Monday 13 June was my evidence day. First, a slot to speak about academia and policy making at the Investigating Academic Impact Conference, then chairing a session back at Institute for Government with Financial Times Undercover Economist, Tim Harford, on his ‘Adapt’ thesis; that success can only start with past failure, with Director of the LSE Centre for Economic Governance, Prof John Van Reenan, acting as respondent.
So what to make of the two sessions?
First, there is no shortage of existing evidence out there to help policy makers understand the problem. Yet it often exists in a form that makes it difficult for policy makers to use. Those nearest the policy decision – generalist civil servants and Ministers – may not have the skills to make sense of it. For those who do, there is a big communication gap that can impede them from synthesising and using this information in a fair way. For those who listened to a recent Today show former Chief Schools Inspector, Sir Chris Woodhead, gave what is a quite familiar political reaction – that evidence reflects the bias of the researcher who asks questions they are interested in and report results that confirm their views.
Second, good policy can often come from sharing and testing the evidence base before decisions are made. The Institute for Government’s case study of the way in which the Turner Commission on Pensions (of which LSE Professor of Social Policy, John Hills acted as commissioner) tested its evidence base before developing options shows a model of effective evidence based policy making.
Third, even when policy decisions are based on the best reading of current evidence there is no substitute for trial and error to see how they play out in practice. The market acts as a fantastic selection mechanism as do ecosystems but too often the impact of politics is to deny policy makers access to this tried and tested model of progress. Testing things out looks weak and as if the policy-makers lack the courage of their convictions (and the Brits don’t like that in their governments). In addition, given the short shelf lives of both Ministers and civil servants, politicians risk losing electoral credit for being the person who transformed any one of a number of public services.
So the temptation is to rush to full-scale implementation whereas the secret of long-term success can be to build up in scope. The national minimum wage is a great case in point. Academic evidence showed that fears a minimum wage would destroy jobs were unfounded (at least for adult workers) and paved the way for the creation of the Low Pay Commission. But setting the LPC up as a standing commission, the Government was enabled to set the first minimum wage low and then as it was clear the move was not having adverse labour market effects, the Government were able to ratchet it up while solving some of the problematic special cases as they emerged rather than needing to get everything right first time.
As our report, Policy Making in the Real World showed, even when governments do the right thing and commit to evaluate a policy too often the process is compromised. The evaluation is not scoped properly at the start as the consumer of the evaluation is the department (and maybe even the official) who owns the polic and the evaluator wants another commission from the department. Trying and failing carries too heavy a political price. All these biases mean there are too many evaluations which are neither used nor useful. Financial Times Undercover Economist, Tim Harford, contrasted this with our attitude to medical practice where we accept trialling and testing as part of the routine. The Cochrane Collaboration exists to pull together all the available evidence – and then push medical practice in a better direction. But what is routine in medicine is the exception in social policy in the UK.
At our evening session, Prof John Van Reenen suggested extending the National Institute of Clinical Excellence (NICE) model to make evaluation more independent. We have similar proposals in Making Policy Better, where a new Head of Policy Effectiveness, a very senior person in the Cabinet Office, would be charged with overseeing government evaluation and making sure they were rigorous and independent and that the results were used. At a recent joint IFG-NESTA seminar, proposals were made for an independent accreditation body to test what works and guide commissioners toward effective interventions.
Academia needs to join the debate not just on what it can contribute directly to individual policies, but on how to build a system which allows policy to be made better. That may mean creating new institutions. It may also mean developing new ways of producing feedback on policies as they are implemented to give policy makers the sort of information they need on shorter timescales. And academia needs to become a big part of improving the standard of public discourse about policy by helping pubic understanding of the value of good failures.
But does that write politicians out of the script? At our event, a final questioner asked Tim Harford if the approach he was setting out would make politicians redundant. His reply was that politicians would still have to make the big political trade-offs and choices but there was no case for making them (or letting them) make those choices on the basis of poor evidence.
Listen to Jill Rutter talk about academics and policy-making with Maria O’Beirne of the Department for Communities and Local Government and James Johns of HP here.
Jill Rutter is Programme Director at the Institute for Government in the UK. This post was originally published on the LSE Impact blog.
Vicky: very interesting comment..and I think you certainly have a good point. not sure on the Big Society — you might distinguish between the big political concept which is hard to evaluate per se vs the opportunity it offers of having multiple different providers/ approaches which does have huge more possibilities for looking at what works (if we ask the question at the start and build it in). Payments by results is in a sense a way of trying to discover what approaches work. one of the big issues in the way we do policy is that we always go… Read more »
I’ve just come back from the Social Policy Association conference which focused on the Big Society. It struck me that one of the key problems in evaluating a policy platform like the Big Society is its size (not to mention its nebulousness…). I guess that politicians treat their big ideas differently to their small ideas: small ideas may one day be evaluated the way you suggest, but big ideas are a separate species. I don’t really know enough about the history… How was National Insurance evaluated? The NHS? Or were these big ideas just implemented? I suspect the latter, and… Read more »