Peachy Panelbase poll puts the SNP on 53%

As far as I can see, this is the first Panelbase poll to measure Holyrood voting intentions since the one commissioned by the SNP themselves in late September/early October - less than two weeks after the referendum.  (That was the weird poll that used a completely different weighting scheme for the Holyrood and Westminster questions, thus leading us to severely underestimate the extent of the SNP surge.)

Scottish Parliament constituency ballot :

SNP 53% (+11)
Labour 22% (-5)
Conservatives 15% (n/c)
Liberal Democrats 5% (n/c)
UKIP 2% (-3)
Greens 2% (n/a)

Scottish Parliament regional list ballot :

SNP 48% (+11)
Labour 21% (-6)
Conservatives 15% (-1)
Greens 6% (-3)
Liberal Democrats 5% (n/c)
UKIP 2% (-2)

By any standards, a stunning advance for the SNP over the course of the last nine months, but at least some of this is the story we already know - most polling firms agreed there was a secondary surge for Nicola Sturgeon's party a few weeks before the general election, over and above the post-referendum gains.  So what isn't clear is how much of the 11% increase in this poll only occurred after the general election.  We won't get a definitive answer to that question from Panelbase, but if Survation and TNS are correct, it may have been quite a bit.

Unsurprisingly, Panelbase have produced the lowest SNP lead of the three firms that have been active in Scotland since the general election.  (It feels very odd to call a 31% lead "the lowest", but it is!)  This continues the pattern we've seen in recent months, with the most Yes-friendly pollster from the referendum proving paradoxically to be one of the most conservative on the subject of the SNP's own fortunes.  This may be partly explained by Panelbase's use of online fieldwork - in the run-up to May 7th, it was TNS (face-to-face fieldwork) and Ipsos-Mori (telephone fieldwork) that gave the SNP their biggest leads.  Alternatively, it may be partly to do with political weighting - TNS haven't been weighting by recalled referendum vote, and Ipsos-Mori don't weight by recalled vote at all.  That sets both firms apart from online pollsters like Panelbase.

At this very early stage of the Holyrood contest, it also looks as if Panelbase are slotting in as a less SNP-friendly firm than their online counterparts Survation, but the difference isn't huge - just 3% on both ballots, which could even be explained by the margin of error.

The Greens will once again be disappointed with their showing, although their 3% drop on the list should be treated with a slight touch of caution, because this poll is not directly comparable with the last one, in which the Greens weren't offered as an option on the constituency ballot.  There's a case to be made that the Greens do a bit better on the list when people aren't given the option of splitting their two votes "in reverse".

There are also numbers on the EU referendum, from both sides of the border -

Should the United Kingdom remain a member of the European Union?

Respondents in England :

Yes 49%
No 51%

Respondents in Scotland :

Yes 66%
No 34%

I find that quite an odd result, because as you may remember a Panelbase poll commissioned by Wings a few months ago showed only a very narrow Yes lead in Scotland.  The disparity between the two countries is no great surprise, but you'd still expect the trend to be broadly the same - ie. if it's the Greek situation that has helped No in England, there's no reason why it wouldn't also have helped No in Scotland as well.  I'll be interested to see if the fieldwork dates for the two sets of respondents are the same, or more or less the same.

Whenever a telephone poll comes out suggesting an "unassailable" lead for Yes across the UK, I always caution that you can't just look at the most recent poll and assume that it trumps everything that's gone before from other firms using different data collection methods.  The same principle applies now that we're seeing a No lead in England for the first time in a while - it's likely to have come about partly because of Panelbase's online method, and it doesn't negate the telephone polls that have shown such a radically different picture.  We still have no real idea what the true state of play is.

UPDATE : Although the datasets for the poll aren't out yet, a little more information has emerged.  It turns out it's not actually a new poll, but instead further questions from the Sunday Times Panelbase poll that was initially published last weekend.  That means the fieldwork for the two samples was indeed conducted simultaneously, but quite a while ago - well before Greece's humiliation, so that can't be the explanation for the strong showing for No south of the border.  And the "English" sample was in fact an England and Wales sample (not "rUK", because Northern Ireland wasn't polled).

*  *  *

I was quite disturbed to see BBC Scotland's Douglas Fraser make this explicitly devo-sceptic comment on Twitter a few hours ago, albeit in a personal capacity -

"I see @theSNP is back to the campaign for Holyrood control of BBC Scotland. To do what with it? Anyone?"

That appears to be a continuation of the attitude problem we saw from the mainstream media during the referendum, with entirely different standards being applied to the two sides of the debate.  London control of broadcasting is seen as neutral, normal and benign, whereas Scottish control is abnormal, and can be assumed to be hostile and harmful until we hear a damn fine explanation.

Why doesn't Douglas ask why London Tories want to KEEP control?  What do THEY want to do to BBC Scotland, and why should they be allowed to do it?

The reality, of course, is that devolution of broadcasting is not really about direct Holyrood control over the BBC, but about BBC Scotland securing meaningful autonomy from its London masters.  That will never happen for as long as broadcasting remains a reserved matter.  And yes, there would be a degree of democratic oversight at a Scottish level of the new arrangements - why the hell shouldn't there be?