Ukip gained a lot of publicity from its journalist friends on the back of a recent poll conducted in the Camborne-Redruth constituency. The poll was conducted by Survation for Exeter University at Tremough. However, there are growing suspicions among political anoraks across Cornwall that this poll may not be fit for purpose.
Specifically, alarm bells are sounding over the question about the way people voted back in 2010. This question depends on memory. So it’s often a bit wrong. But in this poll it wasn’t just a bit wrong; it was massively inaccurate. Look at how the respondents in the Survation poll claimed they voted and how the vote in Camborne-Redruth actually turned out in 2010. (This ignores vote for others). For comparison, the figures are given for the same question from the Ashcroft poll of the same constituency in June.
The 2010 vote in Camborne-Redruth
|Survation poll Vote 2010||Ashcroft poll Vote 2010||Actual vote 2010|
Perhaps there’s an obvious explanation for this. Maybe Survation polled a lot of people who were just too ashamed to admit they voted Lib Dem. For which I suppose we can’t blame them. Or could their poll have been biased towards 2010 Tory and Labour voters and against Lib Dem voters?
On the other hand, if the handful of Lib Dem activists left in Camborne-Redruth want to take heart from this, they’d be disappointed. Both Ashcroft and Survation have the Lib Dems’ Julia Goldsworthy trailing on a similar score of 13-14%. And both put Ukip at around 26-28%.
(Thanks to Rob Simmons for alerting me to this)