Anecdotal evidence for Shy Yes Syndrome

You might remember that ICM tainted the results of their last-but-one referendum poll by asking a creepy introductory question that enquired how "comfortable" people felt about being asked the question they were about to be asked, ie. should Scotland be an independent country.  The firm's stated reason for doing this was that there were "anecdotal" suggestions of the existence of Shy No Syndrome, ie. people telling pollsters that they plan to vote Yes or are undecided when they actually plan to vote No.  The "comfort" question was intended to test whether the "anecdotes" were grounded in reality.

But the obvious question to ask is - where exactly were these "anecdotes" originating from?  ICM are a London-based firm, so it's highly unlikely that their employees had been randomly picking up intelligence on the street.  I was very struck the other night when I read an Alex Massie blogpost that I overlooked a few weeks ago, which paraphrased the No campaign's private spin on the polling situation in terms that were eerily similar to Martin Boon's explanation of what ICM were testing for - ie. this could be 1992 all over again, but with shy Tories replaced by shy Nos.  Could it be that ICM's political contacts in London (who are, let's face it, likely to be overwhelmingly sympathetic to the No campaign) have been whispering in their ear : "Listen chaps, you're making a big blunder here.  You'd better start introducing a Shy No Adjustment, otherwise you're going to have egg on your face on September 19th."

The reality is, of course, that if ICM had the same number of contacts on the Yes side, they'd be hearing an altogether less fantastical message - that it is much, much harder for people to openly admit to supporting independence, and therefore the existence of Shy Yes Syndrome is far more probable than the existence of Shy No Syndrome.  For pity's sake, I'm a pro-independence blogger, and I blog under my real name and a real photo, but even I had to force the words out last year when a group of English people asked for my opinion on independence.

To reinforce that point, it's well worth taking note of something that Scottish Skier told us on a recent thread.  He revealed that the mother of a friend of his had been interviewed by Ipsos-Mori, and had told them she was undecided about how to vote, even though she actually plans to vote Yes.  That's fairly clear-cut anecdotal evidence for Shy Yes Syndrome, and must rival anything that was being whispered in ICM's ear a few weeks ago by the siren voices of the London establishment.  The only health warning I would put on this is that there can't be all that many Yes voters in Ipsos-Mori's sample who are pretending to be Don't Knows, for the simple reason that there aren't all that many Don't Knows in Ipsos-Mori's results, and at least some of them must be genuinely undecided.  But we shouldn't discount the possibility that there also Yes people who are so embarrassed that they pretend to be straight Nos, and indeed Don't Knows who pretend to be Nos.

ICM followed up their initial test by asking the same 'comfort' question in last month's poll, but mercifully shifted it to the end of the question sequence where it couldn't taint the headline results.  But they also kept the results a secret that time, which is mildly disturbing.  The question is - are they just doing this testing for research purposes, or are they seriously toying with the idea of artificially adjusting the No vote upwards in future polls?  After the 1992 disaster, they did introduce a "spiral of silence" adjustment which artificially increased the reported Tory vote - but the difference in that case was that the existence of Shy Tory Syndrome had been conclusively proven by the small matter of a general election result.  If they start playing silly buggers in their referendum polls on the basis of no concrete evidence at all, and if there's clear reason for suspecting that they're doing it after pressure was applied by the No camp, then it will be nothing short of outrageous.

The next ICM poll for Scotland on Sunday should be out either this coming weekend, or the weekend after that.  In the past, methodological changes have not always become apparent straight away, so it will be well worth keeping an eye out for anything in the results that doesn't entirely make sense.

*  *  *

I was canvassed for only the second time in my life yesterday - and I'm delighted to say that it was by Yes Scotland!  It was very brief and to the point - have I given any thought to how I will vote in the referendum, which party do I normally vote for, and thanks for your support.  However, he did identify which side he was on before he asked the questions, so that will obviously be impacting on the results that the canvassing produces.

I was feeling slightly harassed when I came to the door, so my apologies to the man in question, just in case he's reading this!