15/07/2013 11:32 BST | Updated 14/09/2013 06:12 BST

On the Benefits Cap - Evidence, Belief and 'What Works'

Iain Duncan Smith - who happens to be Secretary of State for Work and Pensions as well as my MP - thinks the benefit cap is working. More than that, he has "a belief" that it is working, according to an interview he gave on Radio 4's Today programme. That's OK then, we can all rest easy knowing that IDS' belief trumps the Office of National Statistics' view that IDS was being less-than-accurate in his use of evidence - evidence, lest we forget, produced by his own department.

Perhaps this is a dispute over IDS' definition of "works," in that the ONS criticised him for claiming that the £26k cap had resulted in 8,000 people moving into work (in the absence of any causal link), whereas the Secretary of State might be thinking it "works" rather well as a populist electioneering straregy, what with nearly 80% people supporting it in principle.

Or perhaps it's another (depressingly common) example of inconvenient statistics being sidelined by a Minister, in favour of 'belief' in their preferred policy. Whether or not a policy works shouldn't be an article of faith, it should be a matter of setting up an hypothesis (capping benefits will move people from shirking to striving), testing it in a pilot (as happened here, in London), then learning the lessons. IDS appears to think that if the evidence on the policy is sketchy, or incomplete, or simply contradicts his original hypothesis, it is the stats/evidence that must be wrong, for his belief tells him his policy is right.

This isn't the first time DWP expands a welfare-to-work programme in the absence of supporting evidence. The government's flagship Work Programme failed, miserably so, in its early days - so they did the logical thing, and expanded it. In fact it isn't the first time this month a Secretary of State has sidelined evidence in to preserve political expedience - witness Theresa May's classification of Khat as a Class C substance.

This is to say nothing of Labour's line of attack on the cap, which as George Eaton rightly points out is absurd - to claim that a cap that isn't working is too soft, whilst supporting a broader cap for London, is a strange position to say the least. I tend to agree with Chris Goulden, who says tackling the high cost of housing and other drivers of the cost of living is a better way to bring the benefits bill down - but this is a testable hypothsis I (and no doubt Chris) would be willing to subject to scrutiny in anything ranging from well-designed pilot schemes to (ideally) randomised controlled trials. Sadly IDS and others are not so willing to engage with evidence, it seems.

At the beginning of September, I will start a new job, working on evidence-based policy - and I'll have a particular focus on social policy areas such as crime, justice, education and employment. We cannot continue to tolerate government's failure to respect evidence, nor can we look away whilst belief trumps mere statistics - IDS and his faith in his policies, despite evidence to the contrary, will no doubt keep me busy!