Another poll released in anticipation of the first anniversary of the independence referendum, and another poll that suggests that the result would be different if the ballot were to be held now.
Last week Ipsos MORI released a poll showing that once Don’t Knows are excluded, 55% now say they would now vote Yes to independence, the first poll to put Yes ahead since March. Today TNS BMRB have issued a poll – conducted much like Ipsos MORI’s during the second half of August – which shows that 53% would now vote Yes (once Don’t Knows are excluded). While this result is not quite as spectacular as Ipsos MORI’s finding, the fact that two polls in succession have now put the Yes side clearly adds weight to the suggestion that Scotland has swung in favour of a Yes vote during the course of the summer.
There is still though a need for some caution before coming unreservedly to that conclusion. First of all, as in Ipsos MORI’s case, this is the first time since last September that TNS BMRB have asked people how they would vote in a second referendum. Thus there is no previous post-referendum measure from the company with which we can compare today’s result.
Equally, as was also true of Ipsos MORI’s poll, today’s poll did not ask its respondents how they voted in the referendum. If it had done we would have had direct evidence of how many people have changed their minds, while it would also have been possible for the poll to have been weighted so that how people say they voted in the referendum reflected last year’s result.
That said, today’s poll has been weighted, amongst other things, in line with the outcome of last May’s general election, and the correspondence between how people voted on that occasion and having voted Yes in the referendum is a close one. Most of those who say they voted Yes either say they backed the SNP in May (64%) or else that they did not vote or cannot recall what they did (24%). So the weighting by general election vote should have helped ensure that the poll is representative on the independence question. Moreover, it is clear from TNS BMRB’s tables that the weighting of the poll has in fact reduced the Yes percentage (from 55% to 53%).
Still, there is no doubt that we now eagerly await a poll from a company that has previously asked people how they would vote in a second referendum and which has been weighted to reflect last September’s verdict. If a poll conducted in that way were also to show the Yes side clearly ahead, then we will be as clear as it is possible to be that opinion has shifted.
In the meantime, today’s poll does actually register a slight fall in SNP support for next May’s Scottish Parliament election (and conversely an increase in Labour’s vote). On the constituency vote SNP support is now put at 58%, down four points in TNS BMRB’s previous poll in July, while on the list vote it stands at 51%, down three points. However, these drops simply bring TNS BMRB’s estimates of SNP support more in line with those of other pollsters, and they still leave the SNP well on course for another overall majority in May. It is unlikely Nicola Sturgeon will lose much sleep over that.