DODGY POLL ANALYSIS KLAXON + + + Yes, it's Fraser Nelson again + + + DODGY POLL ANALYSIS KLAXON

It almost seems absurd to home in on this relatively minor detail from an epically offensive article from Fraser Nelson in the Telegraph, but I'll do it anyway -

"The Union does not, yet, look in mortal danger. Barack Obama's hopes for a "united" Britain are likely to be answered; the Yes campaign, which was level pegging seven weeks ago, is now 18 points behind."

Er, nope. The poll showing the Yes campaign 18 points behind (actually it was more like 17 points on the unrounded numbers) was of course this week's Ipsos-Mori poll, which in fact showed the No lead at its lowest level of the campaign so far by some distance. Since the previous Ipsos-Mori poll three months ago, the No lead has plummeted by 7.2% - meaning that almost a full one-third of the gap has been wiped out. The point being that Ipsos-Mori are consistently the most No-friendly of the six BPC-affiliated pollsters.  All of the polls showing Yes at almost level-pegging have been ICM and Panelbase polls, which use a completely different methodology from Ipsos-Mori and are therefore not comparable.

If it was anyone else but Fraser Nelson making such a silly statement (with the possible exception of a certain Cornish sex memoirist), I'd assume he understood the reality of the situation perfectly well and was just indulging in propaganda. But I must admit it's perfectly conceivable that Nelson genuinely doesn't have a clue.

He also gives another outing to the hoary old myth that there is polling evidence that Scotland's schoolchildren (including 16 and 17 year olds who will have the vote for the first time) are opposed to independence. This seems to be entirely based on Jan Eichhorn's Edinburgh University research that has been so comprehensively discredited. As it happens, that research has just been updated and shows a mammoth 8% swing to Yes among young people since last year. Whether that swing is genuine depends on whether Eichhorn's team have got their act together and corrected the almost laughable No-friendly nature of their sampling - if they have, then obviously the figures will not be directly comparable. But if the methodology is identical to last year's survey, the No campaign have good reason to be deeply concerned. And that's notwithstanding this piece of unintentional comic genius from my popular MSP namesake -

"Under 18s poll from Edinburgh University has NO 64% YES 36%. That was before Obama endorsement of the UK staying united."

Which is right up there with -

"And this poll was BEFORE John Barrowman put on a Scottish accent and called Alex Salmond fat."

The headline figures are of course utterly meaningless - there's no way on Earth that 16 and 17 year olds are almost 2-1 against independence, because if they were, credible polling companies like Ipsos-Mori would not be routinely finding the Yes campaign either ahead or only slightly behind with 16-24 year olds.

In case you're wondering what was so offensive about Nelson's article more broadly, it's encapsulated in this passage -

"A similar artists’ bus party was convened before the 1997 devolution referendum – and, back then, I was sold. The argument, then, was simple: countries are best governed by people living in the country, not those living in another country. Alex Salmond is making the same, calm, common-sense proposition now – but with one big difference. It is demonstrably untrue."

It doesn't get more explicit than that - Nelson is arguing for a No vote because Scotland is better off being governed by people who live in another country. His explanation is that the Scottish electorate have time and time again voted against right-wing Blairite and/or Thatcherite "solutions" to the country's problems in policy areas that are covered by devolution - a pattern which would automatically extend to everything else if we were independent. This is not so much a "too wee, too poor, too stupid" argument as it is a plain old "too stupid" argument - we can't trust ourselves to vote Tory, therefore the only solution is for us to contract out our choice of government to a country that can be trusted to impose the Tories on us. But the million dollar question is : are we stupid enough to make that decision even though Nelson has just spelled out the supposed logic for doing so with quite such brutal honesty?

* * *

I see that the UK government have ignored Rev Stuart Campbell's FOI request to be told whether the results of the notorious secret Ipsos-Mori mega-poll were shared with the No campaign. It seems to me the biggest issue here is that both campaigns need to conduct private polling, which costs good money - and that ought to be coming out of the campaign funds. If the UK government are effectively saving the No campaign £50,000 by funding a poll with taxpayers' money and then passing the results on to McDougall Central, that is (not to put too fine a point on it) absolutely bloody outrageous. I keep thinking back to Gary Gibbon's knowing remarks a few weeks ago about how Better Together were characterising the results of "their" much larger-scale private polling - was that in fact the UK government's mega-poll?