Thursday, July 3, 2014

More on Kellner

Just to let you know I've written an article for Newsnet Scotland about this week's ongoing YouGov saga - you can read it HERE.  That's what I was doing last night instead of writing the detailed commentary I'd promised on Kellner's blogpost!  But I'm delighted to say there's no shortage of such detailed commentary from others - I can recommend Survation's very forceful response, Scottish Skier's comments, and of course the piece by Alasdair Stirling that was published here earlier this afternoon.

Are Peter Kellner and YouGov Yes Scotland’s hidden asset?

A guest post by Alasdair Stirling

YouGov's Peter Kellner has written an extraordinary blog article entitled 'Why Do the Polls in Scotland Vary So Much?' in which he puts forward what Sir Humphrey might call a 'brave' proposition. The gist of Kellner’s view is that the Scottish independence referendum opinion polls (other than YouGov's of course) fail to capture the true character of the constituency of Scottish voters that supported the SNP in 2011. He contends that, absent a corrective methodology such as YouGov applies, polling samples will contain too many of what he terms 'Passionate Nats' and too few – again his term - 'Passing Nats' and that that the resulting imbalance skews the forecast Yes/No percentages in favour of Yes to independence.

Now the SNP is not unique in finding its support from both committed loyalists and floating voters; all political parties experience this phenomenon so it might come as a surprise that Kellner does not advocate any similar corrections for poll respondents that supported the other Scottish political parties in 2011. It is not hard to see a rationale for this approach as it may be arguable that changes in voter allegiance amongst the Unionist parties will have little effect on the the binary Yes/No independence voting intention.

However, Kellner is in some difficulty with his policy of correcting only the poll respondents supporting the SNP in 2011 when it comes to forecasts of party support vis-a-vis the 2016 Scottish general election. Clearly if all of the Scottish parties are susceptible to this problem but the correction is applied to only the SNP supporting respondents, then it is likely that the forecast levels of party support for the 2016 election will be wildly inaccurate. Accordingly, by deduction we have to conclude that Kellner believes either: that the problem applies only to the respondents that supported the SNP in 2011; or that in so far as it applies to all parties its effect is only statistically significant as regards respondents that supported the SNP.

Kellner uncovers this phenomenon – lets us call it the 'Kellner Effect' – by investigating the voting record of 2011 SNP supporting respondents and cross-referencing the same respondents' 2010 Westminster election voting record (YouGov of course rely more on their panel data than voter recall). Kellner advocates that SNP supporting poll respondents be disaggregated into two constituent groups (loyalists and floaters or in his terms 'Passionate' and 'Passing') sampled according to quotas established by reference to the 2010 election and then re-aggregated by means of (if necessary, aggressive) political weighting. Kellner, without explanation, limits the application of this methodology – lets call it the 'Kellner Correction' – to SNP loyalists and Labour floating/tactical voters and concludes his blog article with the assertion that any poll that does not apply the Kellner Correction is wrong (and we must suppose just so much toilet paper).

In his UK Polling Report blog Anthony Wells gives a fine explanation of the sampling and weighting methodologies used by the polling industry. In regard to sampling, he points out why a truly random sample is impractical and cogently explains the Quasi-random and Quota sampling techniques that the polling industry uses to achieve viable polling samples before going on to explain the use of weighting to correct discrepancies and bring a viable polling sample more completely into line with the nature and character of the broader community being sampled.

Quasi-random and Quota sampling techniques are therefore the very foundation stones of the modern polling industry and the Kellner Correction can only be necessary if normal Quasi-random and Quota sampling techniques are unable to capture the true character of the community of SNP voters in Scotland. Absent some malign manipulation of the polling process, which Kellner does not suggest or imply, we must take his blog article as challenging the foundations of political polling. For if the Kellner Effect can apply to SNP respondents in Scotland, why not Conservative/Labour/Liberal/UKIP respondents throughout the UK; and if so, where when and why? Can we reasonably consider the merits of any poll (certainly any Scottish poll) that does not rigorously apply the Kellner Correction?

However these considerations are just the start. The broader implications of Kellner’s Effect and Correction are nothing short of staggering! If Kellner is right, then any polling company not applying the Kellner Correction can effectively shut up shop (at least in Scotland) and YouGov are positioned to establish a monopoly of political polling in a revalidated United Kingdom whilst the rest of the industry plays catch-up. On the other hand, if he is wrong it is hard to see that Kellner himself (having allowed his company to depart so fundamentally from the industry’s tried and tested methods) can survive an independence referendum result that falls into line with the forecasts of the polling companies not employing the Kellner Correction. But beyond potentially ruining his own career, Kellner’s is also throwing the dice for YouGov itself as we can only contemplate the detrimental effect on its business if events prove to its customers that YouGov lacks a commitment to the industry’s tried and tested methods, impartial approach and rigorous standards.

Notwithstanding, perhaps the greater implications lie in the referendum debate itself. With other polling companies showing a narrowing of the forecast Yes/No result, the Better Together campaign take much succour from YouGov’s consistently strong forecast of an emphatic No vote (and its effect on the many Poll of Polls type averages). Better Together’s Campaign Director Blair McDougall (not to mention Labour activist Ian Smart) have much tongue-in-cheek fun tweeting ‘no room for complacency’, but if Kellner is wrong then Better Together really do have no room for complacency, but will thanks to Kellner remain blind to the need to alter strategy. So in addition to imperilling his own career and business, Kellner may yet turn out to be massively assisting Yes Scotland in bringing about the end of the United Kingdom.

Wednesday, July 2, 2014

Wisdom on Wednesday : The insult is in proportion to our size

"It's a racist argument that Scots uniquely cannot run a country when there are smaller nations than ours that do it just fine."

Osama Saeed, a communications professional and former SNP parliamentary candidate.

OK, YouGov, your self-imposed purdah period is at an end. NOW will you answer some simple questions about your methodology and lack of transparency?

You might remember that Laurence Janta-Lipinski's excuse last night for refusing to answer any questions about YouGov's bizarre methodology or their lack of transparency was that an upcoming wonder-blog on the subject had not yet been published on the company's website. It was never actually explained why publication of the blog was such an essential precondition for answering extremely basic questions, but luckily that's now an academic point, because the blog has appeared - and it's written by Peter Kellner himself. So now that it is at last unambiguously permissible under the Janta-Lipinski Doctrine to ask questions (admittedly only "reasonable" questions, whatever that means), I've left the following comment under Kellner's article -

I was told last night by your colleague that once this blog appeared (but NOT before!) any "reasonable" questions asked about YouGov methodology would finally be answered in full. Could you therefore please answer the following -

1) Why are you so obsessively secretive about the voting intention breakdown of the two SNP groups you use for weighting? Why do those numbers never appear in the datasets?

2) If there is logic in separating out Labour-to-SNP switchers from 2011, why not also separate out Lib Dem-to-SNP switchers and Lib Dem-to-Lab switchers, both of whom are very large groups? Surely any reasonable person must conclude that the only reason for the inconsistency in your approach is that you're working backwards to produce the headline numbers that "feel right" to you - and that is hardly something you are capable of being objective about, given that you absurdly declared as long ago as 2011 that it was literally impossible for Yes to win this referendum.

3) There was a YouGov referendum poll last year that asked for people's country of birth, which showed a huge disparity between your sample and the 2011 census figures - you had far too many English-born people and too few Scottish-born people. Given that it's been well-established that being born in England is a strong predictor of a No vote, that by definition means that your headline results would have been artificially skewed towards No. What steps have you taken to correct for that bias in subsequent polls? If you have not taken any such steps at all, why not?

No response so far, but admittedly it's very late at night. We should be able to look forward to a very detailed reply, because these were the undertakings that Mr Janta-Lipinski gave last night -

"we'll have blog up later in the week, happy to answer any and all Qs after u've read it, assuming they're reasonable"

"we are publishing a blog, i will answer all Qs after, on here, email, phone, hell, even face to face"

So it'll be interesting to see if the man is true to his word. Frankly, I'm not holding my breath.

I'll be writing a detailed commentary on Kellner's piece and his (let's call a spade a spade) astoundingly arrogant conclusions when I'm feeling less tired - I'm just back from spending the day in Arran. For now I'll just note what amused me the most about it - Kellner specifically cites the eccentric practice of splitting SNP voters into two distinct groups as a reason for thinking YouGov are right and others are wrong. I'm not quite sure where that leaves Mr Janta-Lipinski's claim that it's OK to be ultra-secretive about that part of the methodology because it's just one trivial detail out of "thousands"!

Tuesday, July 1, 2014

YouGov's mask slips as an angry Laurence Janta-Lipinski justifies his company's obsessive secrecy with the most risible excuse ever

Words (almost) fail me.  Laurence Janta-Lipinski is a London-based employee of YouGov, and I first became aware of him a couple of weeks ago when he wrote a rather arrogant piece that strongly implied that no other pollsters than YouGov mattered in the context of the referendum. Because YouGov were showing a wider No lead, that was by definition the true position, and that was what "Alex Salmond" had to recover from.  Well, my post earlier tonight must have caught Mr Janta-Lipinski's attention, because he indignantly took issue with my (frankly indisputable) point that YouGov are less transparent than other BPC pollsters...

Laurence Janta-Lipinski : in what way are we less transparent than other BPC members?

Me : You don't give voting intentions for the two SNP groups you bizarrely separate out, and no raw numbers at all in datasets.

Laurence Janta-Lipinski : of all the many criticisms of us, lack of transparency is a new one

Me : Then WHY do you never provide the breakdown for the two SNP groups? The obsessive secrecy must be for a reason?

Laurence Janta-Lipinski : i didn't know we were being obsessively secretive. What reason do you think we have?

Me : Will you answer the question? Can we look forward to seeing a full breakdown of the two SNP groups in tomorrow's datasets?

Laurence Janta-Lipinski : sorry, 1 year old has a cold and woke up - he's more important to me than the indyref

Me : Oh, come off it. You could have answered my question in the time it took to write that tweet. Utterly pathetic.

Laurence Janta-Lipinski : we'll have blog up later in the week, happy to answer any and all Qs after u've read it, assuming they're reasonable

Me : Will you answer the question I've just asked, or will you continue to be extraordinarily evasive?

Laurence Janta-Kipinski : p.s., you've not *really* answered the Q on why you think we're so secretive

Me : Yes, I have. A fuller answer is here. Can we have a response now?

Laurence Janta-Lipinski : you've clearly never spent time with a screaming 1 year old!

Me : It was a simple question! Will that info yet again be kept secret in the datasets, and if so, WHY?

Laurence Janta-Lipinski : I have, we are publishing a blog, i will answer all Qs after, on here, email, phone, hell, even face to face

Laurence Janta-Lipinski : no, you will not get breakdown. Why? Why should you? We don't put up 1000s of potential xtabs at behest of bloggers

Me : That is the most absurd and offensive answer. How have you got the brass neck to complain about me pointing out your secrecy?

Me : You ARE obsessively secretive, and you've just boasted about the fact.

Laurence Janta-Lipinski : we're blogging on our methodology this week, any Qs after that, happy to answer them

Me : Will you promise to answer the questions you've evaded tonight?

I don't think anything could sum up YouGov's institutional attitude problem better than the above exchange.  We get the snide implications that anyone who criticises the company is rather tiresome and paranoid, but when Mr Janta-Lipinski's questions are answered in straightforward fashion and a concrete example of YouGov's obsessive secrecy is specified for him, what do we see?  Bluster, evasion, passive-aggression, and then ultimately angry defiance justified by the most risible excuse imaginable.  Make no mistake about this - the stuff about "1000s of potential Xtabs" is not merely an insult to my intelligence, it's an insult to the intelligence of every single person in Scotland who takes an interest in referendum polls.  The voting intention breakdown of the two SNP groups is not some kind of trivial detail out of thousands of others - it is ABSOLUTELY FUNDAMENTAL to how YouGov produce the headline numbers they do.  They are the ONLY COMPANY who eccentrically split SNP voters from 2011 into two groups and then weight them separately.  Every other firm provides a comprehensive breakdown of their sample by past vote recall (with the exception of Ipsos-Mori who don't weight by past vote) - and so do YouGov, apart from that ONE area of relentless secrecy.  Why?  If I had any lingering doubts about whether YouGov have got something to hide, they've been removed by Mr Janta-Lipinski's evasion and excuses tonight.

We'll see whether YouGov's upcoming wonder-blog on this subject actually addresses the issue of why they use such an eccentric methodology, but the bluster tonight about it just being one minor detail out of "thousands" doesn't fill me with confidence that the mist will be clearing very much.  And I think we can already safely assume that we won't be getting any satisfactory explanation for why the voting intentions of the two SNP groups are being kept secret.  But rest assured that I'll be on hand to ask the awkward questions again and again (unless they block me!), and I suggest that you are too.

And if I sound angry, it's because I am.  YouGov are playing games with the future of this country, and the least they can do is explain themselves adequately.  I've thought for some time that they're a rather reprehensible company, but I only realised tonight just how boorish their public face can be.  It also appears that they don't get out much, if they honestly don't think anyone has ever accused them of a lack of transparency before!

UPDATE : Mr Janta-Lipinski later responded further, but only with intimations of yet more evasiveness to come...

Laurence Janta-Lipinski : given your reaction to earlier tweets, I'm minded not to promise anything. It's pretty clear your mind is made up.

Laurence Janta-Lipinski : that being said, happy to answer any questions you have later in the week

Me : The man who sneered at complaints about YouGov secrecy, now warns that he may evade reasonable questions. It's laughable.

Me : How can you change minds when you openly boast about your determination to withhold information?

Me : Your claim that splitting SNP voters in two is one trivial detail among "1000s" is an insult to people's intelligence.

Me : You know that as well as I do, and I hope you have the decency to be privately embarrassed by it.

Monday, June 30, 2014

YouGov's controversial methodology has now claimed the mantle of the outright most No-friendly pollster

Apparently there is a new YouGov poll out in The Times tomorrow.  The paper's Nat-bashing night editor has predictably gone straight into "ANOTHER blow for Salmond!!!" mode, but in fact the changes are margin of error stuff from a pollster that has now not merely consolidated its status as one of the most No-friendly pollsters, but has in fact overtaken even Ipsos-Mori to become the outright most No-friendly pollster.

Should Scotland be an independent country?

Yes 39% (-1)
No 61% (+1)

We'll have to wait and see what the headline figures are (ie. not excluding Don't Knows), and in particular what the figures are for those respondents who say they are certain or very likely to vote - in the last poll those were considerably better for Yes.

It's a very odd pattern we've been looking at of late - we had polls a couple of weeks ago from Panelbase and Survation showing the gap between the two sides at its narrowest point to date, and an ICM poll that also showed a sharp decrease in the No lead, but those have been counterintuitively followed by two YouGov polls showing slight increases in the No lead.  It may be that YouGov are the earliest to pick up a new trend and that others will soon follow, but until at least one other pollster shows the same thing I'll be very sceptical.  All of the caveats about YouGov's extremely odd methodology still apply.  They always sharply upweight Labour-to-SNP switchers from 2011, which means they're heavily reliant on the belief that the voting intentions of a very small number of respondents are representative of a very wide section of the electorate.  They're also probably interviewing those people over and over again for each poll (because they can't find enough 2011 switchers), and that can easily lead to respondents becoming 'self-conscious' and just reflexively giving the same answer every time.  We also know from one previous poll that YouGov's sample had far too many English-born people in it, and too few Scottish-born people in it, so if that's an ongoing problem and not one that was specific to that individual poll, then it could be persistently skewing the poll numbers in a No-friendly direction.

UPDATE : We now have the headline numbers, which again show tiny margin of error changes, and thus make a bit of a nonsense of The Times' breathless claims that the poll is somehow "dramatic" and shows a "slump in Salmond's support" -

Yes 35% (-1)
No 54% (+1)

That suggests it may have been quite a close-run thing as to whether the figures with Don't Knows excluded were rounded up to 40/60 or rounded down to 39/61 - but we'll never know, because YouGov are less transparent than all other BPC pollsters, and always keep their raw numbers a closely-guarded secret.

To put it in perspective, tonight's 19-point No lead is just 1% higher than the 18-point lead reported by YouGov at the start of March - which at the time was the lowest lead reported so far in the campaign by this most No-friendly of pollsters.  The lead is still a whopping 11 points lower than it was at its peak last August.

I can't see any mention on the Times front page of the figures for people who are certain or very likely to vote, so we may have to wait until the datasets are published (hopefully tomorrow).

*  *  *


Once again, YouGov's small margin of error changes have failed to make much of a dent on the overall trend of a declining No lead in the Poll of Polls - two updates ago we had the lowest No lead of all-time, in the last update we had the second-lowest No lead of all-time, and in this one we have the third-lowest No lead of all-time - just 10.7% when Don't Knows are taken into account.

MEAN AVERAGE (excluding Don't Knows) :

Yes 43.6% (-0.2)
No 56.4% (+0.2)

MEAN AVERAGE (not excluding Don't Knows) :

Yes 36.5% (-0.2)
No 47.2% (+0.2)

MEDIAN AVERAGE (excluding Don't Knows) :

Yes 43.6% (-0.2)
No 56.4% (+0.2)

(The Poll of Polls is based on a rolling average of the most recent poll from each of the pollsters that have been active in the referendum campaign since September 2013, and that adhere to British Polling Council rules. At present, there are six - YouGov, TNS-BMRB, Survation, Panelbase, Ipsos-Mori and ICM. Whenever a new poll is published, it replaces the last poll from the same company in the sample. Changes in the Poll of Polls are generally glacial in nature due to the fact that only a small portion of the sample is updated each time.)

Here are the long-term trend figures, with updates prior to Easter recalculated to remove the inactive pollster Angus Reid ...

The No campaign's lead in the Poll of Polls mean average (not excluding Don't Knows) :

Sep 2013 - 21.6%
Sep 2013 - 21.4%
Sep 2013 - 19.4%
Oct 2013 - 18.8%
Oct 2013 - 18.4%
Oct 2013 - 18.2%
Nov 2013 - 18.4%
Nov 2013 - 18.0%
Dec 2013 - 17.0%
Dec 2013 - 16.8%
Dec 2013 - 16.4%
Jan 2014 - 14.4%
Jan 2014 - 14.2%
Jan 2014 - 14.2%
Jan 2014 - 15.2%
Feb 2014 - 15.0%
Feb 2014 - 15.5%
Feb 2014 - 15.5%
Feb 2014 - 13.7%
Feb 2014 - 13.3%
Feb 2014 - 14.2%
Mar 2014 - 14.2%
Mar 2014 - 14.5%
Mar 2014 - 14.5%
Mar 2014 - 14.7%
Mar 2014 - 13.8%
Mar 2014 - 13.0%
Mar 2014 - 12.5%
Apr 2014 - 12.5%
Apr 2014 - 12.7%
Apr 2014 - 12.7%
Apr 2014 - 12.3%
Apr 2014 - 11.4%
May 2014 - 11.2%
May 2014 - 11.2%
May 2014 - 11.5%
May 2014 - 13.3%
Jun 2014 - 12.1%
Jun 2014 - 12.1%
Jun 2014 - 11.3%
Jun 2014 - 9.9%
Jun 2014 - 10.3%
Jun 2014 - 10.7%

Cameron's Euro posturing proves to be double setback for the anti-independence campaign

We know that there are two key factors outwith the Yes campaign's control that could potentially help them win the independence referendum - one is the possibility of the Tories opening up a decisive lead in Britain-wide opinion polls, and the other is the possibility of the UK's exit from the EU starting to look imminent.  On the first point, the European elections initially seemed to have worked in the No campaign's favour - UKIP enjoyed a mini-surge of support in the polls that was largely at the Tories' expense, thus enabling Labour to artificially regain the narrow lead they had more or less squandered in the late spring.  And on the second point, there seemed to be enough pandering to Cameron from other European leaders that it was at least plausible to imagine that a cosmetic "reform" package might be cobbled together to keep Britain in the EU.

But the extraordinary Juncker episode seems to have turned the game on its head at a stroke.  Serious commentators in the London media are now talking of the likelihood of an exit from the EU in a way that would have seemed laughable only a few weeks ago.  And, bizarrely, Middle England seems to be lapping up Cameron's pointless display of impotence, isolation and irrelevance on the EU stage, and have rewarded the Tories with a clear lead in the latest GB-wide Ashcroft telephone poll.  Could this double whammy possibly have occurred at a worse moment for the No camp?

GB-wide voting intention for 2015 Westminster general election :

Conservatives 33% (+5)
Labour 31% (-2)
UKIP 15% (-2)
Liberal Democrats 9% (n/c)
Greens 6% (-1)
SNP/Plaid Cymru 3% (-1)

Don't be fooled by the combined share for the SNP and Plaid being down slightly, because the SNP have actually moved into the lead in the Scottish subsample...

SNP 36% (+9)
Labour 27% (-9)
Conservatives 21% (+4)
Liberal Democrats 6% (+4)
Greens 2% (-5)
UKIP 2% (-6)

However, very little can be read into that, because after the turnout filter was applied, there were only 34 respondents left in the Scottish subsample!

Sunday, June 29, 2014

Seducing Rob Shorthouse

This is not the most riveting Twitter exchange I've ever reposted here, but it's still worth looking at, because you really have to constantly remind yourself that the guy I'm talking to is the official No campaign's Director of Communications.  He seems for all the world like the mirror image of me - someone just sparring on Twitter to pass the time.  But he's actually being employed at a very senior level to do this (and as he once pointed out to a bemused audience at a public debate, the most important thing about the referendum is that it's paying his mortgage).  Can you imagine any of the equivalent senior people in the Yes campaign - Dennis Canavan, Nicola Sturgeon, Stephen Noon, Kevin Pringle, etc - wasting their time making snide comments on Twitter to try to get one over on a random No supporter?

It's the sheer lack of class in the ranks of the official No campaign that never ceases to take my breath away.  It's not as if Shorthouse is one bad apple - his boss Blair McDougall acts in much the same way on social media.  It's like the Tories have hired a bunch of overgrown, sniggering schoolboys to help save "the greatest political union in the history of the known universe".

Rob Shorthouse : The #bbcbias protest is in full swing. What they don't know is that #MI5 is about to turn on the mind beam.

Me : Shouldn't they have appointed a comedian as Better Together's Director of Comedy?

Rob Shorthouse : As you wish... (posts a photo of Eddie Izzard apparently auditioning to be the new James Bond)

Me : Yes, Izzard almost certainly could do your job better.

Rob Shorthouse : You must be absolutely brilliant at parties.

Me : Another classic, Rob. All your Christmases came at once when the Tories said they'd pay you to be an internet troll.

Rob Shorthouse : Fancy going out sometime? Just me and you. Go for a drink? Maybe catch a movie?

Me : I think we both know what's happening here, Rob. Don't fight it.

Rob Shorthouse : Steady on pal. I was only asking if you wanted to go for a drink. Take it slow, don't rush me!

Me : Take all the time you need. The best things in life are worth waiting for.

Rob Shorthouse : You are creeping me out a little now.

Me : Don't let irrational fear ruin something special, Rob. (That advice also has broader applications.) #indyref

Bannockburn Live photos

I know it's probably the ultimate nationalist stereotype to attend a reenactment of Bannockburn, but I honestly wasn't planning to go until a couple of nights ago, when I had a look at the musical line-up and suddenly got a bit tempted.  It also struck me that I went to the reenactment of the Battle of Largs last September, and it would probably seem a bit odd on reflection if I'd gone to that one but didn't bother with the biggest one of the lot.  It's been ages since I was last in Bannockburn - oddly enough, I think the last time was in May 2007, just after the SNP originally took office, and on that visit I had the picture taken that I still use for my profile on this blog!

I was astonished at how many Yes umbrellas, T-shirts and flags were in evidence.  There was no sign at all of the opposition, unless you count a couple of half-hearted Union Jack brollies.  But perhaps that's understandable, because - on balance - Bannockburn wasn't really a 'No Thanks' kind of scrap.

Apologies as ever for the terrible picture quality - I'll get organised with a proper camera one of these days.  The musical acts are the Precious Penny Pluckers, Rura, Karine Polwart, Dougie Maclean, Rachel Sermanni and Saor Patrol.  As you can see, a sharp-suited Jean-Claude Juncker dropped by to catch a bit of Karine Polwart in the rain.  You might also just about be able to spot the Red Arrows in the picture of the Bruce statue.  They flew straight over Dougie Maclean's head as he started to sing - I think we can safely assume that wasn't an intentional salute for the writer of Scotland's alternative national anthem, but it was certainly a very dramatic moment, which unfortunately I wasn't quick enough to catch!

The very last picture is of Stirling in the evening as I walked back to the train station - presumably the combination of saltires and Union Jacks was a compromise for the armed forces event that was taking place at the same time as Bannockburn Live.

(Click to enlarge.)