Fascinating figures

UK Polling Report has an analysis of some of the results from a YouGov poll released recently, about voting patterns and voters' beliefs and so on. Anthony Wells focusses on the BNP, and I shall extract some interesting figures there, but here is a starter for ten: voting Labour is hereditary, more so than the Tories.

%	I vote Conservative	I vote Labour	I vote Lib Dem
My parents voted Conservative	47	11	23
My parents voted Labour	25	66	38
My parents voted Liberal/Liberal Democrat	4	3	12

So in terms of priors, given that you parents vote/d Labour, you are six times as likely to vote Labour as Conservative; if you parents voted Labour, you are about twice as likely to support the Tories in elections. The Liberal Democrat figure is something of an oddball, probably because their history does quite not stretch across a generation. This is particularly peculiar because it would suggest that Labour retains votes very well, but presently they are haemorraghing them.

Moving onto the political hue of the country, we have the following two questions: how left/right-wing do you think you are; and how left/right-wing do you think the party you vote for is? Here are the results: negative numbers are left-wing, and positive, right-wing. (YouGov's choice, not mine!) I have added a variance figure [1].

	Conservative	Labour	Liberal Democrat	Green	UKIP	BNP
Personally	26	-29	-17	-29	17	25
Party	32	-24	-15	-34	24	42
Variance	1.81	2.08	0.86	1.17	1.41	3.36

You can see that the variance is reasonable for all the other parties—noticeably lowest for the Lib Dems, which is interesting—but it shoots up for the BNP, with its electorate thinking it far more 'right-wing' than they are personally.

Although the figures go on to suggest that a disproportionate contingent of BNP voters genuinely are racist bigots, as Anthony Wells goes on to explain, this inexplicably huge variance does lend credence to the suggestion that BNP voters are not, in large part, voting for the BNP, but against everyone else. This is only strengthened by the following facts: firstly, that the BNP vote profiles itself as about as different from the BNP as from Labour; and secondly, that the BNP voters scored variances of 1.12 and 1.65 against UKIP and the Tories respectively. In other words, BNP voters would probably vote for UKIP or the Tories if… something were different. [2] What, I do not know.

[1] I derived the variance figure by taking root mean square deviation over the political spectrum found in the full tables. I can supply an Excel spreadsheet if you can prove that you are a fully paid-up nerd.
[2] Readers may wonder whether how I view all this in the context of my view that the BNP is not simplistically right-wing. My answer: the media tends to focus on the racism side of things, and it clearly serves the left-wing end of the media to insinuate that racism is a 'right-wing' attitude. They ignore the economic policies, about which I understand Griffin was expostulating at some length on his election, and showing himself to be as socialist as they come. Clearly, some aspects of the BNP's political positions resonate with the right's own, but equally, some are on the left.