Friday is the anniversary of the Scottish independence referendum, so we’ve had a flux of Scottish polls over the weekend from YouGov, Panelbase and Survation. They mostly covered the same ground – should there be another referendum and how would people vote, and how people will vote in the Holyrood election next year.

Should there be a second referendum?

In the Survation poll 43% of Scots wanted another referendum within five years, 36% said in five years or more time, 20% never. Panelbase found 36% wanted a referendum within five years, 55% did not.

In YouGov’s survey they didn’t ask a timing question, but asked a couple of questions on whether there should be a second referendum. 40% of people thought that a referendum should be a once in a generation event, and there shouldn’t be another for many years. 29% thought that a referendum should not necessarily be once in a generation, and there was no reason why there shouldn’t be another one soon. Inbetween those two groups were 24% of people who thought a referendum should be once in a generation…but that if circumstances change, an early second referendum could be the right thing to do. One potential such change could be the European referendum (15% of people told YouGov they didn’t support a second referendum, but would do if Britain voted to leave the EU), but the SNP could obviously seek to persuade Scots of the case for a second referendum because circumstances had changed in other ways.

How would people vote in a second referendum?

All three polls had very similar findings – people would still vote NO in a referendum tomorrow, but by a smaller margin than they did a year ago. Survation had figures of YES 49%, NO 51%, Panelbase had figures of YES 47%, NO 53%, YouGov had figures of YES 48%, NO 52%.

Holyrood elections

The main story across all three polls was again the same, and largely unexpected: the SNP are headed for another landslide victory in next year’s Holyrood election.

• Survation/Daily Mail had constituency shares of SNP 53%, LAB 22%, CON 14%, LD 6%. Regional shares were SNP 42%, LAB 21%, CON 13%, LD 6%, GRN 11%.
• Panelbase/Sunday Times had constituency shares of SNP 52%, LAB 23%, CON 14%, LD 6%. Regional support stood at SNP 48%, LAB 22%, CON 15%, LD 6%, GRN 6%
• YouGov/Times had constituency support at SNP 51%, LAB 22%, CON 18%, LD 4%.Regional support stood at SNP 45%, LAB 20%, CON 18%, LD 4%, GRN 6%

All three had the SNP just over 50% in the constituency vote, and a little lower in the regional vote (which may be a Green party effect, the Scottish Greens don’t typically stand in the constituencies). The main differences between the polls are that Survation show the Greens doing significantly better than Panelbase and YouGov, and that YouGov have the Scottish Tories doing significantly better.

Tables for the three polls are here – Panelbase, YouGov (1),(2), Survation

This week’s YouGov results for the Sunday Times are up here. Topline figures are CON 35%, LAB 36%, LDEM 8%, UKIP 13% – a Labour lead of one point

28% now think the economy is in a good state, 36% a bad state. It’s still negative, but it’s creeping ever upwards – 28% is the highest YouGov have recorded since they started asking the question in 2010. Similarly on people’s own economic optimism, 19% expect their household finances to get better over the next twelve months, 31% expect them to continue to get worse – the net figure (“the feel good factor”) of minus 12 is still negative, but it’s the least negative YouGov have recorded since 2010.

YouGov also reasked some questions on Gaza that they initially asked a week ago. Public opinion has moved slightly towards the Palestinians – a week ago 23% said they were more sympathetic towards the Palestinians, now it’s 27%, but the broad picture remains the same: most people aren’t more favourable to either side, both sides are considered equally to blame for civilian casualties, and both sides’ actions are considered unjustified.

The Sunday Times also had a new Panelbase poll on the Scottish referendum. Topline figures are YES 41%, NO 48%. Excluding don’t knows that works out at YES 46%. Panelbase use slightly different wordings for their polls for the Sunday Times and for Yes Scotland, and it’s unclear whether they make a slight difference to the results or not. It doesn’t make much difference to the trend either way: since March Panelbase’s level of YES support has been steady at 46%-48%.

Following the intervention of George Osborne into the debate over Scottish independence and what currency an independent Scotland might use there has been an obvious interest in the next Scottish polls and what they might show. Will it have closed or widened the gap, or made no difference? Today there are two new Scottish polls asking about the referendum, sadly neither quite answer the question.

To get the less interesting one out of the way first, TNS BMRB have a “new” Scottish referendum poll, but the fieldwork was actually conducted between the 28th January and 6th February (I can only assume that the long time scale is because the poll was conducted face-to-face… though even then, the fieldwork was completed a fortnight ago). The figures in TNS’s poll are YES 29%, NO 42%, 29% don’t know – entirely unchanged from their previous poll in mid-January. Given the fieldwork was conducted prior to Osborne’s intervention though, this clearly doesn’t answer the question.

More relevant is a new Survation poll in the Daily Mail. This was conducted on Monday and Tuesday, so after Osborne’s intervention and at the time Alex Salmond was actively responding. Topline figures there are YES 38%, NO 47%, Don’t know 16%. Survation’s previous poll was showing YES on 32%, NO on 52%, so prima facie it looks as though there has been a significant shift towards YES. But there’s a caveat – last month Survation weighted their data by recalled 2010 vote, this month they’ve weighted by 2011 Holyrood vote. According to John Curtice Survation’s weighting last month knocked about five points off of Yes, their new weighting has not, raising the possibility that the difference could just be down to weighting. Realistically its not that simple – there is a random element in sampling, one sample will not be the same as the next and, therefore, weighting will have a different effect from one poll to the next, and it seems like a big difference to all be down to weighting to a different election. All we can really be confident in saying is that the two polls are not really comparable, so we should probably hold off on judgement – there are sure to be some more Scottish polls along soon. The tabs should be up on Survation’s site in about half an hour.

Scottish independence referendum polls so far are here.