Panic is mounting in progressive circles in Cornwall. First, a BBC report ‘mapping Ukip’s polling strength’ seems to imply Cornwall is a hotspot of Ukip support. Now we have another poll putting Ukip level with the Tories in Camborne-Redruth. Mind you, those polled also wanted a Cornish Assembly by a large majority, which isn’t exactly Ukip policy. So a little bit of confusion there in Camborne-Redruth.
But you can stop digging your survivalist bunker or stockpiling saffron buns. Keep calm and let’s get this in perspective. The BBC report was based on constituency polling by Lord Ashcroft. As Ashcroft’s polling outfit has been polling marginal constituencies, five of the six Cornish seats have come under his radar, a far higher proportion than in England.
Both Ashcroft and today’s Survation poll ask two questions about voting intention. The first is what the media focus on, their attention waning before they have time to get around to the second. The actual question is ‘If there was a general election tomorrow, which party would you vote for?’ The second question however asks people to think ‘specifically’ about their constituency ‘at the next general election’ and ‘the candidates who are likely to stand’.
Now which of these looks more likely to give the better picture of people’s likely vote at next May’s election?
I’d suggest it’s the second. The first is asking about a theoretical election tomorrow which is not going to happen. People are more likely to respond by reflecting the latest half-baked and superficial press and TV coverage. The second question however forces people to begin to think about the complicated task of putting a cross on a piece of paper next May.
Now none of this would matter very much if the two questions gave the same results. But they don’t. Taking the average of the five Cornish polls mentioned by the BBC and taken from June to September, we get the following distribution. The Tories are in the lead but Ukip looks to be in a healthy second place.
These polls usually have a polling error of around 5% either way once the don’t knows and the don’t cares are stripped out. So, taking that into account, the answers to Question 1 suggest the Tories had a clear lead in North Cornwall, Truro and St Ives, while St Austell and Camborne-Redruth were Tory/Ukip marginals.
But let’s look at the second question.
The broad picture is similar but the Lib Dem share has miraculously increased, especially in those constituencies with sitting Lib Dem MPs. For example, in St Ives the percentage plumping for the Lib Dems in Question 1 almost doubled from 18% to 32% when people were asked to consider voting specifically for their MP Andrew George.
As a result, if we focus on Question 2 only Truro remains a Tory safe bet. St Ives and North Cornwall are transformed into Tory/ Lib Dem marginals. Meanwhile, St Austell becomes a three-horse Con/Lib Dem/Ukip marginal, while Camborne-Redruth was a three way Con/Ukip/Labour marginal.
The two questions in today’s poll display exactly the same thing. The Lib Dem share almost doubles from Question 1 to Question 2. The only problem for the Lib Dem’s Julia Goldsworthy is that this still leaves her languishing in a very poor fourth place. But it gives George Eustice a comfortable 34-28 point lead over Ukip, with Labour’s Michael Foster slipping out of the race to 18%.
That didn’t stop the local (Tory) press reporting today’s poll findings as ‘Ukip leading the way‘ and Ukip ‘topping the poll‘. The Western Morning News even went so far as to predict ‘Ukip on course for first Westcountry [sic] MP‘.
This is of course rubbish. There are other two reasons to be more sanguine. First, when people actually vote, there isn’t much sign in Cornwall of a huge Ukip surge. At the recent Mevagissey by-election Ukip’s share of the vote actually fell. At last year’s round of unitary elections their performance in Cornwall, gauged by the number of seats won, was nothing out of the ordinary when compared with English counties.
Finally, look at the bigger picture. The Ukip bubble will burst; it’s a question of when, not if.