You might remember that we heard from a couple of people at the weekend that a full-scale Scottish poll from Panelbase was underway. I wondered if it might be an internal poll for the SNP, in which case there was a chance the results would never see the light of day. Thankfully, it turns out to have been commissioned by Wings Over Scotland. It's hugely valuable, because unless we get another surprise poll later today, this is our only opportunity to see what impact on public opinion (if any) there was from the revelation last Thursday night that Miliband might be prepared to help Cameron stay in power in some circumstances. Panelbase's fieldwork started on Friday, and concluded only today.
Scottish voting intentions for the May 2015 UK general election (Panelbase, 1st-6th May) :
SNP 48% (n/c)
Labour 26% (-1)
Conservatives 14% (-2)
Liberal Democrats 5% (+1)
UKIP 3% (n/c)
Greens 2 (n/c)
It looks as if there has been little or no change as a result of the events of Thursday. The 1% increase in the SNP lead is obviously well within the margin of error, although it's enough to break yet another record. As I used to say during the referendum campaign, even the smallest of changes is potentially of interest if it takes a pollster outside its previous 'normal range'. However, this result is fully consistent with the trend we saw from YouGov at the weekend - the most plausible narrative is that there was a second telling SNP surge a few weeks ago, and that the position has remained relatively stable since then.
Although there may be no disagreement between pollsters over the trend, what we don't have is agreement on the extent of the SNP's lead. Believe it or not, Panelbase remain the most Labour-friendly of all the pollsters that have been active in this campaign (with the exception of ICM, who haven't reported for quite a while). The 22 point lead in this poll is well short of the 32-34 point leads reported most recently by the two most SNP-friendly firms, TNS and Ipsos-Mori. The divergence we're seeing between the firms is considerably bigger than anything that is going on at Britain-wide level - if that wasn't the case, we might be going to bed tonight not having a clue whether to expect a Conservative majority government, a Labour majority government or a hung parliament.
Given that even the Panelbase result is projected to give the SNP no fewer than 53 seats out of 59, does it even matter which firm is closest to the truth? Of course it does. There could be a very late swing back to Labour. There could be very heavy tactical voting. There could be a 1992-style (or Israeli-style) polling disaster in which all firms turn out to have overestimated one party, and underestimated another. If any of these things are true, the SNP could end up winning significantly fewer seats if Panelbase is nearest the mark. But if Ipsos-Mori turns out to be the best performer, the SNP are already so far out of sight that the result is effectively assured.