Making sense of the polls

Published: 17 October 2013
Author: John Curtice
In theory this week’s SNP conference is a chance for a party in power to take stock at the mid-point of its current term in office. It should be able to look back with some satisfaction at polls that still largely put the party ahead in Scottish Parliament vote intentions and look forward with optimism to how it might win a third term in office in just over two years’ time.
 
But of course, this will not be the theme that dominates the debate on the conference floor or the discussion in the hotel bars. Rather it will be how the party, together with its allies in the Yes campaign, can win the votes needed to deliver the dream that constitutes the SNP’s raison d’être, an independent Scotland. And alas for nationalist activists, here the polls have much less cheery news to impart.
 
The recent onset of the ‘pre-anniversary’ of a year to go to the independence referendum stimulated a flurry of polls on how people proposed to vote next year. All in all some ten polls of voting intention have been published during the last two months – full details are available at whatscotlandthinks.org. To make most sense of these, it is best to take out the don’t knows and won’t says. When we do so those ten polls have on average pointed to a 38% Yes vote in response to the proposition, ‘Should Scotland be an independent country?’, with 62% indicating they will vote No. 
 
Such figures suggest the Yes side still have plenty of work to do. Moreover, there is no sign of their having made any progress in persuading more Scots to join them in their cause. That 38% figure is exactly the same as the average Yes vote in all of the polls conducted between February and May this year. It is actually somewhat down on the 41% average Yes vote in the polls conducted last year in response to the referendum question originally proposed by the Scottish Government, ‘Do you agree that Scotland should be an independent country?’.
 
Yet there was that one recent poll that put the Yes side ahead, albeit by the narrowest possible margin. However, that poll – commissioned by the SNP themselves – was widely criticised for having prefaced the question on how people would vote in the referendum with a couple of questions that seemed to lead people into saying Yes. But even if we leave that poll to one side there have been some big – and persistent – differences between the polls produced  by different polling companies.
 
At one end of the spectrum lie Ipsos MORI, whose most recent poll put the Yes vote at 32%. That was very much in line with their previous reading in May of 34%.  At the other end are Panelbase who estimate the Yes vote to be as high as 44% - very different from Ipsos MORI’s reading, but again almost exactly the same as their own previous findings, including one of 45% in May. 
 
In short the polls point to an apparent remarkable stability in public opinion – yet nevertheless leave us highly uncertain about what exactly are the relative sizes of the Yes and No forces.  We apparently might be looking at a contest in which the Yes side are outnumbered by as much as two to one, thereby making victory in the referendum looking a remote prospect indeed. Or we could be faced with quite a tight battle in which all that the Yes side need to do is to add another five or six points to their existing share of the vote.
 
Either way, just how easily the Yes side might be able to make up the difference depends also on how many people still have to make up their minds.. Despite the differences between them in their estimates of the size of the Yes and No votes, the different polling companies are largely in agreement in finding that less than one in five of their respondents say that they do not know how they will vote. That is far fewer than there were just a matter of two to three months before the May 2011 referendum on changing the Commons voting system, and is almost undoubtedly a lower proportion than the Yes side would like
 
However, one company, TNS BMRB, have put the level of Don’t Knows in their two most recent polls at no less than 28% and 31%. The likely explanation for these much higher figures is that, in contrast both to most other polling companies and their own previous practice, TNS BMRB now ask their respondents how ‘do you intend to vote’ in the referendum next year rather than how they would vote now. Perhaps quite a few people feel that they have an idea what they would do if presented with the referendum ballot paper now, but accept that they might yet be open to persuasion by the other side. The Yes side will certainly hope that this means there are significant numbers of Scots who they could still win over.

It is said that Greenland could be independent by 2037. One of the challenges for this is the number of powers that… https://t.co/SNCo1VDUzB

6 hours ago

Lots of comparisons can be drawn between Scotland and Greenland, including being taught and speaking Greenlandic and Gaelic.

6 hours ago

The most important thing for the Department of Constitution in Greenland is public engagement and speaking to all a… https://t.co/RGkgl2UNO5

6 hours ago

Natuk Lund Olsen starts our seminar ‘The debate on independence in Greenland’ providing a history of Greenland to t… https://t.co/oyoTYSz6kc

7 hours ago