Peter Cuthbertson is a Conservative activist and is Director of the Centre for Crime Prevention.

Scotland is on the brink of independence, the headlines tell us. Our leading commentators ask what Scottish independence will mean for the rest of the UK. Thankfully for those of us who value the Union, there is no need to panic. The overall picture becomes clearer by the day: Scotland will vote No.

It’s not only that across time and history, voters in referendums tend to fall behind the status quo unless there is an overwhelming majority for change.

It’s also – as I show below – that the No camp has a big lead in almost every opinion poll. It’s that this lead is so consistent. It’s that the Don’t Knows would need to break overwhelmingly in favour of Yes. It’s that when the Don’t Knows do move, it’s to the No camp in overwhelming numbers.

I’ve looked at the opinion polls since January 2012, mostly taken from Anthony Wells’ excellent UKPollingReport – 63 polls in total.

A few basic figures:

  • The Yes vote averages 34 per cent, the No vote 49 per cent
  • About three polls in four put the Yes vote between 29 per cent and 38 per cent
  • About three polls in four put the No vote between 44 per cent and 56 per cent

34 per cent Yes to 49 per cent No equates to 41 per cent/59 per cent if you remove the Don’t Knows – a full 18 point gap.

This would hardly be the end of all hope for the Yes camp if they could point to trends moving strongly in their direction. The poll is four months away after all.

But they can’t. The thick lines in the graph below show the trends in the No, Yes and Don’t Know votes over the last 2.5 years (with the dotted lines representing each individual poll). The lines scarcely slope.


With such a large and consistent lead for No, it is a Herculean task for the Yes camp to out-poll the Nos. But perhaps they can win over a big enough majority of Don’t Knows on the day?

Unfortunately for them, the No lead is so great that in 24 of the 63 polls (38 per cent) even if 100 per cent of the Don’t Knows voted Yes, the No side would still win. In only 30 per cent of polls would the Yes side win with less than a 3-1 majority of Don’t Knows. In only one poll could they win with less than half the Don’t Knows – and that one was dubious as it was commissioned by the SNP and preceded by rather leading questions such as “Do you agree or disagree [that] Scotland could be a successful independent country?”.

Let’s charitably assume the Yes camp needs less than 100 per cent of Don’t Knows on the day. Could they do this well among Don’t Knows? Again, the volume of polling evidence is a guide to who does better when the Don’t Know vote drops. (There is not a ‘Don’t Know’ option on the ballot, after all, so it has to fall to 0 per cent on the day).

Only 16 per cent of the fluctuation in the Yes vote is explained by fluctuation in the Don’t Know vote (the r-squared is 0.16). But the majority of fluctuation in the No vote is explained by fluctuation in the DK vote (an r-squared of just over 0.50).

Another way of looking at this is to divide the polls into thirds by how high the Don’t Know vote is. In Group I (20 separate polls) the Don’t Know vote was 13 per cent or less. In Group II the Don’t Know vote is between 13 per cent and 18 per cent. In Group III (21 of the polls) it was 18 per cent or more.

The average Yes vote in Group I is 34 per cent – 2 per cent more than the average Yes vote in Group III. But the average No vote is a full 10 per cent higher (54 per cent instead of 44 per cent). When the Don’t Know vote drops, it breaks down about 5 to 1 in favour of a No vote.

So not only does the No camp have a strong and consistent lead – so much so that if the Don’t Knows don’t vote at all, the No side wins hands down. It’s that Yes needs an overwhelming share of Don’t Knows despite being far worse than the No camp at winning over the undecideds.

The Union is safe, at least for now. If ever there were a sure thing in politics, it’s this.