Plausible hypocrisy
To the ASTONISHMENT OF ALL, the Scottish media has leapt to cover a new poll today. It was conducted on behalf of the cross-party “Devo Plus” group, which we were mildly surprised to discover apparently still exists despite the previous two posts on its website being dated February 2013 and November 2012.
Being far more fair-minded than other news outlets, however, and not ones for bearing petty grudges, Wings Over Scotland is more than happy to run some analysis on it.
Here are a few bullet points that jumped out.
1. The first thing to note is that the poll shows support for independence steady compared to the previous one from the same polling company, and opposition slightly up. Applying the standard margin of error to both polls in fact gives no reliable evidence of movement at all.
That’s not our opinion, but that of media-favourite psephologist Prof. John Curtice:
2. The Sunday Herald claims the 29% pro-independence figure as a “Record poll low for Yes vote”. Which is curious, as in January of this year the Herald was running the headline “Support for independence falls back to 23%”. It’s Sunday morning and we’re still a bit sleepy, but pending independent verification we’re fairly confident in saying that 23% is lower than 29%.
3. The Herald also notes that “Amongst SNP voters, 10% said they would reject the party’s key policy and vote No. A further 8% did not know which box they would tick in the referendum ballot.”
As far as we can tell in the absence of full data tables, this is a highly misleading representation of the actual findings – Devo Plus itself says “One-in-ten SNP voters would be more likely to vote ‘no’ if they got Devo Plus“. (Our emphases.) That’s a highly-conditional “maybe”, presented by the Herald as a definite “No”.
4. The poll also found that just 6.6% of voters would even consider changing their vote from Yes or Don’t Know to No if they believed the Devo Plus model would be adopted as a result, which we imagine the group found a little disappointing.
5. The results confirmed what other polls have found consistently, despite a concerted Labour and media campaign to present pro-independence supporters in the party as a tiny minority – 15% of Labour voters are currently planning to vote Yes, with another 10% undecided.
(Incidentally, as far as we know YouGov still attributes its party affiliations by Westminster elections, so that’s 15% of the 1m people who voted Labour in 2010, not the 500,000 or so who backed the party for Holyrood in 2011.)
And since we’re quite sure that’s the most anyone’s ever written about the wholly-irrelevant Devo Plus – which isn’t backed by any party and has no hope whatsoever of seeing its proposals become policy – since it launched 18 months ago and promptly vanished without trace, we’ll leave it at that.
We’re sure there’s a really good reason why a poll commissioned by a partisan organisation about things which aren’t being offered to the Scottish people and have absolutely no chance of happening is all over the papers, while a different one which concerned itself with issues directly relating to the actual referendum options (and the independence debate itself) got comprehensively blanked.
We’ll get back to you just as soon as we remember what it is.
Dinna worry theres another poll coming oot soon … a few o us hiv done it. It didna hae a question ‘should we aa stop eating Tunnocks Tea cakes’ so I dinna think its yours……
Yup, the sainted Sunday Herald caught lying.
But it’s worrying that some Yes supporters are primed to lend more credence to this partisan – but ‘official’ – poll than to the WoS Panelbase one.
Let’s see:
10% [of SNP voters] said they would reject the party’s key policy and vote No. A further 8% did not know which box they would tick in the referendum ballot.”
Now the last poll had the SNP on 48%, but lets be conservative and give them 45 assuming no change from May 2011 as polls agree on.
So, 100-10-8 = 82.
0.82*45 = 37%
Hmmm.
15% of Labour are Yes
0.15*30 = 4.5%
37+4.5 = 41.5%
And then there’s some libs, greens, independents to add.
Weird huh?
😉
Methodology’s a funny thing….
From what I see one of the reasons is that What Scotland Thinks added it to their graph on the asking of the referendum question. And probably sent out a press release.
If that is part of it, and you have it in the next poll, then they can’t argue that’s why they ignore it, if they do.
As I understand it’s BBC policy to cover all polls that include the referendum question, we’ll see.
There is an old saying that the Herald would do well to take on board. “No man has a good enough memory to be a successful liar”.
Remember, Devo Plus is heavily supported by Tory think tank ‘Reform Scotland’ (or in the words of Reporting Scotland, an ‘independent think-tank’)
Quite how this bunch managed to wangle charity status for their organisation when that status does not allow a political campaigning is frankly a disgrace.
All recent polls have demonstrated a swing towards YES.
The above mentioned poll was instigated by Devo Plus. Another unionist exercise in deception. Vote NO and we’ll get substantial tax powers!
Hardly surprising that the MSM has dived into it like flys to shite. Shite is what it is and should be regarded as such.
Incidentally, where are the VI numbers?
Clearly, this question was asked.
Maybe the results were not great for pro-union parties…
The WOS-comissioned poll needs more coverage. Do you have a single article which summarises the results, which would make it easier to link to and share?
And it’s a YouGov poll…
link to scotgoespop.blogspot.co.uk
Which in Scotland these days isn’t worth much since they got the Scottish election result massively wrong – by 16% in terms of how the regional vote for the SNP would be.
That’s the vote that determines how proportionate the Parliament would be.
They had an SNP lead of 2% (well within those insignificant margin of errors John C mentions). At the end of the day it was 18%.
“What perhaps might have been more effective at persuading him to do so was evidence that large number of No supporters might switch to Yes if they were to come to the conclusion that the unionist parties would fail to satisfy their aspirations for more devolution, thereby taking a large chunk out of the No lead. Alas for them, this seems to have been the vital question that Devo Plus forgot to ask.”
link to blog.whatscotlandthinks.org
Ask this question on your next poll. And be sure not to include any silly space monsters nonsense. Make it a proper professional job so the media have no excuses for ignoring it.
Clearly, the Herald have missed the No campaign’s recent call for a balanced debate:
link to bbc.scotlandshire.co.uk
Handshrimp
Yeah, you are right,the downfall of most habitual liars,is just what my father always said ” to be a good liar, you need to have a very good memory”
Something BT never takes into contention, something that the REV, usually throws back at them. No answer back of course,just a Hmmmmmm !
Panelbase also carried out an Indy last week (23-27 Aug), using the straight question. I assumed that it would be the monthly Sunday Times one, but there’s no sign of it today!
Maybe that had some interesting results, or got conveniently dropped since this YouGov poll gives juicy ammo for the No Scotland headline writers. I’d be interested to see the results from that.
Also there was a Panelbase one yesterday (running 31 Aug – 12 Sept) asking who you voted for in the holyrood and Westminster elections, then the indy question, then ended with did the fact the Lib Dems went into coalition with the Tories make you more or less likely to vote for them (sliding scale 1-10). Another one I’d like to see the indy question results for.
It would be unbelievable in a fair & balanced debate to see a poll getting MSM
coverage whilst one they don’t like gets almost zero coverage.
But this is neither a fair nor balanced debate.
Can’t wait for the next Wings poll!
test
Rev.Stu – re your tweet as a result of Mail article lots of old school friends getting in touch – ever heard the saying “if you fell in a barrel of shite you would come out smelling of roses” well I think that applies to you. I think DH should call it a day I am really pleased it has back fired on them.
Finding it hard to figure out who is asked to take part in a poll other than those who have registered an interest to, with the polling company used for the poll. Does anyone know if that is the case?
@Kevin Lynch
Ask this question on your next poll. And be sure not to include any silly space monsters nonsense. Make it a proper professional job so the media have no excuses for ignoring it.
Kevin – that was Prof Curtice’s explanation. There was silence from the MSM on the WoS poll as there will be on the next one. These polls are for our information and we can then spread the message by the internet and word of mouth.
I see the usual suspects have leapt on to The Herald forums ranting along the lines that its all over for Salmond’s vanity project.
In the not so distant past, I recall pretty much the same from Grahamski and others on The Scotsman forums… Oh, around May 2011 it was.
@ Atypical_Scot
I have been a YouGov subscriber for over 3 years and I have NEVER been selected to complete a Scottish political poll, (have been selected for many others) beginning to think it’s because I completed profile honestly
@cirsium
Agreed ! There should be absolutely no reason that we have to lend currency to anything Curtice says by seeking his approval of our crowd sourced polling.
The poll is for our edification in order that we have clear visibility and interpretation of the publics intentions.
The fact that we can get a bit bent out of shape when the MSM fail to mention our positive indy poll should come as no surprise to anyone…and frankly, the reason I became an avid indy supporter was due to the fact that along with the UK gov, the MSM DON’T reflect my views, and constantly spin the unionist line.
link to tinyurl.com (pls forward to as many people as you can….Wingslanders Rool ! 🙂
Dunphail says:
1 September, 2013 at 1:39 pm
@ Atypical_Scot
I have been a YouGov subscriber for over 3 years and I have NEVER been selected to complete a Scottish political poll, (have been selected for many others) beginning to think it’s because I completed profile honestly
Ditto Dunphail.
@Dunphail;
I don’t know if I’m right here, but I’m thinking that the outcome of polls are subject to a bias – those who make the decision to register interest in taking part in polls.
I have recently registered interest to do so. But I can’t help but wonder if the number of Yes voters there actually are, is reflected in the number of registered pollsters?
What is the political leanings of the majority of registered pollsters? I only ask because when I registered an interest to take part, I felt a sense of adherence to a system I usually avoid – as in the current UK collective sense, as most polling companies ask UK wide surveys. I’ve always liked the stats, but never considered ‘taking part’ which, in a way is effectively casting a vote if selected.
What if every Yes voter registered interest in taking part, would that affect the poll outcome?
scottish_skier
The reason that it looks as if the numbers don’t add up is because the percentages you are using are those of the people who gave a voting intention (presumably for Holyrood), not of all people who replied to the survey. Since VI Don’t Knows aren’t included – and they’re more likely to DK on the referendum too – both Yes and No votes will make up a higher percentage of those with a VI than of the population in general.
It should be clearer when we get the tables which I would expect to see on the YouGov Archive some time tomorrow morning.
What is Devo-Plus anyway? Any specifics on that, if you’re reading this Reform Scotland? Also any concrete/cast iron/titanium commitments in writing from the Unionist parties that they will actually do what they are hinting that they’ll promise?
It might change my vote you know, if they’re to be trusted that is, and if I don’t suspect that this is just a self-serving and dishonest attempt to hoodwink the Scottish people into voting no with every intention that we will be roughly shafted for all eternity for foolishly putting our trust in the spoken or written word of the Westminster parties and their MSM lickspittles.
No, thought not.
YouGov the pollster of choice for unionists.
YouGov is definitely an outlier because all the other pollsters not only show convergence but have consistently shown the No vote under 50% recently.
Dramfineday and Dunphail
Not odd at all. By my reckoning there are about 30,000 YouGov panel members in Scotland and they usually only need 1,000 people for each poll. There’s only been two YouGov Scottish polls that asked VI in the last year (plus one for Ashcroft that only asked re the referendum). So you would only expect to be asked every 10-15 years. You’ve probably been asked UK-wide polls though because they do a lot more of those, even though they only need 150 Scots each time.
It’s because of the very small number of Scottish polls that get commissioned. No doubt because your media are impoverished and incurious (unlike your bloggers).
Like a pack of snarling rabid dogs we witnessed Scotlands hacks
Rounding on the monkey who has climbed upon their backs
Their foaming mouths spilled out their bile, they know no other way
To try to stop us hearing, what the good Rev has to say
Each day he questions what they write and this they do not like
For they all know that many more will say, go take a hike
They’ve compromised their honesty and their journalistic skills
And when they sit with heads in hands and have to pay the bills
I wonder if they’ll have a wish their time to have once more
Or if they’ll just be happy to fade into folklore
The Rev’s become the itch that their scratching wont relieve
And the people have decided who they choose now to believe
@Gavin: The WOS-comissioned poll needs more coverage. Do you have a single article which summarises the results, which would make it easier to link to and share?
Full results, see link to panelbase.com
Also:
link to wingsoverscotland.com
link to wingsoverscotland.com (voting intentions compared to last vote in WGE 2010)
link to wingsoverscotland.com (powers devolved in event of no vote)
Stuff devo+ and just go for full independence which read as devo max + no interference from Westminster.
If Devo+ was on offer how would you vote in the referendum?
Was that the question posed?
I for the life of me cannot imagine how anyone could get up on the 18th September 2014 and think, I have waited all my life for this day, the day I vote for, eh, continued Westminster destruction of Scotland. I’m sure my grandkids will thank me one day.
Atypical_Scot says:
I don’t know if I’m right here, but I’m thinking that the outcome of polls are subject to a bias – those who make the decision to register interest in taking part in polls.
I have recently registered interest to do so. But I can’t help but wonder if the number of Yes voters there actually are, is reflected in the number of registered pollsters?
What is the political leanings of the majority of registered pollsters?
All pollsters weight for demographics (gender, age, class etc). Most weight for political bias (though MORI and Opinium don’t) either based on how people voted at the last election or which Political Party they identify with. There’s a good article on weighting on Anthony Wells’ UKPR site.
Online panels also use the political data they have on their members so they invite people to participate in proportion to the profile they are matching too so as to keep weighting as small as possible. (Telephone and face-to-face polls use random phone numbers or addresses). All this means that even if there is a difference between the make-up of the panel and of the population as a whole this should be corrected for.
There could be a problem with online panels when it comes to referendums (there may have been with AV) which leads to overestimating the Yes vote, because Yes voters may be more politically aware and so more likely to be panel members. But I haven’t seen a sign of it in the Independence polls this year – telephone polls (MORI) and face to face (TNS) seem to produce similar results to online ones.
There is however one online pollster (Panelbase) producing different figures form the others which I think may be a related glitch. When Panelbase invite you to take part in a poll they state what the subject of the poll is before you start (‘Cars’, ‘Shopping’ or whatever). It could be that No voters and especially the uninvolved are more likely to see ‘Politics’ and not bother to reply to that survey. Some other pollsters tend to ‘hide’ political questions in bigger consumer surveys and this may mean that such people are more likely to take part.
ScotgoesPop has a good take on this poll – link to tinyurl.com
Could it be the BT see a trend moving towards the YES Camp, and have decided to spout the Devo Max Plan, in order to reel some voters back in.
Even though theirs not a snowballs chance in hell, of anymore significant powers
being devolved to Holyrood…Hmmm! stinks of “Jam Tomorrow” and 1979.
The danger is that the unionist parties will announce a devo plus policy just before the referendum ,that policy wont be specific of course and will never be implemented or only on meaningless powers .
TheYes caMpaign have to be ready with the answer to or exposure of their deceit
In my opinion as the offer of Devo ‘anything’ is off the table never to return before 18th September 2014 then we should totally ignore anyone who claims to come from a Devo ‘whatever’ think tank.
The offer of Devo ‘whatever’ was forced off the table by the UNIONISTS, they therefore can not carry on with the lies that Devo ‘whatever’ is still an option that will be available to the electorate. it will not, get over it Better Together and ‘friends’ and move on!
I believe the time has come for everyone on the pro independence side to really pull Better Together and ‘friends’ over hot coals for their constant lying about the possibility of Devo ‘anything.’ Better Together constantly come out with ‘promises’ of more powers but amazingly they are well short of any detail, this wouldn’t be because the majority of Scots believe there will be NO more powers for Scotland but a removal of existing powers would it?
No matter what anyone says we all know there are no Devo ‘whatever’ powers sitting in the wings and it is up to us to spread the word to everyone we meet that this is the case. The referendum is pure and simple ‘do you want Independence’ or ‘do you want the status quo?’
@Dunphail
Me too, while I was a member. I gave up in disgust after several too many commercial surveys without a ‘none of the above’ or equivalent option meaning I could not honestly complete it. Entered into correspondence on the matter, got assurance would do better, did they shite.
The reason that it looks as if the numbers don’t add up is because the percentages you are using are those of the people who gave a voting intention (presumably for Holyrood), not of all people who replied to the survey.
Thanks but I know this.
If you know how Yougov weight in Scotland I’d be interested. They don’t match any other poll being a complete outlier from 2007 on. Only MORI has come close and it’s largely out on its own too. What’s weird is the other online polls (ICM, AR, panelbase) all match each other well. Then, as expected, non-anonymous / landline approach TNS/BMRB show higher Yes and lower No. Only then at the far reaches of plausibility do you get the lonely BT stall Yougov results.
@Roger Mexico ;
I’ve registered with Yougov and been asked to complete a survey with no prior screening. They know nothing about me but my e-mail address.
The Devoplus people did not ask the question that is on the ballot paper next year. On twitter someone has posted the questions:
”Scotland should become a fully independent country seperate from the rest of the UK”
“Scotland should remain part of the UK WITH THE INCREASED POWERS OUTLINED IN THE DEVO PLUS PROPOSAL”
“Scotland should remain part of the UK with the same devolved powers it has at present”
– wait till the 41% who picked no 2 realise that what is being proposed is devo max +
I would be interested to know how many Yes signatories have registered with a polling co. for example, or maybe how many Wings readers Rev?
The fact that Prof Curtice was quick to back major elements of this ‘rogue poll’ stinks in my view. Major thrust points of note before he mentions very few caveats in his comments; very BBC to me.
It is just not credible in light of recent polls and events that the opposing sides are this far apart at this stage as suggested. No way; something is seriously tainted here. Curtice’s ready availability for comment & involvement smacks of partisan representation on this one. Look how long it took for Panelbase. This Devo + group is just another clever smoke & mirrors, save the Union – stalling tactic splinter group. Checkout their website (very limited) There will be no new powers people! Salmond tried to get this on ballot paper .. the No side said no; what does that tell the undecided….come on.
Curtice has inadvertently shown his hand on this one.
Arbroath 1320 says:
Better Together constantly come out with ‘promises’ of more powers but amazingly they are well short of any detail, this wouldn’t be because the majority of Scots believe there will be NO more powers for Scotland but a removal of existing powers would it?
Im with you on this – Their dishonesty is crucial; we need to out them in some way.
Scotland should become a fully independent country seperate from the rest of the UK?
As opposed to;
Should Scotland be an independent country?
Now if any pro-independence sponsored poll were to do the same the question posed would be criticised and the result questioned. Devo+ seems to a unionist spoiler offering something that is not on the table and never will be.
Gillie says:
1 September, 2013 at 5:06 pm
Devo+ seems to a unionist spoiler offering something that is not on the table and never will be.
Spot on
Nice to see Stewart Hosie come out on twitter and show support/appreciation for Wings, Newsnet, Scotlandshire etc.
“Nice to see Stewart Hosie come out on twitter and show support/appreciation for Wings, Newsnet, Scotlandshire etc.”
Yeah, that meant a lot today. I see it didn’t take long before the Smear Squad started hounding him on Twitter for it.
I’m now very confused about this poll. Here’s the page from the What Scotland Thinks site where this poll is plotted on the referedum question graph, this poll at the end, as suggested by others there is preamble and then the ‘Should Scotland be an independent country?’ question.
link to whatscotlandthinks.org
Arbroath 1320 says:
The referendum is pure and simple ‘do you want Independence’ or ‘do you want the status quo?’
We need to hammer this home AS OF NOW
Just what have BT said to date about post Sept 2014 vote.
Nothing, zip, zero, nil and they are getting away with it by clever supportive press!
So is this a 3 option poll that has been interpreted as a 2 option poll?
”Scotland should become a fully independent country separate from the rest of the UK”
“Scotland should remain part of the UK WITH THE INCREASED POWERS OUTLINED IN THE DEVO PLUS PROPOSAL”
If these were the questions, then the poll seems pretty dubious.
The independence question gives the impression we would be turning into some sort of isolationist North Korea, not keeping close links with the rest of Britain, or choosing to keep the pound.
And if there was no explanation of what the Devo-Plus proposal actually means, then the question is useless.
From what I understand, many people would be happy for now with Devo-MAX, generally defined as the Scottish Government controlling everything apart from Defense and Foreign Affairs.
I could well imagine that many voters confuse that with ‘Devo-Plus’ , when in reality that will turn out to be a pig in a poke… where extra powers over income tax alone are compensated by a reduction in the block grant, probably leaving us worse off than before.
I imagine the tactic will be to fob off voters with unspecific proposals of ‘more powers’ / DevoPlus, thinking there will be enough confusion that it is actually something meaningful.
and Kevin Williamson too.
New reason to vote YES in 2014 : Councillor Terry Kelly Labour
Onwardssays:
“I imagine the tactic will be to fob off voters with unspecific proposals of ‘more powers’ / DevoPlus, thinking there will be enough confusion that it is actually something meaningful”.
Thats it … scary but effective.
I’ve registered with Yougov and been asked to complete a survey with no prior screening. They know nothing about me but my e-mail address.
Yes, I’m registered with them but I don’t believe I ever told them where I live other than ‘Scotland’. Yet I’ve answered voting and indy questions before. How did they regionally weight? Or did they bother? It’s not shown in tables for Y/N polls so presumably not. Should be there if they have so you can see it was done correctly. That would be a huge fcuk up not to do that; take a look at an election result map…
Also, I comment every time at the end that lots of questions regularly don’t apply as I don’t live in England. Doesn’t make any difference. That potentially makes 8.3% of the results for a UK poll wrong / not exactly right.
Also, I really hope they don’t weight based on newspaper readership in Scotland. That’s a big thing in their polls which they think’s cool / works well. It would be nuts to do that in Scotland unless they’ve developed a system adapted to the different country and it’s newspaper reader demographic that Scotland is.
When you compare Yougov’s approach with e.g. ICM (or all the others for that matter), you think ‘this looks a bit cheap and low quality’.
Could it work the other way, ie a party could stand for election in 2016 on a platform of Defense and Foreign Affairs being a joint matter? (just asking)
“Could it work the other way, ie a party could stand for election in 2016 on a platform of Defense and Foreign Affairs being a joint matter? (just asking)”
I can’t see why not.
Without seeing the results tables, I can already guess the question was something like this:
“If the referendum on Scotland leaving the United Kingdom and becoming a separate country was held tomorrow, how would you vote?”
Are you going to make comment/article about the DR piece today?
Some further v/good points raised on this @ Scot goes Pop!
Personally, I think they could have done better with the referendum question.
Can’t remember where I read it, but in surveys the word ‘independent’, wasn’t seen as overwhelmingly positive – It is too easy for opponents to give the impression that it means separate or isolationist.
Whereas, if the question was more along the lines of ‘Should the Scottish Government have sovereign powers to make decisions for the people of Scotland?’ , then most people actually agree with that.
By the time the referendum came around, everyone would have know exactly what sovereignty meant… The word would have framed the whole debate – Either we have the power to make decisions, or someone else does.
Paula Rose
That is of course the weak point of any devo plus or federalism they are punting. Defence and Foreign Affairs ( ie Trident and Iraq ) remain with London and of course oil revenues.
These are the biggest reasons why we must be independent but some soft witted people seem to be easily charmed by such daft proposals
Yes Dave McEwan Hill I agree, but what I’m asking is what’s to stop the devo plus/max supporters standing for election in 2016 and offering that?
Scottish Skier
I too (there are no doubt many others) who make robust comments via You Gov each time as to what they mean by “the NHS” or “this country” etc. but in common with your good self – no reply is forthcoming. I note that they always link SNP and Plaid Cymru when asking about political preferences. I only persist in answering their interminable surveys as to my coffee drinking habits in order to make this repeated point!
Roger Mexico says:
1 September, 2013 at 2:29 pm
Dramfineday and Dunphail
So you would only expect to be asked every 10-15 years. You’ve probably been asked UK-wide polls though because they do a lot more of those, even though they only need 150 Scots each time.
Thanks for that Roger.
Atypical_Scot says:
I’ve registered with Yougov and been asked to complete a survey with no prior screening. They know nothing about me but my e-mail address
It probably wasn’t one of their proper surveys where they pay you for taking part. If you were fully registered they would have asked you what Party you identified with along with age, job and all the rest of it and they keep this data. If they just have your e-mail, so you can get updates or whatever from them, it’s probably just ‘fun’ surveys that aren’t for a client and they’re hoping to tempt you into joining their panel.
Paula Rose
I would suggest we start to support the concept of “confederal” Britain. That is brought about when countries already independent come together freely and cooperate in areas of mutual interest much like the Nordic Union does.
This has the same easy appeal as the federal or devo plus nonsense which would make our task to achieve independence
scottish_skier
YouGov don’t weigh by Scottish region as far as I know, though oddly enough they do for Welsh ones when they do Welsh polls. I think the truth is that someone asked them to do regular Welsh political polling so they bothered to set something up. When they’ve done Scotland-only polls they’ve usually been one-offs and they’re not going to put the work in unless it’s a regular gig. In the end they’re a commercial organisation.
Scottish commenters on UKPR regularly complain about the irrelevance of some of the questions, but because they ask Westminster VI and that needs to be of all Britain, Scotland usually gets included in questions about the NHS etc. They remember sometimes, they didn’t ask about GCSEs this year for example.
Actually the whole way handle Scotland is a bit odd. They rely on knowing wht GB Region you are in from your records rather than checking each time they send you a poll and they then offer SNP and PC as a main option allong with Con, Lab and Lib Dem. This means pick SNP or PC as your voting intention wherever you live (other pollster seems to check where you live and offer you the appropriate list). As a result SNP/PC regularly score 1 or 2% in London or the North (of England). Today Holyrood, tomorrow…
They do use newspaper readership for weighting, though I get the impression that even they don’t really think much of it. It’s a real problem for all pollsters how you adjust for ‘class’ at the moment. When they are doing Scottish polls though they do weight to Scottish demographic etc, so I assume they would use the appropriate National Readership Survey.
I don’t know that I would say that YouGov were consistently different from other pollsters in Scotland. They did overestimate the SNP by a lot in 2007, but seem to be in with the pack since. true they got the 2011 Regional vote wrong, but so did nearly everybody else and in the same direction. But they do have less experience and I don’t think they have offices in Scotland for example unlike some.
I don’t know that I would say that YouGov were consistently different from other pollsters in Scotland.
Agreed on many of your points. With regard to the above, I was talking about straight Y/N indy polls (they’re not terrible for Holyrood VI, but their UK subsets for Scotland are weird compared to the rest).
Up to 2007, Yougov matched well within variance all the others in terms of Y/N. Then, suddenly, after that they went to become the extremes of Y/N with lowest Y and highest N consistently right up to now. A real outlier of the sore thumb type. Even MORI who do outdated landline only and really go hard on likelihood to vote can’t match Yougov for high No / low Yes scores. I’m puzzled as to what happened here; how can they match the rest, then become such an outlier almost instantly? Not only that, but results hardly change at all. The others shows peaks and falls in response to events but not Yougov in any big way. At best they might mirror a little within variance, but way out as an extreme.
I could only conclude it was something to do with weighting (party affiliation / previous UK GE vote) associated with the rise of the SNP voter for Holyrood who tactically voted Labour in UKGEs but was tempted by SNP for both; Yougov down-weighting people who say SNP considerably in their UK VI polls.
Oh and thanks on the regional weighting thoughts. Would be a big mistake not to do that in Scotland. As a country, it has significant variations by region just like any other. The highlands / NE have long been the heartlands of the SNP / libs with the central belt Labour territory until recently. Then you have the southern borders where the rural Tory vote holds on…
Everywhere north of the forth/clyde is likely to vote Yes. The borders tight. The central belt will carry things, notably Glasgow.
“notably Glasgow”
Indeed. Glasgow and West Central Scotland operate in a significantly different political landscape from the rest of Scotland and this is critical and may well determine the result.
Glasgow and West Central Scotland operate in a significantly different political landscape
Looking at the Glasgow BT launch, I’m not so sure things are that different!
😉
Those who turned out for the BT launch were in no way typical of Glasgow voters
scottish_skier
Everywhere north of the forth/clyde is likely to vote Yes. The borders tight. The central belt will carry things, notably Glasgow.
Well you’d think so (to be honest I thought so too) but then I looked at the last TNS poll because they do a regional breakdown. And that shows a lot of uniformity in the Holyrood regions with the exception not what you’d expect and probably due to them having the two lowest samples (and you might expect Central and West to be the most similar not the furthest apart).
Now these regions are based on weighted sample size of 130-ish, so should be treated cautiously, but it illustrates how we let ourselves get misled by first past the post and single colour expanses it produces. Obviously there will be some real variation, but it’s not a great as people assume.
On your other points small samples also explain part of the oddities with the Scottish cross-breaks in the YouGov ‘daily’ polls, but there are also systemic problems which I think are due to the polls being weighted to GB targets. This seems to boost the Conservatives’ numbers and reduce the SNP. The SNP in that column may also be reduced by ‘losing’ some votes to English regions as explained before where YouGov have the wrong region down for the respondent. Admittedly it’s difficult to check all this because we haven’t had a Westminster VI poll for Scotland only this year from anyone as far as I can tell.
YouGov’s weighting methods should actually get round the problems of split voting because they use Party-id rather than actual vote in 2010 and the odds are most splitters probably don’t identify with either Party and so don’t weighted to either’s targets. There may still be some Scottish problems due to tactical voting, but using id is probably better than the alternative (past vote and doing nothing) providing you have the data.
I didn’t look at older Y/N YouGov polls and they could well have had early problems with high values for Yes if they had a lot of SNP supporters on the panel – especially if unacknowledged. Their 2007 high figures might suggest that too. I also think there may have been weighting problems as well – I know they changed the way they dealt with the SNP a couple of years ago.
Well done, Rev.
I must admit, when I read this yesterday in the Herald, and then an hour later, Scotland on Sunday had leapt onto the story and were screaming it from the rooftops too, I began to feel a dread.
However, there is still a year to go, and who knows what may happen. I just hope Yes Scotland get it right when it finally does produce the White Paper.
Anyway, well done. Reading that and realising that once again, it is a misleading scare story, has just settled me once more.
Keep up the brilliant work. Your a Man who deserves full praise (and then some) should the day finally be won!
Note I’m basing my thoughts on 10+ years of polls as people don’t change their underlying wishes/hopes/dreams(?) on such a subject to readily; it’s not like elections where manifestos get put forward, then dropped etc, leaders come and go.
People may get cold feet, but what they would like deep down doesn’t change, at least not from Yes to solid N. That would be weird. I;ve met plenty N to Yes (When they looked into it), but never a Yes to solid N / love the union.
TNS concerns me as it is face to face hence has a potential shy factor. Do they not also arrange interviews by landline? That would potentially skew the demographic considerably (e.g. see the last MORI which was well off on the Moreno and what the result of that was!). Both it and MORI (landline) showed similar values to the other polls (mainly online) in 2011, then both went off to a larger peak of No with trough of Yes in 2012-early 2013 which were far more accentuated that online polls, i.e. the non-anonymous ones are far more reactive.
When you XY plot Y vs N, N, vs DK and Y vs DK for TNS, you find the strongest correlation is between Y and N, with the next between N and DK and none between Y and DK. This implies people are sometimes saying Yes and at other times saying no; i.e. jumping straight from one opposite to the other. Why do that? You’d have thought people would go from N to DK (which some do) with cold feet Yes going to DK.
In contrast, for online polls which don’t show extremes of reaction but much more gentle changes, while the strongest correlation is still between Y and N, there’s no correlation between N and DK but one between Y and DK. So here, whilst again people who like the idea of independence say no depending on the circumstances (cold feet), many swap to DK from Yes as might be expected. In contrast data imply very few swap from N to DK.
Why the difference in the way people behave depending on how they are asked? Anonymity versus being asked to their face or over the telephone? Anyway, data suggest people using landline telephones tend to be more reactive and say No when they’ve previously said Y (and vice versa) far more than those interviewed online in a more private manner.
I’d also find it very odd that someone who has said Yes, i.e. would like independence, would actually go out and vote No on the day.
Even TNS shows close to parity for Y and N at times, e.g. 2008-2009 and late 2011 suggesting that even this group skewed to likely older generations, young still living with parents etc and partially missing the biggest supporters of independence (mid 20’s to late 40’s who are big internet users / lower landline use) are capable of voting Yes and have even being saying nearly that quite recently.
MORI confirmed much of this when they said their polls – which are good for No – actually only had 36% definitely set on No with 44% still open to Yes in addition to core Yes. Initially I thought that odd until I looked at the tables and there it was. So people are jumping back and forth between Yes and No. If we take MORI as a higher no due to methodology, then solid no is 1/3 which is just about the lowest No values we see historically. 1/3 is also seen as the general base Y. This ties in with what you find on the doorsteps; 1/3 Y / 1/3 N / 1/3 DK. The No has the lead, but it’s not a solid one it would seem.
In terms of regional variations, breakdowns historically show the NE and highlands have the strongest yes. That’s from all the polls, not one.
Oh and the 1/3 thing ties in with Indy-Devo-status Quo. Presumably the devo folk like the idea of independence, just don’t like it being called that / see it as a big step. Will they vote No though? I don’t feel so.
Anyway, the latest panelbase adds a bit of spice to the mix.
http://wingsoverscotland.com/credibility-and-incredibility/