YouGov for The Sun: Coalition Monkey Business

Couple of problems with this new YouGov poll for Murdoch's super soaraway The Sun. Some of the ConDemNative bloggers and tweeters are loving the nation's response to the first question. But to these eyes that looks rather rigged and just a tad unscientific. Knowingly so. Though as a series develops it should become more meaningful. In the meantime watch out for screaming headlines that miss the reality by a country mile. Data and details.


If you want a fair test of the range of people's feelings about X or Y it's usual to give them a scale with two or three positive answers of different strength and two or three negative answers of different strength with a "Nichevo" option in the middle. Being the Russian for [shrugs].

Looks like Q1 in this poll did not have that option. There are some Don't Knows but given the relatively small number I'd guess this option wasn't on the list. So most respondents accept the force and answer positive or negative, And this can be "strongly" felt or the much weaker "tend".

Overall 60% of respondents said they felt positive about the ConDem Coalition two days in. But when we interrogate this we find that three-quarters of these are in the "tend" camp. 45% of the respondents may basically be no more than giving the thing a chance. With a further 19% in the "tend" in the negative direction. 64% with no strong view, 7% refusing to be forced, 71%. Just 29% strongly one way or the other and to all intents and purposes splitting 50:50.

Some readers also seized on the finding that 25% of Labour supporters (c 60/242) were positive. But here the weakness is even greater with just 5 people (2%) strongly positive - possibly happy to escape the full-on Tory alternative, the question does not distinguish - and about 55 (23%) just tending.

A large proportion of the "tenders" in my experience - all Opinion and MR companies will ask such Qs now and then - would take the "Nichevo" or "Mu" option if it were there on the scale. Which is why News International opt for the force. More of a story perhaps.

But the underlying real story is that for every hundred people asked 71 had no strong feelings, 15 strongly approve the coalition and 14 strongly disapprove. Chucking in the margin of confidence which will be +/-3% might as well say the jury's out.


To my eyes the second question with fairly clear time bands and an overt "Don't know" is a fairer and more revealing one. 62% of respondents predict the coalition will last about two years or less, 28% saying less than a year. With 13% responding DK there are just 25% of the sample believing this thing will last three, four, five years or more.


The third and fourth questions ask respondents to gauge the effect of the coalition on two of the three parties. Why they didn't ask about, or publish the third party I don't know. It would have been an interesting check if nothing else. But with rounding it appears that for the two parties involved around two thirds think the coalition parties will benefit while opposition (Labour) supporters go about 2-1 the other way.

All well and good. But as this approval/disapproval is tracked the things to look for are the "strongly" not the "tend", and also the regional differences with disapproval and predictions of shorter spans strongest in the Lib-Lab battlegrounds - the North, Scotland, London.

YouGov poll supremo Peter Kellner provides his own commentary on the election and the coalition HERE and you can join their online panel RIGHT HERE. Click on charts to enlarge. Or follow first link for the original pdf.