The poll from TNS BMRB released today has been widely reported as further evidence that the No lead is narrowing. In truth, it would appear to be further evidence that the No lead has stabilized around the narrower level it has now been at for some weeks.

First, though, a note about the timing of the poll.  Interviewing actually finished a fortnight ago, well before last weekend’s SNP conference, so it certainly does not tell us anything about what impact that event might have had. In fact the poll is older than both of the two most recent Panelbase polls as well as the most recent one conducted by Survation.

Those three polls all suggested that little had changed in the last month or so. The same is true of TNS BMRB’s poll too.  True, at 29% the Yes vote is up one point on the company’s previous reading, while the No tally is down a point to 41%.  When the Don’t Knows are excluded the Yes vote is on 41%, up one point.

But in themselves such movements are certainly statistically insignificant. Moreover, they simply restore the Yes tally of 41% back to where it was in TNS BMRB’s poll in February, conducted just before the currency intervention.  True, no TNS BMRB poll has put the Yes vote higher than 41%, so today we have further confirmation of the progress Yes made during the winter – but there is no sign here that their advance is continuing further.

Meanwhile, of course we have a reminder (if indeed we needed it) of the continuing substantial divergence in the estimates produced by different companies. In the last fortnight Panelbase have twice made the Yes tally 47%, while now TNS BMRB suggest it is no more than 41%.  Both companies agree that little or nothing has changed in recent weeks, but they paint a very different picture as to how close the race might be.

One of the continuing distinctive features of TNS BMRB’s polls has been a much higher level of Don’t Knows than in other company’s polls.  Typically (and as in today’s poll) TNS BMRB find that 30% say they do not know what they will do, twice the level usually found in other polls. However, today’s poll casts a little further light on this difference.

In response to a new question on the extent to which people have made up their minds, only 15% say that they have not made any decision at all about how they will vote – in other words roughly the proportion of Don’t Knows reported by most other companies.  However, in addition there are another 14% who say that ‘I have an idea of how I will vote, but I have not yet made a final decision’.  Most of these 14% did not give TNS BMRB any indication of whether their ‘idea’ was to vote Yes or No.

Evidently when asked as they are by TNS BMRB what they intend to do in September, this group of voters say they Don’t Know. Perhaps when other companies ask similar voters what they would do if the referendum were held now (or what they think they might do in September) respondents are answering with the ‘idea’ they have of how they will vote.

In any event it is clearly a mistake to think of an undifferentiated  group of undecided voters who are clearly set apart from their fellow citizens who have all made up their minds. Rather voters lie at various points on a spectrum of indecision. Just over half (52%) have definitely made up their minds. The rest lie somewhere between having no idea at all what they will do and having a fairly firm inclination but might just be willing to change their minds.  What remains to be seen is whether either side can secure any further success in persuading them to do so.

About the author

John Curtice is Professor of Politics at Strathclyde University, Senior Research Fellow at ScotCen and at 'UK in a Changing Europe', and Chief Commentator on the What Scotland Thinks website.