Wednesday, March 26, 2014

Breakthrough for pro-independence campaign in landmark new YouGov poll

Although the No lead has been slipping in recent YouGov polls, the pace of change has been frustratingly sluggish compared to most other pollsters - a state of affairs that even allowed Peter Kellner to pen an extraordinary reality-defying article in which he implied that public opinion had essentially remained static, and that the apparent movement was an illusion caused by normal sampling variation. Well, those comforting words for the No campaign have finally been blown away by new numbers from YouGov, which corroborate the strong pro-Yes trend shown recently by Survation, Panelbase and ICM.

If there was a referendum tomorrow on Scotland's future and this was the question, how would you vote? Should Scotland be an independent country?

Yes 37% (+2)
No 52% (-1)

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

Yes 42% (+2)
No 58% (-2)

And among respondents who say they are certain to vote, the figures with Don't Knows excluded are -

Yes 43% (+2)
No 57% (-2)

A 15% No lead may still look substantial, but it's well below the normal range for YouGov, which has consistently been one of the most No-friendly pollsters (indeed for a time it was the outright most No-friendly firm). 37% is the highest Yes vote they've shown over the entire course of the campaign so far, breaking the record of 35% that was set only last month.  The latter detail is also very much in line with the trend detected by ICM and Survation, which both showed Yes at a new high watermark in their most recent polls - although in those cases the figure was 39%.

Most polling analysts seem to be charitable to Mr Kellner and his colleagues by only comparing any new referendum poll they produce with other YouGov polls that have appeared since September, when a huge methodological change was introduced. Everything before that seems to be as mythical as Oceania being at war with Eurasia. However, we shouldn't entirely lose sight of the fact that in the final poll produced under the old procedures in August, No had a massive lead of 30% - and for the only time in their lives the London media were actually justified in referring to it as a "2-1 majority against independence". So if we take the headline numbers at face value, the gap has literally halved over the last seven months, with another six months still to go. And that's assuming that Yes actually need to be in the lead with YouGov by polling day, which may well not be the case - they could quite easily finish the campaign ahead on the polling average while still being as much as 5% behind in the final YouGov poll.

Until recently it looked as if the pattern across all the pollsters pointed to there having been significant movement to Yes between September of last year and January of this year, followed by a few weeks of consolidation thereafter.  But recent polls have pointed to the possibility of a further surge for Yes, and even Professor Curtice suggested that two more polls showing a similar trend would more or less confirm that.  It seems we're now halfway towards receiving that confirmation.  It's still conceivable that the uniform direction of travel shown by Panelbase, ICM, Survation and YouGov is coincidental, and that all four pollsters are misleading us due to margin-of-error effects.  But that does look increasingly unlikely.

Even with this new poll, YouGov remain significantly more No-friendly than all other online pollsters.  It's unlikely (although not impossible) that the composition of their panel is significantly different to the others, and of course they've finally got their act together to some extent with the wording of the preamble, so what's going on?  The most probable explanation is the unusual practice of splitting respondents who voted for the SNP in 2011 into two distinct groups, and weighting them separately.  One group always seems to be upweighted massively, while the other usually seems to be downweighted, which may indicate that the balance between the two groups is being misjudged, or that the whole process is just totally misconceived.

Who knows, there may be method in YouGov's madness, but at face value it does look very, very odd.

*  *  *


For the second time in just a few short days, the Yes vote has soared to a new high watermark in this blog's Poll of Polls, and the No lead has slumped to yet another new low.

MEAN AVERAGE (not excluding Don't Knows) :

Yes 35.6% (+0.3)
No 48.1% (-0.2)

MEAN AVERAGE (excluding Don't Knows) :

Yes 42.5% (+0.3)
No 57.5% (-0.3)

MEDIAN AVERAGE (excluding Don't Knows) :

Yes 42.0% (n/c)
No 58.0% (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, and that adhere to British Polling Council rules. At present, there are seven - YouGov, TNS-BMRB, Angus Reid, 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.)

The median numbers remain unchanged for what seems like the umpteenth time because Angus Reid (who haven't reported for months) are still the mid-point, although YouGov are closing in on them fast. TNS-BMRB have now overtaken YouGov as the second-most No-friendly pollster after Ipsos-Mori.

Here are the updated long-term trend figures...

The No campaign's lead in the Poll of Polls headline figures :

Sep 2013 - 20.2%
Sep 2013 - 20.0%
Sep 2013 - 18.4%
Oct 2013 - 17.9%
Oct 2013 - 17.5%
Oct 2013 - 17.4%
Nov 2013 - 17.5%
Dec 2013 - 17.1%
Dec 2013 - 16.3%
Dec 2013 - 16.2%
Dec 2013 - 15.8%
Jan 2014 - 14.2%
Jan 2014 - 14.8%
Feb 2014 - 14.8%
Feb 2014 - 14.7%
Feb 2014 - 15.1%
Feb 2014 - 13.6%
Feb 2014 - 14.0%
Mar 2014 - 14.0%
Mar 2014 - 14.3%
Mar 2014 - 14.3%
Mar 2014 - 13.6%
Mar 2014 - 12.9%
Mar 2014 - 13.0%
Mar 2014 - 12.5%

And last but not least, here are the updated averages for the four online pollsters that have reported so far this year (YouGov, ICM, Panelbase and Survation) -

MEAN AVERAGE OF ONLINE POLLSTERS (not excluding Don't Knows) :

Yes 38.8% (+0.5)
No 47.8% (-0.2)


Yes 44.8% (+0.4)
No 55.2% (-0.4)


Yes 45.4% (n/c)
No 54.6% (n/c)


  1. Thanks for the breakdown. I wonder how it compares with the poling before the last Scottish election? (as I remember all the pollsters underestimated the scale of the SNP win)

  2. Actually, a lot of the pollsters were pretty close to the mark in their final 2011 poll - but YouGov was the big exception. Having been the most accurate pollster in the 2003 Holyrood election, they really performed dismally in 2011.

    However, they've changed their methodology since then, so we shouldn't jump to too many conclusions based on 2011 alone.

  3. As people are beginning to search out information, they are being converted to Yes. Simples!