Saturday, June 14, 2014

Pro-independence campaign surge to 48% in spectacular new Panelbase poll

The Sunday Herald have just tweeted the results of this weekend's new Panelbase poll (with Don't Knows excluded).  It shows the Yes vote at an all-time record high for the firm -

Should Scotland be an independent country?

Yes 48% (+2)
No 52% (-2)

I say an all-time high, but of course that's only the case if we exclude the Panelbase poll late last summer which put Yes in an outright lead.  The legitimacy of that poll was later called into doubt due to an unusual question sequence (a little like last month's ICM poll).  Because this new poll was supposedly commissioned by Yes Scotland before being passed on to the Sunday Herald, we'll just have to be slightly cautious until we get firm confirmation that the referendum question was asked first.  But my guess is that it probably was - it was certainly asked first in the two Panelbase polls commissioned by the SNP and Yes Scotland respectively earlier this year.

Assuming that the methodology for this poll does indeed turn out to be sound, what does it mean?  Well, the most obvious thing to say is that two regular referendum pollsters (Panelbase and Survation) have both shown Yes at an all-time high within the space of a few days.  That could be coincidence but it's unlikely.  Oddly enough, my heart sank when I initially saw the Survation datasets, because there were a couple of unusual patterns which made it somewhat more likely that the swing to Yes might have been an illusion caused by random sampling variation.  But the fact that Panelbase have independently found exactly the same trend sets my mind to rest on that score to some extent (although obviously it's still possible that ICM might show something different later tonight).

Oh, and it's probably also worth making the small point that if Survation and Panelbase are getting their methodology right, it's now clear that the Yes campaign have every chance of winning this referendum, and Scotland could well be on the brink of becoming an independent country.  Even if the more No-friendly pollsters are closer to the mark, there does now appear to be renewed momentum towards Yes, so the somewhat larger No leads reported by the likes of YouGov can't be regarded as in any way safe.

This is the third poll to be published during the campaign that can be called a 'statistical tie', meaning that if the standard margin of error is to believed, there is a greater than 5% chance that the side that appears to be behind is actually in the lead.  The previous two were the aforementioned controversial Panelbase poll last summer, and the ICM poll published on Easter Sunday.

The new ICM poll is now out - details of that, plus a landmark Poll of Polls update, can be found in a fresh post here.

Coming soon, to a blog near you...

In case you haven't heard, we'll shortly be discovering the results of two new referendum polls - the monthly ICM poll in the Scotland on Sunday, and a Panelbase poll.  If the political editor of the Record is to be believed (obviously a big question mark over that), the Panelbase poll will appear in the Sunday Herald but was commissioned by the SNP.  That would be consistent with what we heard from Calum Findlay earlier in the week - he took part in the poll and said it had obviously been commissioned by a pro-independence client.  That won't in any way detract from the credibility of the headline numbers as long as the referendum question was asked first - and that was indeed the case in the Panelbase polls commissioned by the SNP and Yes Scotland earlier this year.  But we'll just have to wait and see.  Presumably the numbers must be reasonably good for Yes, otherwise the SNP wouldn't have passed them on to a newspaper.

I presume there can't be anything dramatic in the ICM poll, because the Scotland on Sunday front page leads with some characteristically irresponsible spin about the results of the supplementary questions.  We also haven't seen the usual suspects on either side drawing attention to the poll on Twitter (with the partial exception of Nicola Sturgeon), so I would imagine it's a middling sort of result.

Don't forget to check back later for what will hopefully be an all-time high for the Yes vote in the Poll of Polls.

Friday, June 13, 2014

Is it time for the Yes campaign to play hardball against gutter journalism?

VICTORIA : It's a fossil.

THE DOCTOR : Victoria, be careful. Let me see that. Yes, it's certainly inactive, but it's not a fossil. Wait a minute.

(The Doctor consults his 500-year diary.)

THE DOCTOR : Yes, here we are. It's a Cybermat.

VICTORIA : What's a Cybermat?

THE DOCTOR : It's one of those. I'd leave it alone if I were you. Come along.

(Victoria pops the Cybermat into her handbag.)

*  *  *

Kit Pedler and Gerry Davis have got a lot to answer for.  With those lines in the 1967 Doctor Who story The Tomb of the Cybermen, they unwittingly coined a term of abuse that (in very slightly modified form) has ended up being used several decades later to dehumanise any supporter of Scottish independence who happens to have an internet presence.

We all know the truth.  Any of us who have engaged for any length of time with anti-independence activists (or even just anti-independence people in general) on online forums and social media is well used to being subjected to such extreme abusive behaviour that we almost fail to register it after a while - it becomes background noise.  And yes, there are idiots on both sides, but the statistical evidence from a scientific opinion poll for the Sunday Express speaks for itself - 21% of Yes supporters have suffered abuse, compared to only 8% of No supporters.

The idea that supporters of independence are disproportionately responsible for any abuse is one of the most brazen 'Big Lies' that any political campaign has ever attempted to propagate - and an independent mainstream media worthy of the name would quickly challenge it and put a stop to it.  Instead, journalists are enabling it and colluding with it.  Today's front page of the Daily Record may well be the most despicable thing I've seen in this campaign outside the pages of the Mail.  They've done exactly what Iain Macwhirter pointed out journalists were doing - gone actively searching for an example of abuse in the sure knowledge that it must exist somewhere, but they've gone looking for it on the Yes side only.  They've then seized upon the most extreme example they can find, and used it to smear all independence supporters.

My feeling is that the time has come for the Yes campaign to fight fire with fire.  I know they probably won't do that, because the mantra is that Yes can only win by being relentlessly positive.  But the whole point about this particular smear campaign is that it's intended to undermine the public's perception of Yes as positive - and in particular, it's a cynical attempt to foster distaste for the Yes campaign among undecided female voters.  There comes a point where you have to take some brief time out to defend your reputation against people slinging mud which otherwise might stick.  I wouldn't mind seeing a succession of morning press conferences featuring the personal stories of people who have been abused for their pro-independence beliefs - and that spectacle should carry on relentlessly, day after day, until the media are shamed into giving the stories at least one-tenth of the coverage they've lavished on the Clare Lally nonsense.  By all means questions should be asked about whether the abuse was personally ordered and coordinated by David Cameron, Alistair Darling and Blair McDougall, and those men should be challenged to "call off the dogs" without delay.  Yes, that would be complete garbage, but it's precisely the sort of garbage we have to put up with as a matter of routine, so I fail to see why the other side should get a free pass.

And with the benefit of hindsight, I think it was a mistake to ask Campbell Gunn to make such an overly-generous apology to Lally - he should have simply made a curt apology for a minor factual error, and stuck to his guns about everything else.  Being all goody-two-shoes about it was interpreted by the media as a sign of weakness - and as the saying goes, "no good deed goes unpunished".

Wednesday, June 11, 2014

Pro-independence campaign close the gap to just 5% in sensational new Survation poll

Based only on some very insubstantial straws in the wind, I've had a slight feeling that the direction of travel in the aftermath of the European elections may have been in the favour of Yes.  Last night's TNS poll didn't really offer any clues because the fieldwork mostly predated the announcement of the European election results, and although last week's Ipsos-Mori poll was conducted during the right period, it was hard to interpret because it was the first poll from the firm for three months.  But the latest monthly poll that has just been released from Survation may - and I do emphasise the word may - be the first proper indication that something has indeed been happening.  It shows the No lead at just 5% - that's a 5 point drop on last month, and is 3 points lower than Survation have shown at any point in the campaign to date.

Should Scotland be an independent country?

Yes 39% (+2)
No 44% (-3)

With the Don't Knows stripped out, the No lead has been cut in half from 12 points to 6 -

Yes 47% (+3)
No 53% (-3)

Survation have thus joined ICM and Panelbase as one of three BPC-affiliated pollsters that have shown the No lead at 5 points or lower (ie. close to what the Americans call a "statistical tie") at some point during referendum year so far.

It's important to stress at moments like this that one swallow does not a summer make.  We know that there will be another ICM poll in a few days, and there's a rumour about a Panelbase poll in the works as well, so if those polls fail to replicate tonight's trend then it's possible we're just looking at a poll that is unusually flattering for Yes due to margin of error effects.  But if one or both of those upcoming polls show something out of the ordinary, then we may just be in business.  And I can't emphasise enough that last night's TNS poll in no way contradicts Survation, simply because the fieldwork took place considerably earlier.

At an absolute minimum, this poll represents the final nail in the coffin for the idea that the increase in the No lead suggested by the last ICM poll was in any sense real.  It's clear now that, at worst, the position has been static in recent weeks, and at best there has been sufficient movement towards Yes to take us to the brink of victory with more than three full months to go.  The latter point is naturally contingent on the belief that relatively Yes-friendly pollsters like Survation and Panelbase are closest to the truth - if by any chance Ipsos-Mori are the most accurate then Yes will still require significant gains (although of course even Ipsos-Mori are saying that the No lead has slumped from 28 points to 18 since last September).

Tonight's poll was commissioned by the Daily Record, who are actually making a bigger splash about one of the supplementary questions. When people were asked how they would vote in the referendum if they were sure that David Cameron would remain Prime Minister, this is how they responded -

Should Scotland be an independent country?

Yes 44%
No 38%

With Don't Knows excluded, that works out as -

Yes 54%
No 46%

That finding is an absolute hammerblow for the No campaign.  Doubtless John Curtice will caution us that people are generally bad at accurately answering hypothetical questions, but in a sense that's not the point.  What these numbers prove is that there are sufficient people out there who are at least open to the idea of voting Yes without requiring any implausible financial guarantees (ie. "if you could be sure that you would be X amount of money richer per year").  There's a path to victory for us here.

And as it happens, this linkage of David Cameron's future with a possible majority for Yes coincides with a dramatic improvement in the Tories' fortunes at Britain-wide level.  Perhaps helped by the Newark by-election result, they seem to have regained the ground they lost to UKIP in the aftermath of the European elections, and have been just 2% behind Labour in both of the last two YouGov daily polls.  If that trend continues, it won't be long until there is unequivocal polling evidence that Cameron is heading for another five-year term as Prime Minister - and then we'll see if Survation's respondents mean what they say about opting for independence in those circumstances.

A final important detail - the fieldwork for this poll took place entirely after President Obama's rather half-hearted (or should that be quarter-hearted?) "endorsement" of the No campaign.  So it appears that Cameron's latest desperate attempt to enlist the help of a world leader has either had no effect at all, or has catastrophically backfired.  But I'm sure all of us are far too mature to even dream of finding that highly amusing.

*  *  *

SCOT GOES POP POLL OF POLLS

In spite of the fact that the Poll of Polls sample still includes what increasingly looks like a rather questionable poll from ICM, the No lead has practically returned to its all-time low on the average that takes account of Don't Knows (the second batch of figures below).  Incidentally, I had a brief discussion on Twitter yesterday with Ipsos-Mori's Steven Hope about the merits of various averaging methods.  He felt that the FT's methodology is the best one, although he did go on to add that no method is perfect, and that it's important to look at various different averages to get the complete picture.  Frankly, my own view is that what we're seeing tonight illustrates perfectly the misconceived nature of both the FT and the Curtice averaging methods.  Over recent weeks, the FT Poll of Polls has seen the No lead increase from 7 points to 14 - and that's happened without the slightest scrap of evidence from any individual pollster (with the possible exception of ICM) that a trend towards No was actually underway.  It came about simply because a sample full of polls from Yes-friendly firms like Panelbase was gradually replaced with a sample featuring a greater number of polls from No-friendly firms like Ipsos-Mori.

Only the method used to produce the figures below is more or less guaranteed to reflect the real trends, because the respective weighting given to Yes-friendly and No-friendly firms does not change from one update to the next (except in the rare circumstance of a new firm entering the fray).

MEAN AVERAGE (excluding Don't Knows) :

Yes 43.2% (+0.4)
No 56.8% (-0.4)

MEAN AVERAGE (not excluding Don't Knows) :

Yes 36.0% (+0.3)
No 47.3% (-0.5)

MEDIAN AVERAGE (excluding Don't Knows) :

Yes 42.3% (n/c)
No 57.7% (n/c)

(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 the updates prior to Easter recalculated to exclude 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%

*  *  *

Why the hell didn't anyone bother to tell me that STV's former current affairs show Scotland Tonight has been relaunched as a comedy series?  If only I'd known I would have tuned in long before tonight.  It really is the freshest, most rib-tickling new comedy to appear on our screens since Bonjour La Classe made its triumphant debut back in 1990.  In case you missed tonight's episode, the hilarious plot concerned a woman who was on the brink of tears because someone had accurately pointed out that she's a member of Labour's Shadow Cabinet, and had also inaccurately suggested that she's the daughter-in-law of Pat Lally.  In spite of the fact that the "wrong-doer" had for some inexplicable reason issued her with a grovelling apology for doing practically nothing at all, she continued to ramp up the victimhood for the cameras and told the nation that it felt more like an insult than an apology.  It seemed that nothing short of Campbell Gunn committing a Japanese-style ritual suicide would have sufficed.

I think what made the production work so brilliantly was the utter seriousness with which many of the participants treated the woman's complaints - there wasn't even the slightest hint of a giggle at the fact that the "hate attack" upon her largely consisted of being wrongly called Pat Lally's daughter-in-law.  The highlight of the episode (which is already assured of its place in the pantheon of iconic TV comedy moments alongside Del Boy and Rodney dressing up as Batman and Robin) was the horrified intake of breath from a former Labour aide when Stuart Campbell accurately described the woman as being a "hard-core political activist".  The exquisiteness of the comic timing was, I felt, rather reminiscent of the "even if he does say Jehovah" moment in Life of Brian.

Question - in the brave new Scotland unveiled by our mainstream media today, shouldn't the Labour party be issuing a grovelling apology to Pat Lally for suggesting that being called the daughter-in-law of Pat Lally is something that requires an apology?

*  *  *

This is a real conversation that occurred just a few hours ago -

Family Member A : Is it true that people have been attacked on the streets just for being No supporters?

Family Member B : No, no, I think they were talking about verbal attacks.

Family Member A : I was sure I heard something about physical attacks...

Moral of the story : It's high time that journalists stopped using such irresponsibly inflammatory language when they refer to minor (and sometimes non-existent) Twitter spats.  Not everyone hangs around to hear the details.

TNS-BMRB poll : Support for independence stands at 45% among definite voters after 'undecided leaners' are taken into account

As I said last night, the aspect of the new TNS-BMRB poll that will be of most interest to the Yes campaign (indeed, for the opposite reason it will be of equal interest to the No campaign) is the position once 'undecided leaners' are added in.  TNS-BMRB haven't collated those numbers themselves, but fortunately their datasets (unlike YouGov's) are detailed enough that it's possible to make the calculation.  First of all, here are the figures for the 72.5% of the sample who say they are certain to vote in September -

Should Scotland be an independent country?

Yes 38.1%
No 47.3%

Excluding the hard-core of undecideds who express no leaning at all, it works out as -

Yes 44.6%
No 55.4%

If TNS had collated those numbers for publication, they would of course have been reported as Yes 45%, No 55%.  That's a strikingly small gap coming from a pollster that doesn't rely on a volunteer online panel.  However, we do have to bear in mind that most predictions for the turnout in September now exceed 72.5%, so the more meaningful figures may come from the 82% of the TNS sample who say they are either certain or very likely to vote -

Yes 37.6%
No 47.4%

With the hard-core of undecideds excluded, it's -

Yes 44.3%
No 55.7%

Those aren't too shabby either.  Admittedly the position is a touch less encouraging among the whole sample, working out at Yes 42.0%, No 58.0% after non-leaning undecideds are stripped out, but whether that actually matters is open to question, given that no-one is predicting a 100% turnout.

Having said that, Scottish Skier always points out that a particularly important figure is the percentage of the entire sample who say they are certain to turn out to vote and who currently plan to vote No - that currently stands at just 31.9%.  It rises to 35.6% when you take into account people who say they are very likely to turn out to vote, and who plan to vote No.  So much for the No campaign's theory of a year or two back that the "quiet but determined anti-independence majority" would make this a triumphal procession for them.

*  *  *

As Tom "work ethic" Harris perversely found it so "nauseating" that a decent, hard-working, law-abiding (etc, etc) couple decided to use some of their lottery winnings to help the cause of Scottish self-government that they've believed in all their lives, I wonder how he'll feel about a fabulously wealthy celebrity from South Gloucestershire donating money to the anti-independence campaign that she accrued after an Edinburgh cafe owner took pity on her and let her write stories about wizards in comfort and warmth?  As you'll probably have seen in Yes Scotland's email this afternoon, it doesn't take magic to fight back against Ms Rowling's attempts to further unbalance the playing field - you can make a small donation to the official Yes campaign by clicking here.  Other very important fundraisers running at the moment are Generation Yes, Business for Scotland and Green Yes (I'm sure there are plenty of others as well, but I struggle to keep track of them all!).

Wisdom on Wednesday : The more things change, the more they stay the same...

"The United States appear to be destined by Providence to plague America with misery in the name of liberty."

200-year-old words from the forefather of Latin American independence, Simón Bolívar.

TNS-BMRB poll suggests Yes campaign have retained their big gains among definite voters

This month's edition of the TNS-BMRB referendum poll has just been released, and it shows an almost identical position to last month.  Among respondents who are certain to vote, the Yes campaign have retained three-quarters of the significant gains they made in the previous poll, when the gap was slashed from 13% to 9%.

Should Scotland be an independent country? (Definite voters only)

Yes 34% (-1)
No 44% (n/c)

A 1% slippage in the Yes vote is of course statistically insignificant.  Although TNS don't use the figures for definite voters as their headline numbers, other pollsters would do (notably Ipsos-Mori), so it's striking that on this crucial measure the Yes campaign have once again hit the kind of heady heights with a face-to-face pollster that they've previously only been accustomed to reaching with online firms.  We'll have to wait and see what the exact figures are after Don't Knows are stripped out, but a rough calculation suggests that they'll be somewhere in the region of Yes 44%, No 56%, which would be very much in line with the picture painted in recent times by the relatively Yes-friendly online pollster Survation.  [UPDATE : The datasets confirm it is indeed Yes 44%, No 56% after rounding.  Apologies for the incorrect update earlier.]

The headline numbers which take the whole sample into account show that Yes have managed to stay on 30% - their high watermark of the campaign so far (or at least the record high since TNS introduced a radical methodological change which hugely increased the reported number of undecided respondents).  The No lead remains stuck at 12% - which is similarly the record low for the campaign so far.

Should Scotland be an independent country?

Yes 30% (n/c)
No 42% (n/c)

The rough calculation on those numbers suggest that with Don't Knows excluded it's most likely to be Yes 42%, No 58%.  However, I said exactly the same thing last month, and it turned out that the Yes vote had fallen victim to the rounding, and was actually still at 41%.  It would certainly be a big psychological boost if it turned out to be 42% this time, because it would be the highest Yes vote that TNS have reported in the campaign so far on that particular measure.  [UPDATE : Unfortunately, John Curtice has revealed that it's once again Yes 41%, No 59% on the rounded numbers.]

As is always the case with TNS-BMRB, the fieldwork for this poll is ridiculously out of date.  It took place between the 21st and the 28th of May, which presumably means that well over half of the interviews were conducted before anyone had even heard the results of the European elections.  So if there was any swing to Yes directly generated by revulsion towards UKIP's victory south of the border, we wouldn't necessarily expect it to show up in this poll.  However, the good news is that the fieldwork did take place well after last month's ICM poll, and the fact that it has failed to show any increase in the No lead at all is another powerful indication that the ICM poll was probably reporting a misleading trend. Indeed, if we look at this poll in combination with the most recent Panelbase and Survation polls (and to some extent last week's Ipsos-Mori poll, although that's harder to interpret because it was the first for three months), I'd say we now have reasonably conclusive proof that the increase in the No lead reported by ICM was an illusion.  However, it's still open to question whether it was caused by a particularly extreme example of margin of error "noise", or by the absurd methodological change which may have unsettled respondents before they were asked the main referendum question, or indeed by a combination of both factors.

TNS have done in this poll what ICM and Ipsos-Mori do as a matter of routine, and pressed undecided voters on how they are most likely to vote.  The results are highly encouraging for the Yes camp -

Yes 19%
No 15%

The remaining two-thirds of Don't Knows were still unable to say which way they were inclined to jump.  But it's important to remember that TNS always report a far higher number of undecided voters than any other pollster - Don't Knows make up 28% of the entire sample in this particular poll.  So a 4% lead for Yes among this group is certainly a matter of some significance.  Presumably we'll shortly find out what the figures for the whole sample are when 'undecided leaners' are added in, but my rough calculation (this is becoming a habit!) suggests that the position will probably be something like Yes 35%, No 46% - or Yes 43%, No 57% after the hard-core of undecideds are excluded.

I couldn't help but raise a smile at the revelation on the front page of the Herald that the No campaign are dismissing the lead for Yes among undecided respondents, with "a small sample size" being the stock excuse for the evening.  It's only ONE WEEK since Blair McDougall was crowing about the fact that Don't Knows in the Ipsos-Mori poll were breaking slightly more for No - even though that was based on a smaller sample.  They really have absolutely no shame.

[UPDATE : I'm slightly confused after reading John Curtice's blog, because he suggests that undecided voters are breaking exactly evenly in this poll.  I can only assume therefore that the 19%-15% split is among definite voters only.  Still no word on what the numbers for the whole sample are once 'undecided leaners' are added in, although presumably among definite voters the Yes vote will be lifted up to either 44% or 45%.]

* * *

SCOT GOES POP POLL OF POLLS

No change in the TNS headline numbers naturally means no change in the Poll of Polls, so this update is mostly a very simple cut-and-paste job!

MEAN AVERAGE (excluding Don't Knows) :

Yes 42.8% (n/c)
No 57.2% (n/c)

MEAN AVERAGE (not excluding Don't Knows) :

Yes 35.7% (n/c)
No 47.8% (n/c)

MEDIAN AVERAGE (excluding Don't Knows) :

Yes 42.3% (n/c)
No 57.7% (n/c)

(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 the updates prior to Easter recalculated to exclude 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%

Tuesday, June 10, 2014

Breaking : ICM change their question sequence again

A week has passed since the last full-scale referendum poll was released, but we should certainly hold on to our hats, because at the very least a new ICM referendum poll is on its way - which probably means it will appear in Scotland on Sunday at the weekend.  The reason I know is that someone whose wife took part in the poll yesterday very kindly sent me screenshots of all the questions (apart from the routine demographic questions at the start).

The big news is that the absurd introductory "comfort" question from last month is still there, but it's been moved to the very end of the question sequence where it can't do any harm.  It would be interesting to know whether ICM have reflected on their blunder from last month and realised it may well have distorted the results on the main referendum question, or whether it was always part of the plan to do a "split test" to see if the results are the same regardless of the order in which the questions are asked.  If the latter is the case, I must say it seems extremely unprofessional of ICM to have compromised the integrity of a high-profile published poll simply to enable them to conduct some speculative testing - it's a bit like a football stadium using the middle of a World Cup match as an opportunity for a fire drill.

If we learned anything from last month's experience, it's that we should be extremely cautious about the headline results of any new ICM poll until we find out whether any other methodological changes have been surreptitiously made - and that might not become clear until a day or two later.  For a third month in a row, ICM have asked for people's country of birth, so it'll also be important to see whether it remains the case that Scottish-born respondents are significantly under-represented in the weighted sample, and English-born respondents are significantly over-represented.  And will the bizarre under-representation of male respondents in the last poll have been corrected?  All of these factors played a role in suppressing the reported Yes vote in last month's headline figures.

The standard supplementary questions about the economy, equality, pensions and the future of devolution are asked, but there are also a few additional questions -

* Respondents are asked to assess how risky they think it would be to choose independence or to remain in the UK, on a scale of 1 to 5.

* Respondents are asked to agree or disagree with a series of six statements, one of which indulges the bee Martin Boon has in his bonnet about the possibility that people might think they'd be seen as unpatriotic if it was known they were voting No.  (In my view a much more interesting question would be whether some people feel blackmailed by their employers to vote No.)

* Respondents are asked the question that comes up with tedious predictability every four years - do they want England to do well or badly at the World Cup?  This is about as pointless as asking people how much they've given to charity in the last year, or how many sexual partners they've had in their life - in many cases they'll have given the answer they feel casts them in the best light, rather than telling the truth.

Monday, June 9, 2014

Whatever else may be true about this campaign, it's not true that "Better Together are in the lead"

Of the many objectionable things that journalists tend to do in their coverage of opinion polls, perhaps the silliest is their habit of summarising a poll showing a No lead as being "a lead for Better Together".  A referendum is a vote for or against a proposal. It's not an election or even a popularity contest, and the idea that people who are still sceptical about independence are in any sense giving a vote of confidence to the God-awful campaign run by God-awful Blair McDougall is frankly barking mad.  That point has been driven home in extremely timely fashion by a supplementary question that has just been released from last week's Ipsos-Mori poll -

Based on what you have read, seen or heard, which of the campaigns do you think has been the most effective so far, if any?

Yes Scotland 51%
Better Together 23%

Incredibly, among currently undecided voters the lead is even bigger -

Yes Scotland 50%
Better Together 14%

And it's almost as big among the 'underlying undecided' group, which includes those who gave a voting intention but admit they may yet change their mind -

Yes Scotland 45%
Better Together 23%

Even among No voters, it's practically a dead heat -

Yes Scotland 33%
Better Together 36%

When you take that in combination with the fact that Alex Salmond and Nicola Sturgeon have much higher personal satisfaction ratings in the poll than their counterparts in the No campaign, it's absolutely clear that if this was a contest for elected office, the Yes campaign would be winning by a landslide.  They'd even have a fighting chance if the election was restricted to voters who are opposed to independence!

Sunday, June 8, 2014

Those in glass houses and all that...

Kenny Farquharson of Scotland on Sunday, in a tweet about the Populus poll which so far has been retweeted 26 times -

"The SNP there, heralding a new #indyref poll that has 548 respondents and doesn't ask the actual #indyref question."

Hmmm. I might have more respect for smug comments of that sort if Kenny hadn't withheld crucial  information about the jiggery-pokery that took place in his own publication's last ICM poll, while at exactly the same time leading people up the garden path with a tweet about how that poll would have been worse for Yes if a far less important methodological change had not been made.

In any case, one half of his objection to the Populus poll doesn't even make sense - there is no problem with a sample size of 500-odd. No problem at all. That kind of sample would be considered absolutely routine in the US, and indeed two out of Angus Reid's three indyref polls last year had samples of about 500. No, the only issue with the Populus subsample is that it probably wasn't weighted to Scottish demographic target figures - and that would still have been an issue even if the sample had been ten times as large.

As for the indyref question not having been asked, I wonder if Kenny had any sarcastic comment to make about Better Together's antics in respect of the last Progressive Partnership poll, which a) didn't use the indyref question, and b) was claimed to be a YouGov poll even though it wasn't? Perhaps he did say something, but he can't have said it very loudly.