What caused the riots? For perhaps the first time this year, the Sun and the Guardian were in agreement - or at least their pollsters were. The Sun's YouGov poll asked for the 'main cause of the riots' and top of the list was 'criminal behaviour' (42%). The Guardian's ICM poll asked the same thing and there it was again 'criminality on the part of the rioters' (45%). Two polls. Two newspapers. One finding. The public must have decided.
But there is something very odd about the idea that criminality caused the riots: rioting is criminal. In fact it's hard to think of more archetypal criminality than kicking in the window of Curry's and making off with a clutch of MP3 players and some nice headphones. Saying that criminality caused this criminal behaviour is a bit like saying that I'm hungry because I'm peckish. I think there is a way of parsing that logically (I'll leave it to you to fathom) but it is not exactly clear. Could it be that the two polls - despite agreeing with each other almost to the decimal point - were in fact measuring confusion about what the question meant rather than real attitudes? I decided to find out.
There is a sensible way in which criminality can cause riots. It could be that the people who were rioting were criminals on the look out for an opportunity. When the Old Bill aren't around, they have a propensity to nip into Debenhams and help themselves to track-suit bottoms and trainers. I think that is the sort of thing someone could possibly mean when they say that 'criminality' caused the riots. So I put it into our poll.
But the poll needed more than one possible cause, so we added some others that have been floated in the media. We added an explanation around greed, one around irresponsibility, one around opportunity and one around cuts. You can see the full question wording below. And because we know we don't have a full list of possibilities (Where is racial tension? Where is recessive policing? Where is the shooting of Mark Dugan) we added an 'other' option. Then fielded the poll over two days and looked at the results.
Suddenly criminality isn't top. It isn't second. It's third - just behind greed and way behind irresponsibility. And if you let people pick two causes rather than just one, the gap grows. More than 50% pick the greed option. More than 50% pick the irresponsibility option. And just 37% pick the criminality option.
You could argue for a long time about why criminality is less salient here than in the Guardian and the Sun polls. My view is that it is to do with their criminality answer option which (unlike ours) had the dual character of explanation and expletive. And if you're measuring the emotive power of words that's fine. It's just different to actually trying to find out what people blame for the riots.
What does this show? Two things - one substantive and one about how to look at polls. The substantive point is that the public tend to agree with city analyst Dr Tim Morgan when he argued that the riots were a result of "out-of-control consumerist ethos". The national conversation about the riots has to connect to this and look at how we build a greater sense of responsibility and also ensure that the status that comes with the latest fashion isn't the only sort of status people can plausibly aspire to.
The point about polls is this: look very carefully before you trust what they appear to say. Public opinion isn't a single crystalline structure. Every individual, let alone every public, can simultaneously hold mutually contradictory views without much difficulty. As a result, two pollsters with two methodologies can find two totally different facts about public opinion. Both can be true in the sense that they can be replicated. The art of the pollster isn't just getting to the truth, its getting to the right truth.