A few weeks ago, I attended a British Academy workshop on ‘Urban Futures’ – partly focused on research priorities and partly on research that would be useful for policy makers. The group consisted mainly of academics who were keen to discuss the most difficult research challenges. I found myself sitting next to Richard Sennett – a pleasure and a privilege in itself, someone I’d read and knew by repute but whom I had never met. When the discussion turned to research contributions to policy, Richard made a remark which resonated strongly with me and made the day very much worthwhile. He said: “If you want to have an impact on policy, you have to lower the bar!” We discussed this briefly at the end of the meeting, and I hope he won’t mind if I try to unpick it a little. It doesn’t tell the whole story of the challenge of engaging the academic community in policy, but it does offer some insights.
The most advanced research is likely to be incomplete and to have many associated uncertainties when translated into practice. This can offer insights, but the uncertainties are often uncomfortable for policy makers. If we lower the bar to something like ‘best practice’ – see preceding blog 42 – this may involve writing and presentations which do not offer the highest levels of esteem in the academic community. What is on offer to policy makers has to be intelligible, convincing and useful. Being convincing means that what we are describing should evidence-based. And, of course, when these criteria are met, there should be another kind of esteem associated with the ‘research for policy’ agenda. I guess this is what ‘impact’ is supposed to be about (though I think that is half of the story, since impact that transforms a discipline may be more important in the long run).
‘Research for policy’ is, of course, ‘applied research’ which also brings up the esteem argument: if ‘applied’, then less ‘esteemful’ if I can make up a word. In my own experience, engagement with real challenges – whether commercial or public – adds seriously to basic research in two ways: first, it throws up new problems; and secondly, it provides access to data – for testing and further model development – that simply wouldn’t be available otherwise. Some of the new problems may be more challenging and in a scientific sense more important, than the old ones.
So, back to the old problem: what can we do to enhance academic participation in policy development? First a warning: recall the policy-design-analysis argument much used in these blogs. Policy is about what we are trying to achieve, design is about inventing solutions; and analysis is about exploring the consequences of, and evaluating, alternative policies, solutions and plans – the point being that analysis alone, the stuff of academic life, will not of itself solve problems. Engagement, therefore, ideally means engagement across all three areas, not just analysis.
How can we then make ourselves more effective by lowering the bar? First, ensure that our ‘best practice’ (see blog 42) is intelligible, convincing and useful; evidence-based. This means being confident about what we know and can offer. But then we also ought to be open about what we don’t know. In some cases we may be able to say that we can tackle, perhaps reasonably quickly, some of the important ‘not known’ questions through research; and that may need resource. Let me illustrate this with retail modelling. We can be pretty confident about estimating revenues (or people) attracted to facilities when something changes – a new store, a new hospital or whatever. And then there is a category, in this case, of what we ‘half know’. We have an understanding of retail structural dynamics to a point where we can estimate the minimum size that a new development has to be for it to succeed. But we can’t yet do this with confidence. So a talk on retail dynamics to commercial directors may be ‘above the bar’.
I suppose another way of putting this argument is that for policy engagement purposes, we should know where we should set the height of the bar: confidence below, uncertainty (possibly with some insights), above. There is a whole set of essays to be written on this for different possible application areas.