A distinguishing feature of the polls that ICM conduct for Scotland on Sunday is that they repeatedly ask many of the same questions each month. As a result, apart from academic exercises such as the British Election Study, they provide us with a unique insight into how much difference the referendum campaign is making to attitudes to some of the key issues at stake. The answer appears to be very little indeed.

The headline voting intention figures in this poll are Yes 34% No, 45%, while Don’t Knows are on 21%. No are up two while Yes are down two on last month, but the result is almost identical to that in May (Yes 34, No 46). Once the Don’t Knows are excluded Yes are on 43%, down two on last month but up a point on May. These changes look no more than the kinds of random variations to which all polls are subject, even when in reality little or nothing has changed at all, while today’s reading leaves our poll of polls unchanged.

Indeed, this poll even fails to confirm the impression conveyed recently by both TNS BMRB and Survation that the numbers of undecided voters may be beginning to fall. ICM’s latest 21% figure is exactly the same as it was last month.

Even more remarkable is the lack of movement in people’s views about some of the key issues in the campaign. Now 35% think that independence would be good for the economy, unchanged from last month and exactly the same as it was as long ago as January. At 33% the proportion that believe there would be less inequality in an independent Scotland has slipped just a point on last month, and is no different now from what it was in February.

Equally, 63% now say that in the event of a No vote the Scottish Parliament should become primarily responsible for taxation and welfare benefits in Scotland, up just two points on last month and down just one on the position in January.

However, in one rare exception to this picture of stability, there are some signs that the possibility that the No side might lose a crucial proportion of its vote because of fears that the unionist parties will fail to deliver more devolution has receded somewhat.

First of all 56% of No voters now actually think that the powers of the Scottish Parliament would be increased if Scotland votes No, six points higher than at any time since ICM first asked this question in March. Secondly, no less than 86% of those No voters who say they want more devolution say they will still vote No even if they were doubtful that more powers would be on their way. That figure too is an all-time high (previously, 81%) since the question was first posed as long ago as last September.

But if the No vote has indeed hardened somewhat, this simply suggests that it could prove even more difficult for either side to shift the balance of opinion in the next two months than it has been in the referendum campaign so far.  As the campaign behind in the polls  that of course is unwelcome news so far as the Yes side is concerned.

However, the Yes side might argue that as there are apparently still plenty of people who have yet to make up their minds after all, they can still hope to win in September by winning over the undecideds, many of whom Yes believe are gradually edging towards backing independence.

However some doubt is cast by today’s poll on the degree to which the undecideds are inclining towards a Yes vote. The Yes side’s claims about how voters are gradually moving in their direction have been based in part at least on what they say is the pattern of responses they have obtained when voters have been asked to indicate their attitude towards independence on a scale from 1 to 10 – with 1 meaning they are completely against independence and 10 that they are completely for it. The detailed evidence upon which this claim is based has never been published by Yes Scotland, but the question has now been replicated in today’s poll.

Two points stand out. First, not that many voters are apparently on the cusp of becoming potential Yes supporters.  Just 7% place themselves at point 5 on the scale. Even if every single one of them were to be won over (a demanding proposition), the Yes side would still be a point away from victory.

Second, not only are as many as three-quarters of undecided voters willing and able to put themselves on this scale, but, as one might anticipate, nearly two-thirds of those that do put themselves somewhere towards the middle of the scale, that is somewhere between 4 and 7.  However, the balance of inclination amongst the undecideds is clearly tilted towards opposition to rather than support for independence. Twice as many put themselves at between 1 and 5 on the scale as position themselves at between 6 and 10.  It seems that even on the evidence of its own measure, the Yes side still have a lot of persuading left to do.

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.