I don’t think we’ve had any GB polls today (not doubt there will be the usual flurry for the Sunday papers tomorrow), but we did get a YouGov Scottish poll for the Times, their first of the campaign.

Topline voting intentions were SNP 41%, CON 28%, LAB 18%, LDEM 7%. At only 41% the SNP are lower than in the Survation and Panelbase polls last weekend, but YouGov also show the Conservatives doing significantly less well than that Panelbase poll that had them on 33%. If these figures were repeated at the general election then the Conservatives would take seven seats from the SNP, the Liberal Democrats would take two.

Voting intention on Scottish independence stands at YES 45%(+2), NO 55%(-2) (and that’s without 16 and 17 year olds, so reality might be slightly more pro). Asked about a second Indyref, YouGov asked both about the principle of it and the specific timing – on principle, 42% of Scots want a referendum in the next five years, 51% do not. Asked about specific timing, a referendum before Britain leaves the EU is marginally more popular than one afterwards: 37% would support a referendum once negotiations are complete but before Britain actually goes, 35% would support one after Brexit.

Full tabs are here.

Some Bregrets?

The tables for last night’s YouGov/Times poll are now up here.

The result that has got the most attention is the question on if people think Britain was right or wrong to vote to leave the EU. 43% said right, 45% said wrong. YouGov ask that question most weeks and almost invariably it shows people either think Britain made the right decision, or are evenly split. As a result, a lot of people have got rather excited about today’s figure, when they really shouldn’t.

As regular readers will know, all polls have a margin of error. I try not to fixate upon the specifics because the margin you always seen quoted (plus or minus 3% for a 1000 sample) is based on a pure random sample with no accounting for weighting or design effects. However, it is a good rough guide – polls are not precise, there is some degree of random variation from poll to poll.

So far this year YouGov have asked the right or wrong to Leave question fifteen times. On average the result has been Right 45%, Wrong 43%, a two point lead for “right”. As with all polls, it varies from week to week, so sometimes it has spat out a lead of four points, sometimes it has been neck-and-neck, and how it’s produced one finding with wrong ahead.

Looking at the figure over time I can’t really be confident in any trend. The gap is smaller than in January, but it’s not as if there’s a steady decline there, it looks more like noise:

  • Jan the average was Right 46%, Wrong 42%
  • February the average was Right 45%, Wrong 44%
  • March the average was Right 44%, Wrong 43%
  • April the average was Right 45%, Wrong 43%

My expectation is, as I’ve said before, the people will probably more towards “Bregret” to some degree, simply because Brexit will require some compromises and some people’s high hopes will be disappointed. However, there’s scant sign of it yet and people’s opinions are often much harder to shift than you’d think.

As ever, YouGov will ask the same question next week, and the week after than and so on. If that too shows people think it’s wrong to leave (and other polls start showing the same thing too) then we can start taking about a cross-over in opinion. As things stand, I really wouldn’t get too excited/worried yet.


The latest YouGov/Times poll has topline figures of CON 45%(-3), LAB 29%(+4), LDEM 10%(-2), UKIP 7%(+2). Fieldwork was Tuesday to Wednesday. While sixteen points is still a very solid Conservative lead it’s down from the towering twenty-points-plus leads that most polls have shown since the election was called. As ever, it’s wrong to read too much into a single poll and this may turn out to be just random sample variation, but it may be that that the immediate boost the Conservatives got from the election announcement has started to deflate. We shall see.

While I’m here, tomorrow’s Mirror is quoting some “Labour insider” saying that Labour’s private polling is showing them at 20%. I’ve written about private polling before – people tend to get very excited about it when they shouldn’t. Just because a poll is “private” doesn’t make it any more accurate or reliable. In fact, given we can’t see it and check whether it is actually true, or something that’s got wildly exaggerated through rumour and half-truths, the information is far less reliable. My advice is always to ignore rumours of “private polls” unless the person responsible actually coughs up some tables for us to look at.

YouGov have put out their first Welsh poll of the campaign, conducted for ITV Wales and Cardiff University. Topline figures, with changes from the previous YouGov Wales poll in January, are CON 40%(+12), LAB 30%(-3), LDEM 8%(-1), Plaid 13%(nc), UKIP 6%(-7). Fieldwork was Wednesday to Friday last week.

These are, it’s fair to say, fairly startling figures. A twelve point increase for a party over a relatively short length of time is extremely unusual, but the direction of travel is the same as Britain as a whole. GB polls had the Tories around forty percent at the start of the year, and have them pushing towards fifty percent now. As in Britain as a whole, the reason seems to be largely the UKIP vote collapsing decisely towards the Tories.

The result is remarkable though because of Wales’ history – it is a Labour heartland, even more so than Scotland was before the SNP landslide. Wales has been consistently won by Labour since the 1930s. The only time the Tories have won Wales in modern political times is the 2009 European elections.

If these shares are repeated at a general election then on a uniform swing the Conservatives would gain 10 seats (taking them to 21, an overall majority of the seats in Wales), Labour would lose 10, there would be no change for the Lib Dems or Plaid. The Tory gains would be much of North East Wales, including Wrexham, both the Newport seats and two Cardiff seats, pushing Labour back to little more than the South Wales valleys.

Roger Scully’s write up is here.

There was also a new ICM poll for the Guardian out earlier today, with fieldwork conducted between Friday-Monday. Topline figures are CON 48%, LAB 27%, LDEM 10%, UKIP 7%, GRN 3% – full tabs are here.

ITV’s Peston show had some fresh ICM polling this morning. Topline voting intention figures were CON 48%(+2), LAB 26%(+1), LDEM 10%(-1), UKIP 8%(nc), GRN 3%(-1), so don’t show any significant change since their snap poll on the day of the election announcement, fieldwork was Wednesday to Thursday.

Thre was also a poll in the Sunday Express. Voting intentions were reported as CON 42%, LAB 26%, LDEM 10%, UKIP 8%, GRN 6%. The poll was conducted by Norstat – a company that normally polls in Norway and Denmark, but who I don’t think have previously released British voting intention figures. There are not yet any details of methodology, how it was sampled or weighted, how turnout was dealt with and so on.

Finally, the YouGov/Sunday Times poll from last night is now up on their site here.