Iain Dale, and others who oppose 42 day detention are greeting with great joy a new ICM poll for the Joseph Rowntree Reform Trust that suggests 60% of people think the limit on the length of time terrorist suspects can be held should be less than 42 days. I’m afraid I have to rain on their parade.

Firstly, some bloggers are interpreting this poll as showing that that David Davis’s campaign has swung public opinion against 42 day detention. I’m afraid it doesn’t – what it does show is that if you ask the question in a drastically different way, you often get a very different answer. If you want to discern a change in public opinion, you need to compare like with like.

On PoliticsHome they have done just that, asking their panel indentically phrased questions on the 20th June (2 days after David Davis resigned as an MP) and this Monday (7th July). Back in June at the start of Davis’ campaign the figures were 65% supporting extending the period terrorist suspects can be held without charge from 28 days to 42, with 31% opposed. This Monday the figures were 66% support, 30% oppose. There is quite obviously no significant change at all in opinion.

So, if there is no shift in opinion why does the ICM/JRRT poll show such a different answer? Because of the way the question was asked. It prompted people to begin with by reminding them of the long standing traditions of British justice: “Britain has long-standing rules and principles that have been put in place to protect people from being arrested and wrongly held for an indefinite time in custody.” This is perfectly true, but isn’t necessary for people to answer the question and risks skewing answers.

It then asked about offences in general, then murder suspects, then terrorist suspects – this firmly grounds the treatment of terrorist suspects as part of the wider legal system in the UK, when in political and media discourse it is often talked about as if it is a separate thing. It also made the fact that the people being held may be innocent far more explicit than most questions.

So what do people actually think? When designing surveys with clients something I often need to point out is that a poll is designed to measure public opinion as it is, not as we would like it to be. If the public are ignorant of the arguments about something, then generally speaking a poll should not try to educate them about it, because it only serves to make them more informed than the wider public they are supposed to be representative of!

The JRRT poll suggests that when people are primed to consider 42 day detention for terrorist suspects within the context of Britain’s legal traditions, comparison with treatment of other suspects and that those people might be innocent then support is lower. In reality do people really consider those arguments when forming their opinions about 42 detention? Probably not, or unprompted polls wouldn’t show such different answers. If people thought more about these particular angles upon the issue then no doubt they would have different opinions (and that is a positive sign for those campaigning against it – opinion on this can be moved), but if you give people a straight yes or no on 42 days then around about two-thirds of them consistently say they support it, and there is nothing to suggest that has changed yet.

I have a long post up on PoliticsHome looking at how Labour got into the position of being 20 points behind, based on the daily data produced by their Phi5000 panel.

It’s very clear that the point when their reputation really crumbled was during the 10p tax row, when approval of the government collapsed, they were seen as increasingly divided and the last few shreds of Gordon Brown’s reputation for competence fell away. In contrast, the Conservative increase then is still mostly just a result of them looking good when compared to Labour – their negatives have reduced, but there is no obvious positive boost for them. The exception is Cameron himself, whose reputation has improved significantly since the local elections and mayoral elections, particularly in terms of competence and effeciency.

Still, that is how we got here. If you look at the graphs over on the post you can also see a couple of wobbles in the Conservative ratings over the last week or so, look particularly at the way the percentage of panellists who think the Conservatives are united has dropped, and those who aren’t clear what they stand for has increased since David Davis’s resignation.