq

For sure. They're fucked at least 2 ways. First, in that no (major) governing party I recall has won more seats after 1 term than it did to get in. Second, they'll be caught between tacking right to UKIP with increasingly ugly politics and the fact that such politics alienates the relative centre where elections are said to be won, and marginals certainly are.

 

[brogdale]

People are incredibly bad at estimating quantities. Guesses can be thrown off by tiny, apparently unrelated facts about how the question is asked. For example, you can influence the value someone will give by generating a random number in front of them and asking them to think about how much higher or lower than that number their answer is.

Yes. I'm not sure what methodology lies behind the Ipsos polling, but I think that I read on Political betting that multiple choice options were involved, so that could potentially produce such an effect.

 

[taffboy gwyrdd]

well, they aren't mistaken are they? they're being asked to take a stab at a figure that most of them have no real idea what it actually is. so they'll mainly think 'well, it's less than 50%. dunno, about 30?' hence why it's hovering at around 30% for all of them.

I'm of the opinion they have no purpose except to make people who read the guardian feel a bit more right. but I bet they'd all guess at around 30% too.

What if they just knew the answer as a matter of general knowledge? It's not advanced or secret knowledge.

 

[Santino]

Yes. I'm not sure what methodology lies behind the Ipsos polling, but I think that I read on Political betting that multiple choice options were involved, so that could potentially produce such an effect.

I would hazard a guess that asking 'how many people in a hundred' would also produced a different result from asking 'what percentage', simply because the word 'hundred' acts as an anchor.

 

[brogdale]

well, they aren't mistaken are they? they're being asked to take a stab at a figure that most of them have no real idea what it actually is. so they'll mainly think 'well, it's less than 50%. dunno, about 30?' hence why it's hovering at around 30% for all of them.

I'm of the opinion they have no purpose except to make people who read the guardian feel a bit more right. but I bet they'd all guess at around 30% too.

Well....they are mistaken in their estimations. Certainly for the first 5 columns; for the last one I would suggest that the Ipsos polling might be nearer the mark than the 'reality' figure.

 

[killer b]

What if they just knew the answer as a matter of general knowledge? It's not advanced or secret knowledge.

so what? most people don't memorise this shit, so it'll just be a vague stab. hence 30%.

 

[Santino]

Or (b)

 

[Delroy Booth]

Yeah I would like to have a look at how they asked that question. All that graph shows to me for is that when questioned to give precise numbers for demographics people will fudge it by playing it safe, which then averages out at "20-something" for nearly everything.

You don't need statistical evidence as flimsy as that to show how the media influence people's perceptions

 

[killer b]

Well....they are mistaken in their estimations. Certainly for the first 5 columns; for the last one I would suggest that the Ipsos polling might be nearer the mark than the 'reality' figure.

of course they're mistaken. I don't think there's necessarily any sinister reason for them being mistaken though. it's just how people deal with figures, rather than us all being racists.

 

[taffboy gwyrdd]

so what? most people don't memorise this shit, so it'll just be a vague stab. hence 30%.

Ok, so they take a vague stab that nearly a third of people in the place are muslim? Come off it, that's pretty fucking weird.

 

[taffboy gwyrdd]

of course they're mistaken. I don't think there's necessarily any sinister reason for them being mistaken though. it's just how people deal with figures, rather than us all being racists.

I don't think it has to be about race. But it can be about being wantonly misinformed via right wing and establishment media, which is sinister at the media level, if not the consumer level.