Monday, September 9, 2013

'Besides, as one of my aunties put it, that Ruth Davidson has been a bit of a disappointment, hasn't she?'

One observation I made in Michael Greenwell's podcast the other day was that there is, in fact, something positive that campaigners can do about polls, in the sense that (if you have the money) you can simply commission one that pushes your own priorities, and in which the dice are loaded in your favour in terms of getting a favourable result. The findings can then be used to shape the media narrative. Lord Ashcroft has just proved that point by effortlessly getting his latest 'research' featured as the lead story in the Scotsman, although I'm sure it's just pure coincidence that a billionaire No-supporting Tory peer of the realm was able to pull off that trick where crowd-funding Yes supporters failed just a few weeks ago.

So how do we know that this poll was tactical, rather than a dispassionate attempt to uncover the views of the electorate? Well, apart from the obvious point that Lord Ashcroft is Lord Ashcroft, we need look no further than this part of his commentary on the results -

"I also found many voters deeply sceptical about the idea of giving the Scottish Parliament more powers"

Er, no. If you want to know whether voters are deeply sceptical about giving the Scottish Parliament more powers, what you do is ask voters whether they think the Scottish Parliament should be given more powers or not. Lord Ashcroft mysteriously (well, let's face it, not very mysteriously) failed to ask that question, and instead chose to grill his respondents about a string of Tory hobby-horses such as the likelihood of fiscal discipline under devo max. Those results may well be of great interest to him, but shouldn't be of much interest to the rest of us - unless of course we've been hoodwinked by his spin into thinking they mean something that they don't.

But if that part of his commentary is a touch cynical, Ashcroft's closing line is just plain risible -

"Besides, as one of our participants put it, Alex Salmond has quite enough power as it is."

Which is about as meaningful as me saying that I met a guy called Barry down the pub who made some cutting comments about Jackson Carlaw. Seriously, your lordship, a general rule of thumb in polling is that one person's opinion is not statistically significant, no matter how pleasing that opinion may be to your own ears.

Unsurprisingly, the Scotsman has also misrepresented the poll to some extent -

"Researchers also found voters believe the Scottish Government cares more about independence than issues such as jobs, the economy and the NHS, and say this priority is wrong."

Not true on the latter point. We know that 61% of respondents said that the Scottish government should have a different priority from the one it currently has, but since less than half of respondents actually thought that the Scottish government's priority is independence, that isn't sufficient to substantiate the Scotsman's claim.

The referendum voting intention figures from this poll can also be safely filed away as little more than a historical curiosity, as the fieldwork is between four and seven months out of date. (The main referendum question was also asked FIFTH, which ought to instantly destroy the poll's credibility in the eyes of certain unionist commentators if they are interested in maintaining at least a semblance of logical consistency.) The Scotsman valiantly attempt to breathe some relevance into it by pointing to the unusually high sample size, but the reality is that a normal-sized sample of 1000 only carries a very slightly greater margin of error. Meanwhile, the Holyrood voting intention figures (showing SNP leads of five points on the constituency ballot and twelve points on the regional list ballot) are "only" three months out of date.


  1. Most polls in the main are just interesting at this point. As you have noted it is getting the appropriate coverage of the polls that metters in many ways. That is where the battle at the moment is being lost, the media are so anti-independence that getting an honest review of polls is impossible and getting a fair hearing even more difficult. I do feel that it is all to play for and one major paper may well come out for YES nearer the time, I am thinking news international possibly but the SNP have pulled it out the bag before and their expereince should help the YES campaign as we get down to the nitty gritty. Another possitive is that I feel the no campaign have fired all their main shots and the YES campaign hasn't even really started yet. Times will be getting interesting in the new year.


  2. "I also found many voters deeply sceptical about the idea of giving the Scottish Parliament more powers".

    Strange that.

    Other polls have variously shown that the Scottish public want devo- max (ie more power for the Scottish Parliament, just short of independence ...Channel Islands style), and that they trust the Scottish parliament more than they trust the UK parliament, with matters such as taxation and social security (which might lead one to believe that they wished that the Scottish government had powers in these areas).

    "Besides, as one of our participants put it, Alex Salmond has quite enough power as it is."

    At least he had the grace to say "one of our correspondents", even if he didn't go so far as to call him Barry.

    Maybe Mr Ashcroft knows the statistical importance of one person thinking something and weights it highly. If so, it may be a kindness on my part to tell him that I heard someone in Morrisons this morning say "the whole House of Lords should be blown up with all the useless plonkers in it".

    Don't thank me Your Grace.

    It is interesting that the Scotsman made such a importance of the English nobleman's self-funded opinion poll, but chose to ignore completely one self funded by a commoner.

    Ah... the antiquated ways of the good old upper classes...