The evidence divide

Earlier this month I attended a conference organised by the STEPS centre at the University of Sussex. The topic of the conference was the use of evidence, especially scientific evidence, to inform the development and implementation of policy. There was a great line-up of speakers, and the organisers should be congratulated for making enough space in the programme for discussion and debate.

A huge array of issues were explored at the conference, but one feature in particular stood out for me: the presence of a real divide between two markedly different ways of thinking about the relationship between scientific evidence and policy-making.

One the one hand there is the social science view. Based on empirical evidence and thoughtful consideration of current and past events, this view presents a nuanced consideration of how science should interact with policy. It values scientific evidence, but also emphasises the importance of other forms of knowledge and evidence, like the knowledge of practitioners working on the ground, who have real insight to bring to tackling complex problems. This view values considering a diversity of possible policy solutions, and considers evidence to be very much about presenting options within which rightly political decisions can be made. It values dialogue as a means to ensure that a wide range of views are included in the framing of problems and the evidence required to solve them, and cautions against the deficit model.

This view sits in contrast to the alternative view espoused by some members of the scientific community. While this community was not well represented at the conference, there was a strong sense that there is a very different world view in play. In this world, scientific evidence is regarded as having a special primacy, even to the extent that there is a call for more ‘scientific’ approaches, like randomized control trials, to be used right across the policy domain. For this community the problem is very much about a ‘deficit’ either in the public or policy-makers or both. The solution is either to fill the gap with more or better information, or to allow scientific evidence to dictate decision-making in the policy sphere.

There is ample scope for speculation about the reasons for these different world views. But there is also a central practical question raised: how do we deal with the differences? If the social science view is accepted, we are making errors in the policy-making process, not because of some inadequacy of the evidence (although there are surely areas where the evidence is inadequate), but because of the approach that is being taken, and the narrowness of the evidence base that is being used.

For me the answer to this dilemma lies with a third group of people, the advisers who sit within policy-making organisations, translators who bridge between the science and policy worlds. These people are of central importance because they can ensure that the plurality of evidence sources are included, that policy problems are framed in a broad way, through the appropriate use of public dialogue, and that research tackles the questions that policy-making needs, rather than those that the scientific community wants to address.

As the pressure builds towards smaller and smaller Government there is a risk that these crucial posts will become increasingly rare. Superficially, the intermediary role can seem wasteful and unnecessary. But removing translators means that scientific advice could excessively dominate, and within that advice the framing of questions might be derived from too limited a set of perspectives. We need to protect the translators, if we want policy making to be the best it can be.