It’s month three of our rolling sequence of polling, conducted as usual by Survation, in partnership with the Daily Record and Dundee University’s 5 Million Questions. Last month’s figures are here, and the Record have the indyref polling story here (in brief: Yes 44, No 56, i.e. no change on last month, but they also got numbers suggesting Salmond is unpersuasive).

This month, like last month, I’m comparing vote shares to the previous month’s figures: but seat numbers are still shown as the change on the 2011 result. Seat projections are from Scotland Votes again. The major caveat with the seat projections is that UKIP are shown at a point where they would almost certainly win a handful of regional list seats, but the Scotland Votes site doesn’t include them, so it is unclear at whose expense they would come. The model does show one independent: it’s not clear if this is a legacy result based on Margo or a seat based on the large “others” list score, mostly UKIP. My guess is they’d pick up perhaps five on this result, and a rough guess would be that each of the existing Holyrood parties might win one fewer each than shown here. So here are the figures, with that caveat.

Parties Constituency Region Total
Vote share (+/-) Seats (+/-) Vote share (+/-) Seats (+/-) Seats (+/-) %
SNP 43.7 (-1.2) 49 (-4) 39.1 (-0.7) 12 (-4) 61 (-8) 47.3
Labour 31.7 (-0.4) 19 (+4) 26.4 (+1.3) 18 (-4) 37 (±0) 28.7
Conservative 15.4 (+1.9) 3 (±0) 11.3 (-0.9) 11 (-1) 14 (-1) 10.9
Liberal Democrats 4.9 (-0.8) 2 (±0) 6.1 (-2.9) 4 (+1) 6 (+1) 4.7
Scottish Greens 1.0 0 (±0) 8.7 (+1.4) 10 (+8) 10 (+8) 7.8
UKIP 2.1 0 (±0) 7.1 (+3.1) 0 (±0) 0 (±0) 0
Others 1.2 (+0.5) 0 1.3 (n/a) 1 (±0) 1 (±0) 0

The SNP remain in unquestioned pole position: a small dip in both votes still leaves them winning more than sixty seats overall. As per last month, if this were repeated they could either return to an even more comfortable version of minority rule model they used from 2007 to 2011, or they could pick any one of the three smaller parties as a coalition partner. On these numbers, especially if we assume UKIP would take one further seat from each, the Lib Dems would provide them with the narrowest possible majority, something the SNP might well be reluctant to contemplate.

Labour have edged back up a little on the list and fallen less far on the first vote, and as a result would pick up four constituencies from the SNP while losing four of their list seats. No overall progress since they bore the brunt of the “tartan steamroller” in 2011, in short. Or slightly worse, if the ‘kippers were to take one off them. If I were advising Labour, I’d say it’s clear there’s something wrong with the team, or with the policies, or with the message, or with a combination of all three.

The Tories are broadly holding their position since 2011 here, with just one more seat lost. Managed decline, one might say. And the Lib Dems would win just one more than 2011, with their list vote showing the sharpest decline over the last month.

Which brings me to my favourite part of this result: the strongest Green list vote we’ve seen since this polling sequence began, indicating a Green group five times larger than that which currently sits at Holyrood, primarily because Labour would no longer be so substantially under-represented in the constituencies.

My guess is 10 Green MSPs would mean two each in Lothian, Glasgow, and Highlands & Islands, plus one each in North East, Mid Scotland & Fife, South, and West, although one in Central or even a third in Lothian at the expense of the second H&I seat is another possibility (again, as with the Coalition parties above, note the UKIP caveat here). A result like this would lead to the biggest celebration the Greens have ever had, and a strong hand in the session to come.

Whichever way the referendum goes in September, this result would also see a pro-independence majority at Holyrood almost as large as that elected in 2011, although clearly Holyrood polling this far out is pretty damn speculative, especially over that particular event horizon.