Sunday, 19 February 2012

UKIP on 6% in YouGov poll

YouGov's latest daily poll has UKIP on 6%, just 1% behind the Lib Dems again.

You wouldn't know UKIP was trailing the Lib Dems by just 1% as it has pretty consistently for almost a year by looking at the headline figures because YouGov still insists on putting UKIP in the "other" bucket, despite snapping at the tail of the Lib Dems for months and the closest "other" party having half of UKIP's vote.  Unlike Angus Reid, YouGov has yet to recognise that UKIP is virtually level pegging with the Lib Dems on a national level and could conceivably replace them as the third party in Westminster at the next election if the prediction of them losing all but 9 of their MPs is correct.

Some other statistics in the poll are interesting and make encouraging reading.  Despite the perception that UKIP is a eurosceptic version of the Tories, 5% of people who voted Lib Dem in the last election and 1% who voted Labour intend to vote UKIP in the next election compared to 8% of Tory voters - UKIP attracts votes from left and right in similar numbers, including people who traditionally vote for the rabidly europhile of the LibLabCon.  UKIP also attracts more women voters (7%) than men (6%), putting paid to the myth that it is an old boys and blazers club and attracts more working class voters (7%) than upper and middle class (5%).  Support for UKIP is pretty even across England except for the Midlands where it falls short, although this is probably down to the "Wales effect" (YouGov, quite inexplicably, combines the Midlands and Wales which produces some strange and often irrelevant data).

YouGov's ignorance of devolution and national institutions has skewed some of the statistics.  Indifference towards the British government's £9k a year university tuition fee regime in England has been artificially inflated by including Scottish opinions on a tuition fee regime that doesn't apply to Scottish students and support for grammar schools in England has been depressed by including Scottish opinions on a grammar school system that doesn't affect Scotland.  The number of people having the opinion that the Church of England plays a valuable role in "Britain" has similarly been depressed by including the opinions of Scottish people whose national church is, of course, the Church of Scotland (still protestant but importantly, not English).

Most encouragingly, though, are the opinions on UKIP policies.  A whopping 88% agree with UKIP's policy of increasing the tax threshold to take low earners out of the tax system altogether and 56% support UKIP's policy of rewarding families and married couples through the tax system.  UKIP's policy of supporting and expanding the grammar school system is supported by most voters, as is UKIP's policy that universities should admit students based on academic ability rather than quotas.

Comments (4)

Loading... Logging you in...
  • Logged in as
The Church of Scotland is not Anglican but Presbyterian. And Scotland may have certain policy matters devolved but Scottish voters still vote for the UK Parliament. Plus some of these issues may impact on how Scottish voters vote, so it's perfectly legitimate to ask the questions.

As for Wales it's too small to break out into a separate sub-section that would be meaningful and Welsh polling is an underdeveloped science.
Sorry, should have said protestant not anglican.

Scots aren't going to have the same strength of opinion - or even the same opinion - on unpopular legislation in England such as tuition fees. It doesn't apply to them, including their opinions materially affects the headline outcome and if you look at the tables you'll see how the Scottish results are out of step with the English results.

As for Wales - yes it's small by comparison but it needs separating from the English midlands.
2 replies · active 683 weeks ago
A separation of Wales would confuse more than help - with such small numbers the margin of error would be huge, the swings between polls enormous, and the numbers would give a false impression.

The breakdowns of the poll allow the differences to be seen so asking the Scots makes no difference overall beyond the headlines - but polls are produced to track public opinion not provide convenient headlines. A lot of voting is based more on impressions and feelings rather than specific policies so devolution makes less of a difference than it first seems. But when the results are significantly divergent it will be noted and more likely to impact on policy.
A larger per capita sample size for Wales would make the results more accurate, not less. They sample around 1,000 people per poll as that's the minimum sample size to be representative. The sub-samples are all inaccurate because they don't have large enough sample sizes. YouGov breaks GB up into five "regions". Scotland is the smallest with a population of 5m, London is second with a population of 7.5m. The other 46.5m is divided up between the south of England excluding London, the north of England and the English midlands/Wales. When three of the polling regions are three times the size of one and double the size of another, can you really argue that Wales with a population of 3m people, its own devolved government and nationalist political party can't be treated separately?

Of course polls are used to track public opinion rather than produce convenient headlines but if the polls are skewed because the polling company is asking Scots for opinions on English-only policies or combining the opinions of Welsh people with those of the English midlands then it devalues the survey.

Post a new comment

Comments by