The Panelbase poll for Wings over Scotland


  By Dave Taylor
The recent poll carried out by Panelbase on behalf of the political website Wings over Scotland (WoS) contained some very intriguing responses to some very interesting and clever questions.
Some readers will already have seen the results of this poll in the series of articles already published by the site itself, they make for interesting and entertaining reading.

The variety of questions asked provides a rare opportunity to look at some oft ignored aspects of the independence debate and extrapolate into others.  Given this, it is disappointing though perhaps not surprising that the media in Scotland appear to have ignored this survey completely.

Referendum voting intentions (asked indirectly) comes out as Yes – 34% : No – 36% : Don’t Know – 30%.  This compares with the Panelbase poll of two weeks previously which directly asked the referendum question and produced a voting intention of Yes – 33% : No – 45% : Don’t Know – 20% (including all respondents – not just those most likely to vote, where the figures were Yes – 37% : No – 46% : Don’t Know – 17%.)

The difference is probably more apparent than real.  In the Wings poll, Yes and No supporters were asked how their voting intention compared with the referendum policy of the party that they say they voted for in 2010 at the Westminster election.

There was, however, no option for respondents to answer that they couldn’t remember who they had voted for in 2010, or that they had no clue as to what policy such a party might have!  The only choice for those beset by doubt on the question was “I don’t currently know how I plan to vote in the referendum”.

Interestingly though, few Yes supporters seem to have been confused, compared with 9% of prospective No voters who did express a doubt.

WoS also asked which powers should be devolved to Scotland in the event of a No vote.  As has been the case for some time now, the survey revealed a majority of Scots want Holyrood to control welfare (60%) oil revenue (53%) and taxation (52%) if we remain in the UK.  Only 19% wanted “no new powers” and a mere 6% wanted Holyrood’s powers to be reduced.

Note that the wording didn’t ask what powers respondents would like to see devolved, but what powers SHOULD be devolved.  In short this could be viewed as a line in the sand for many Scottish voters and a clear message to the Unionist parties currently dithering over what ‘more devolution’ might mean.

What emerged in the WoS poll, however, is the cynicism recorded among the population that these matters WOULD be devolved by Westminster, if there is a No vote.  According to the survey 48% thought that none of the powers indicated would be further devolved, and 19% thought that powers would in fact be reduced.

In an clever twist to the referendum question, panel members were asked: “If Scotland was currently an independent country, would you vote to surrender control of taxation, welfare, defence and oil revenues in order to join the Union?”

Responses to hypothetical questions are always hard to gauge.  Although the referendum debate should have provoked some thinking as to whether the advantages of the UK Union outweigh Scottish control of these issues, we know that many people are not fully engaged with the issues.

Only 18% (fewer than wanted no enhanced devolution) said they would rejoin the UK if Scotland were currently independent, while 55% said they would not want to enter a Union.

The survey also asked about the role of the media in the debate.  While 49% said their coverage “will NOT significantly influence how I vote” this is well below the 90% of GB respondents who said that the media wouldn’t influence their decision in the 2010 election.

Only 6% said the media “will significantly influence how I vote”, but in reality we all like to think of ourselves as being above being influenced by advertising or media manipulation.  The huge amount spent on successfully advertising products suggests that we are often fooling ourselves!

It isn’t surprising that many Yes supporters (regardless of party) would be critical of the media’s referendum coverage.  Committed supporters of any side in a critical debate will notice biased reports against their stance.

62% of those agreeing that the media is “biased against independence” identify as SNP supporters – that’s the largest party and the one most Yes voters support.  However, less than half of SNP voters agreed.  More significantly, 15% of Tory and Labour voters shared that view.

The view that the media is “biased in favour of independence” had much less support – 15% of Tory voters felt that way, along with 11% Labour and 12% of LDs.  Only 4% of SNP voters thought that was the case.

The lack of trust in the main stream media may explain why, for the Scottish political websites listed, more SNP voters had heard of them compared with voters for other parties.  The reach of these sites remains low when compared with the likes of the BBC.  Perhaps surprisingly, only 9% of respondents (13% of SNP voters) had heard of Newsnet Scotland!

The question: “Which of these do you think represents a significant threat to Scotland in your lifetime?” produced an interesting response.

52% chose “Conservative governments elected by the rest of the UK”, and that included 9% of Tory voters!  A terrorist attack, at 34%, was in 2nd place while few seriously considered an attack from another country as realistic.

58% said “None of the above” to the question “Which of these threats do you think the siting of nuclear weapons in HMNB Clyde (Faslane) provides Scotland with a practical defence against?”.

Around a quarter of Tory voters saw it as protection against attack by North Korea, Iran, Russia and China.  It would be fascinating to know how the 22% of Tories, 17% Labour, 16% LDs and 13% of SNP voters think Trident could be a defence against terrorist attacks!

In few polities are politicians well regarded, and this is true in Scotland as well.  The poll asked “On the basis of what you’ve personally seen and heard, which of these people do you think are telling the truth about independence?”

Respondents were given four key names from each of the campaigns.  Alex Salmond, Nicola Sturgeon, Denis Canavan and Blair Jenkins on the YES side, together with Michael Moore, Alastair Darling, Anas Sarwar and Blair McDougall from the NOs.

Among the options were “Don’t know who they are” (DK) and “Haven’t heard them talk about it” (HH) to measure their likely influence on the referendum we can add these (DK + HH) to give a referendum invisibility rating (IR).

  • McDougall – (56% + 17%) 73%
  • Jenkins – (55% + 17%) 72%
  • Sarwar – (40% + 20%) 60%
  • Canavan – (36% + 22%) 58%
  • Moore – (34% + 20%) 54%
  • Sturgeon – (9% + 12%) 21%
  • Darling – (8% + 17%) 25%
  • Salmond – (4% + 7%) 11%

Not surprisingly, the person most people have seen and heard is the First Minister Alex Salmond.  Alistair Darling is just ahead of Nicola Sturgeon.  The two non-political figures, Jenkins and McDougall, will be slightly concerned that they remain virtually anonymous.

To the invisibility rating, we can then add the bullshit rating (BS) – the difference between those who have heard them and think they “rarely or never” tell the truth and those who think they “always or mostly” tell the truth.

The highest BS rankings are for Moore (20%), Sarwar (19%) and McDougall (17%).  Jenkins has the highest BS rating for the Yes side (9%), while Canavan and Darling both share an 8% BS rate. Sturgeon at 4% and Salmond at 3% are thought to have very little BS!

This gives us a measure of the total positive influence (PI) (for their side) of these individuals by the formula inverse IR + inverse BS = PI.

In order of PI they are Salmond (86%), Sturgeon (75%), Darling (67%), Canavan (34%), Moore (26%), Sarwar (21%), Jenkins (19%) while McDougall trails the field at only 10%.

The WoS website has added a series of brief “Data mining” reports based on the cross-breaks in the tables – especially those based on their recalled 2011 vote.  In addition to the usual warning that the margin of error in these can be enormous, it’s worth looking at another aspect of “recalled vote” that many pollsters use.

In 2011, 50% of the electorate voted yet 80% of the respondents to this poll named the party they had voted for.  This isn’t unusual.  In the latest Angus Reid and ComRes polls 81% said they had voted, and a similar phenomenon appears in many countries.  The reality of “false recall” is also well known.  With no ill-intent, people often “adjust” their memories to recall who they wished they had voted for (instead of who they did … if they did).

On that basis, we should expect that the distribution should more resemble current voting intention than what actually happened in May 2011.  Indeed, that seems to be the case.

Here are the actual voting percentages of the whole electorate in the 2011 constituency vote (with the “inflation” in the poll in brackets).  The higher the “inflation”, the greater numbers thinking they actually did vote for the party they named.

  • Did not vote 50% (- 35%)
  • SNP 23% (+14%)
  • Lab 16% (+10%)
  • Con 7% (4%)
  • LD 4% (+2%)

An alternative explanation could be that those responding to polls are actually more likely to be people who vote, but that requires a more generous view of human nature than this author is inclined to have!

Congratulations to Stu Campbell of Wings over Scotland for giving us some fascinating data to chew over.

The full data tables are here http:

The relevant Wings articles are:
We did a thing,
In, out, and shaking it all about,
Looking for someone to trust,
The fourth and fifth estates, 
Looking ahead,
The war of the worlds.