1 November 2011

YouGov: That latest independence poll in full...

Since May's Holyrood election, we've had a three polls on independence, with TNS-BMRB publishing the findings of their June and September polls on Scottish opinion on this testy topic, and Ipsos-MORI contributing a third poll, also in September.  Over the weekend, the Scotland on Sunday publishing the results of a YouGov survey, conducted over the 26th and 27th of October.  After a delay to allow the paper to get its money's worth, and to frustrate the amateur psephologist, the full tables are now were available here earlier today but appear now to have vanished in a reshuffle of YouGov's website.  The texture of the findings got rather more attention in the paper than is typical, delving beneath the topline, and pulling apart the overall picture to reveal the interesting, often neglected details across different demographic groups.  Ane Corbie has already cast her beady eye across the implications of the data along similar lines. In usual style, I thought it might be helpful and interesting for folk if the busy detail of the full table was reduced to its composite elements.

But first, the question. With options of "I would vote YES/NO/wouldn't vote/don't know", YouGov asked its 1,075 Scottish adults...

"The SNP wishes to hold a referendum on Scottish independence in due course. Voters would be asked whether they agree or disagree "that the Scottish government should negotiate a settlement with the Government of the United Kingdom so that Scotland becomes an independent state". How would you vote if such a referendum were held tomorrow?"

For a start, you'll recognise that YouGov are putting the ridiculously circumlocutory SNP question to would-be voters. The explanation for this wending legislative language isn't Alex Salmond's idiosyncratic and unaccountable preference for the language of negotiation, but entirely dervies from the legal problems dogging the Holyrood referendum, on which I have now bored for Scotland, most recently here. By contrast, the two other pollsters who have recently generated findings on attitudes to independence phrase their questions rather differently. Jettisoned is the talk of negotiation. Instead, Ipsos-MORI asked:

"... whether you agree or disagree with a proposal to extend the powers of the Scottish Parliament to enable Scotland to become an independent country, separate from the UK." 

While TNS-BMRB enquired of its informants...

"The SNP have outlined their plans for a possible referendum on Scottish independence in the future. If such a referendum were to be held tomorrow, how would you vote?"

However, the talk of negotiation recurred in the answers TNS-BMRB solicited from its respondents, with folk being invited to take a view on whether or not they agreed that "the Scottish Government should negotiate a settlement with the Government of the United Kingdom so that Scotland becomes an independent state."  Inquisitorial differences aside, last Sunday's YouGov poll generated the following findings:

How would you vote if such a referendum were held tomorrow?" (total)
  • Yes ~ 34%
  • No ~ 52%
  • Wouldn't vote ~ 3%
  • Don't know ~ 12%

And the gendered profile...

How would you vote if such a referendum were held tomorrow?" (men)
  • Yes ~ 41%
  • No ~ 50%
  • Wouldn't vote ~ 2%
  • Don't know ~ 6%
How would you vote if such a referendum were held tomorrow?" (women)
  • Yes ~ 27%
  • No ~ 54%
  • Wouldn't vote ~ 3%
  • Don't know ~ 17%

Just like May's election, these figures clearly imply that a strategy to connect with female voters will be absolutely vital, if the referendum is to be carried off for the nationalist opinion. Female support for independence lags a stonking 14% behind their male fellow citizens, while their opposition is 4% points higher. On the brighter side, there are more women to be convinced than men, with only 6% of gentleman -  apparently careful to project a robust appearance of decisiveness - willing to avow indecision about independence at this stage, while indecision amongst female respondents ran a mighty 11% higher.  As Kate notes in her piece, it seems eminently likely that it will be easier to sway waverers towards independence than invert existing preferences, turning opposition into support. It is to women, then, that the advocates of independence must particularly attend. Thirdly, we have the social gradings...

Middle class (ABC1)
  • Yes ~ 31%
  • No ~ 56%
  • Wouldn't vote ~ 1%
  • Don't know ~ 11%
Working class (C2DE)
  • Yes ~ 36%
  • No ~ 48%
  • Wouldn't vote ~ 3%
  • Don't know ~ 13%

Data is also given on how respondents would vote, based on their Westminster, Holyrood constituency and Holyrood regional voting intentions. I don't proposed to comprehensively examine these - if interested you can look for yourself. There are a couple of curiosities however, that I thought I'd flag up. Firstly, YouGov found no more than 49 respondents out of 1075 who intended to vote Liberal Democrat - some 3.6% of their sample considering that Tavish Scott managed a mighty 7.9% and 5.2% in the most recent Holyrood election. We might be cautious, therefore, about reading too much into the surprising topline statistic that 21% of the 39 Liberal voters for Westminster supported independence.

Finally, a word on age. Of late, would-be LOLOTSP ("Leader of Labour outside the Scottish Parliament") Tom Harris has been suggesting that the Nationalists' desire to emancipate sixteen year olds to vote in the referendum represented a perfidious scheme, cozening independence and fixing the result by sheer democratic malice. The Maximum Eck has talked about the "independence generation", while Peter Murrell's Your Scotland, Your Future website is fronted, at present, by a saltire-bearing bairnlet. Interestingly, YouGov's findings don't present such a stark differential in support for independence on the basis of age, but decidedly records hardening hostility as respondents grow older. 

Support for independence (by age)
  • 18 - 24 year olds ~ 36%
  • 25 - 39 year olds ~ 30%
  • 40 - 59 year olds ~ 37%
  • 60+ years old ~ 31%

As you can see, at least in this poll, middle-aged folk are the most supportive group, while the most ancient cohort were a smidgeonwise more positive about the proposition than those between twenty five and thirty nine years of age. Realistically, nothing can be read into such small deviations, but the findings do at least suggest that Tom Harris' fears of the young aren't built on strong data presently in the public domain. Chances are, only the SNP are presently in the position to engage in that sort of polling, and their findings are unlikely to find the light of day. Now compare this with opposition to independence by age across the same four groups...

Opposition to independence (by age)
  • 18 - 24 year olds ~ 46%
  • 25 - 39 year olds ~ 50%
  • 40 - 59 year olds ~ 51%
  • 60+ years old ~ 57%

Contra Alex Salmond's invocation of "the independence generation", perhaps Harris et al should start talking up "the Unionist Generation" - the youngest and freshest spirits of this vintage being laid down by 1951, children and teens of the 1960s and none-too swinging-things of the 1970s?


  1. Am I right in thinking that the way YouGov works is that people volunteer to be part of their interviewee pool, to be polled as and when required, and are weighted according to the information provided by the interviewees themselves without any external checks?

  2. Anonymous,

    YouGov set out their methodology here.

  3. LPW

    Thank you. Have you any idea what percentage of SNP members are YouGov volunteers compared to other political parties?

  4. No bother, Anonymous.

    As to your second question, I'm afraid I haven't the foggiest.

  5. YouGov principally weight their sample (after the obvious demographic factors) by "political ID", so it doesn't matter how many supporters of a party are signed up to YouGov.

    YouGov select the sample that they are going to ask, not the other way round!

    Peat Worrier - you are actually being a tad cynical (who? me?) in assuming the reason for the delay in SoS's publication of the poll.

    They didn't commission the poll. All they did was to commission a single question which was piggy-backed on a poll which someone else was paying most of the cost of. Hence, why the voting intention figures weren't released to SoS.

  6. oldnat said...
    'YouGov principally weight their sample (after the obvious demographic factors) by "political ID",'

    So if I was running things for the SNP, and I wished to influence opinion polls like YouGov, could I let it be known, sotto voce, that members should consider joining the panel, but me imaginative and creative when stating their political allegence as they fill in the application form?

  7. Oldnat,

    Cynical, moi? It was a daft wee aside. As you say, the real reasons for delays in publication are generally much more quotidian.