Although the last full-scale Scottish poll from TNS-BMRB was ludicrously misreported as showing a narrowing of the gap when it showed no such thing, it was nevertheless fair to say that it offered some small hope to Labour, because it meant there was at least one polling methodology that suggested they were vaguely within striking distance of the SNP. That comfort blanket has now been dramatically snatched away with the publication of tonight's new poll from TNS, which shows a much bigger SNP lead of 16% - very much within the range we've been used to seeing from all other pollsters apart from Ipsos-Mori.
Scottish voting intentions for the May 2015 general election (TNS-BMRB, 30th January - 22nd February) :
SNP 46% (+5)
Labour 30% (-1)
Conservatives 14% (-2)
Greens 3% (-3)
UKIP 3% (+1)
Liberal Democrats 3% (-1)
Is there a chance that TNS have tweaked their methodology, I wonder? They very foolishly introduced the discredited 2010 weighting procedure in their last poll, and if they've now reversed that decision, it would be extremely easy to explain the apparent movement towards the SNP. Assuming there hasn't been any methodological change, though, there are basically three possible explanations for tonight's result -
1) There has been a genuine big increase in SNP support over a short space of time (probably no more than a couple of weeks). That seems unlikely, given that the recent Survation poll showed a slight drop in the SNP lead.
2) The last TNS poll was a freakish outlier, and we're now seeing that the firm's methodology will typically produce results very similar to most other firms.
3) We're looking at an extreme example of normal sampling variation, with Labour being flattered by the last poll, and the SNP being flattered by this poll. If that's the case, the TNS methodology will typically produce a lower SNP lead than other firms, but the divergence won't be quite as great as the last poll suggested.
But whichever explanation you favour, it's murderously hard to reconcile this poll with there having been any upward trajectory for Labour of late. At best, their deficit has remained fairly static, which self-evidently isn't good enough with the clock ticking away fast.
TNS are of course the only pollster that still use the old-fashioned face-to-face data collection method, which is the reason we were always particularly excited when they produced a good result for Yes during the referendum campaign - it dispelled the worry that the only encouragement we were getting came from pollsters that were reliant on volunteer online panels. Fortunately we don't have to fret in that way at the moment, because by far the best results for the SNP in recent months have come from a "real world" pollster - namely Ipsos-Mori, who poll by telephone. Nevertheless, it's still reassuring to have proof that the face-to-face method is also capable of producing an enormous SNP lead.
UPDATE : As usual, it turns out that some of the TNS fieldwork is already quite a bit out-of-date - it started towards the end of last month, which (bizarrely) means that it overlaps to a small extent with their last poll. Admittedly it didn't conclude until four days ago, so we've seen a lot worse than that.
* * *
Having seemingly given up on the unionist parties getting their act together in time for the election, Kenny "Devo or Death" Farquharson has a brand new line in wishful thinking, which goes like this : Don't worry, it doesn't matter if the SNP win a truck-load of seats, because a mysterious forcefield will prevent them from having any actual influence. He's pinning his hopes on a couple of assumptions - a) that good old British fair play will not permit Labour to take office if they are the second-largest party, and b) if Labour are the largest party, they will choose to do a deal with the Liberal Democrats and the DUP rather than the SNP.
There are, you won't be surprised to hear, a few small flaws in this reasoning -
1) If Labour and the SNP hold a majority of seats between them, and yet the Conservatives are deemed the "winners" in the London media due to being the largest single party, something will have to happen for David Cameron to remain Prime Minister. The SNP will sure as hell be voting against the Tories' Queen Speech, so if Labour simply do the same, Cameron will be constitutionally obliged to resign, and the Queen will immediately invite Ed Miliband to form a government. The ONLY alternative is that Labour abstain on the Queen's Speech - which will be tantamount to voting in favour of a Tory government taking office. Are we really expected to believe their rank-and-file members (let alone their power-hungry MPs) will be sanguine about them doing that, particularly in an era of fixed term parliaments when they will be effectively condemning themselves to five more years in opposition?
2) The idea of a three-way Labour-Lib Dem-DUP pact is just barking mad. It's a complete non-starter. The only way the Liberal Democrats will countenance a formal deal is if it produces a stable majority, thus ensuring the concessions they win in coalition negotiations are worth the paper they're written on. Most likely, that means they will only enter a deal with Labour if the two parties between them can command an absolute majority. If there is to be an additional partner, though, the DUP will be just about at the bottom of anyone's list of preferences. They can't realistically be relied upon to stick to anything beyond a very short-term deal, but that's not the half of it - if Kenny genuinely doesn't understand the stigma they carry, he's beyond all hope.
3) There's no reason whatever to assume that Labour, if they're the largest party, will have an alternative to dealing with the SNP available to them even in theory - the arithmetic may well ensure that the only other option would be a grand coalition with the Tories.