On science, expertise, and policy

The Wellcome Trust has had an interesting pubic debate on expertise recently. This got me thinking about my own experience of science and policy, and what counts as ‘expertise’. None of this is new: Chris Tyler said it all better here.

I was lucky enough to do a masters and doctorate in neuroscience funded by the Wellcome Trust. About a year into my doctorate I realised that I didn’t want to stay in academia afterwards. There were lots of reasons for this: I didn’t particularly enjoy the practical aspects of my research field; I wanted more variety in my work than I thought I would get from an academic career; I didn’t fancy my chances running the gauntlet of fellowships and post-docs that bedevils early academic careers. I’d been interested in the interface between science and the wider world for a while, so I thought policy might be the thing to do. Again, I was lucky and got on the Civil Service Fast Stream.

I spent the next nine years doing policy and project management jobs in a few different government departments. Most of my jobs had nothing to do with my scientific training. This came as a bit of a surprise, but policy work (at least as I experienced it) is quite a general discipline that needs a range of skills. Subject matter knowledge is important, but not sufficient for making good policy. Evidence on what works or is possible, often supplied by the research community, needs to be combined with evidence on what is legal, what is affordable and the best use of money (not always the same thing), and what is acceptable to the public and their elected representatives and other interested groups. This is why some people prefer to talk about ‘evidence-informed’ rather than ‘evidence-based’ policy. Very few people can honestly claim expertise in all aspects of a given policy area.

Here’s one way I think about policy making: In the beginning, there is the space of exciting possibilities, which contains all the possible ways to solve a given problem. Next, you throw out the things that definitely won’t work, the things you can’t afford, the things that are illegal, and the things that will definitely loose the next election. This leaves the zone of miserable compromises. The most effective option is rarely affordable or politically feasible. The most popular option usually isn’t one that would work. Sometimes it is possible to nudge the edges of the zone of miserable compromise. Legislation can make things legal (or illegal), money can be found if the need is great enough. But the end result almost always feels like a compromise. In this context, an ‘expert’ can sometimes come across as a zealot if they emphasise the research base to the exclusion of all the other constraints on policy.

In a democracy, decision making rests with politicians. People vote for politicians and their parties for lots of reasons — values, self-interest, because of their nice hair. ‘Doing exactly what the research evidence says’ isn’t all there is to it. In my role as an official I was always aware that it wasn’t me who was accountable to the public, and who faced losing his job every five years. If a political leader chooses a policy that is at odds with the evidence base, then they can expect to have to defend that decision, including from the experts in academia and elsewhere.

There’s a lot more that could be said on this topic: about how the evidence gets constructed, how the interests of the people who decide on the evidence base influence this, and how what counts as evidence relates to what you value in the first place. But that’s for another day.

I’ll end with a few musings about how government might help to make the interface between researchers and policy wonks a bit better.

  • The RCUK policy internships are a great idea. We should do more, and for researchers at all stages of their careers.
  • We should also have the reverse: fellowships to allow policy wonks to try their hand at research, maybe funding for a research masters.
  • Some one-day courses in how government and policy work and how evidence can influence it for researchers would be good.
  • More safe spaces for policy wonks to talk to researchers under the Chatham House rule would be good.
  • The Civil Service may need to find ways to bring people in mid-career (my experience was that it’s good at bringing people in early in their careers, and directly into senior positions, but less good at bringing people into mid-level jobs).

I’m a fan of the practitioner-researcher model. I think this helps to get evidence into practice, and makes the evidence that is produced is more relevant. This has had a lot to do with my move into public health, which seems to be an area where there are opportunities to mix research and practice. It may be a model that can be applied more widely.