Merciless electorate metes out more misery to Kalamity Kez as bombshell Survation telephone poll puts Labour third

What I presume will be the last pre-election edition of Survation's regular series of polls for Cleggy and the Vow-Meisters has been released tonight.  As they've done before, the Record have chosen to splash out on a much more expensive telephone poll just before election day.  That's a very useful exercise, because it means we no longer have all our eggs in one basket as far as phone polling is concerned (the basket being Ipsos-Mori).  But it also means that the percentage changes listed by David Clegg on his Twitter feed, showing among other things a seemingly dramatic drop in the Green list vote, are completely meaningless.  Even if other aspects of the methodology have remained the same, you just can't make any sort of direct comparison between the last Survation online poll and tonight's phone poll.

Constituency ballot :

SNP 49%
Labour 21%
Conservatives 19%
Liberal Democrats 7%

Regional list ballot :

SNP 43%
Conservatives 20%
Labour 19%
Greens 7%
Liberal Democrats 6%
UKIP 2%

As far as the battle between Labour and the Tories is concerned, the new figures are uncannily similar to the recent Ipsos-Mori telephone poll, which also had Labour just ahead on the constituency ballot, but trailing on the list.  This underlines the foolishness of journalists who seized upon a single Panelbase poll at the weekend as proof that Labour had more or less won the fight for second place.  However, it remains the case that there is a bigger question mark over list polling than constituency polling, and that there tends to be a strong correlation between a party's fortunes on the constituency ballot and the list ballot - meaning, in my opinion, that this poll still points to a Labour advantage, but only just.  If it really is as close as it appears, differential turnout could decide it - and although that's factored into the headline numbers of the polls, there's no way of knowing for sure that it's being factored in accurately (as the events of last year demonstrated).  Tory supporters are generally more motivated to vote, but Scottish Parliament elections also tend to be 'away fixtures' for the Conservatives.

My own feeling remains that Labour will hold on to second place, probably with a bit to spare, but it's getting harder and harder to claim that's an assessment supported by compelling evidence from the polls.

Since I wrote the opening paragraph of this post, I've had a chance to look at the Survation datasets, and it turns out that data collection method isn't the only methodological factor that differentiates this poll from the last online Survation poll - the firm have finally addressed the long-running concerns over the way they pose the regional list question, which until now has seemed to directly lead a minority of respondents to wrongly think they're being asked for a second preference, thus artificially boosting the reported vote shares for the Greens and UKIP.  The change that's been made is subtle, but it could well be enough to explain the fact that the Greens are 4% lower than in the last Survation poll.  That said, arguably the real benchmark for the Greens should be the impressive 10% they received in the Ipsos-Mori phone poll, so from that point of view tonight's findings are still disappointing for them.  OK, if they actually get 7% of the vote on Thursday, they stand to gain several seats - but we know from past elections that late polls have a habit of overstating the Greens' support.  As things stand, I wouldn't be surprised if the Greens win only two or three seats, but neither would I be surprised if they win seven or eight.

This poll is also another massive warning shot for any SNP supporter who is tempted by the idea of "tactical voting on the list", because it's only the third poll in the last twelve months to put the SNP below 50% on the constituency ballot - and two of those three polls have come in the space of the last 48 hours.  Two of the five regular pollsters are now saying the SNP constituency vote is just 4% higher than it was in 2011, when the party required a minimum of twelve list seats for an overall majority.  We're now well into territory where only a relatively small polling error, or a relatively small late swing (or a bit of both) could cost the SNP their overall majority if their own supporters abandon them in sufficient numbers on the list ballot.  I've said it before and I'll say it again for the benefit of the likes of Kevin Williamson, who was mocking my stance before this poll was released - the idea that "everyone knows" the SNP will win a majority on constituency seats alone, and that there is polling evidence that can convincingly support that contention, is fatuous.  It may happen, but it's very much in the balance.

It'll be interesting to see (and we may have to wait 24 hours to find out) whether an independence question was asked in this poll.  If by any chance it was, it would - incredibly - be only the third independence poll conducted by telephone since the referendum in September 2014.  The previous two were both carried out by Ipsos-Mori, and both showed a Yes lead - so we really need to find out whether that's a phone effect or an Ipsos-Mori effect.  (There are undoubtedly people who already know the answer to that question, because it's an open secret that Survation have been commissioned on multiple occasions to conduct private telephone polls on independence and other subjects - although we don't know who the client was.)

*  *  *

SCOT GOES POP POLL OF POLLS

Constituency ballot :

SNP 50.7% (-0.6)
Labour 21.2% (+0.5)
Conservatives 17.7% (+0.4)
Liberal Democrats 6.2% (n/c)

Regional list ballot :

SNP 44.7% (n/c)
Labour 19.8% (+0.3)
Conservatives 18.3% (+0.3)
Greens 7.7% (-0.6)
Liberal Democrats 5.5% (-0.2)

(The Poll of Polls is based on a rolling average of the most recent poll from each of the firms that have reported Scottish Parliament voting intention numbers over the previous three months, and that adhere to British Polling Council rules. At present, there are six - Panelbase, Survation, BMG, YouGov, TNS and Ipsos-Mori. Whenever a new poll is published, it replaces the last poll from the same company in the sample.)

Explanatory note : As a fond tribute to the mainstream media's restrained take on the GERS report, Scot Goes Pop headlines will feature 50% bonus hysteria for an indefinite period.