Just a quick post about last night's Question Time from Dundee (a pro-independence city that mysteriously always seems to do an impersonation of Buckinghamshire whenever the BBC pay a visit). When Joanna Cherry pointed out that the last three opinion polls had shown a majority for independence, the host Fiona Bruce made two forceful interruptions -
"It's within - it's within the margin of error, Joanna, let's be clear about that."
"They're within the margin of error, so they can't - they're not as strong as you'd like them to be, I'm sure."
You'd think Ms Bruce might be a bit wary about making interventions on the subject of opinion polls, given that she had to issue a humiliating apology a year ago after she "corrected" (in rather mocking fashion) a comment Diane Abbott made about the polls, only for it to turn out that the "correction" was inaccurate and that Ms Abbott's original claim had been entirely right all along. But was Ms Bruce on stronger ground last night?
Well...up to a point, Lord Copper. It's true that, on an individual basis, all three of the recent polls show a Yes lead that is within the margin of error. But when several polls all show the same thing, the equation changes somewhat. If, for example, there were eight or ten polls that all showed Yes in the low 50s, the margin of error would no longer be an alibi, because that wouldn't be happening by chance. (It would still be possible that there was a systemic methodological error across the polling industry causing Yes to be overestimated, but the point is that it wouldn't be happening due to random sampling variation.) With only three polls, it's a bit less clear-cut - it's less likely that the margin of error is causing an illusory Yes lead than would be the case if there was only one poll, but it's still possible. So, yes, Ms Bruce technically had a valid point, albeit a weaker one than she probably thought.
But the wider issue here is the double-standard. There have been long spells over the last few years when the No vote in the polls was averaging out at around 53%, or 52%, or 51%. And it's scarcely been unusual for unionist politicians on Question Time and other programmes to make the point that "polls show the people of Scotland are still opposed to independence". Can you ever remember, even once, a host or interviewer jumping in to say "but it's within the margin of error"? Nope, me neither.
I gather Jackson Carlaw has been all over the airwaves today falsely claiming that the overwhelming majority of Scots don't want a second independence referendum. Did any interviewer correct him by pointing out that the YouGov poll shows that most Scots in fact want a referendum within five years, and that the Scot Goes Pop / Panelbase poll shows that most Scots think Brexit is a big enough change of circumstances to justify a referendum, and that Holyrood should legislate for one even without a Section 30 order? Probably not.