On the need for evidence based research policy - a response.

Thanks to Ben Johnson and Pedro Serôdio  for commenting  (here) on my paper in Research Fortnight (see the post below).

Theirs’s isn’t in fact a response to my paper, but rather an illustrative example of the problems I was trying to get at. They don’t recognise the issue I was raising and seem to think that I am denying that R&D is important. This is not the case.  The  science base, R&D, innovation are all profoundly important at global level, but the relations to the national case are not straightforward, for reasons which should be obvious but are not.  What is true of the world economy, or of say the USA may not be true of the UK.  Not seeing this is a fundamental flaw in science policy, as I know.  Hence my reference to blasphemy.

That my key point is not appreciated is revealed by the fact that the evidence, and the papers cited, about spillovers and the like are not (with one exception) about the UK. The study of spillovers is for R&D done in industry in the United States.   Another uses OECD and French data and shows a very low defence  R&D effect on US growth. Another looks at Korea, another at the US.  One is a survey of theories and points which fail to address the obvious empirics of national R&D and growth I was trying to put on the agenda (yet more grist to my mill).   The only one on the UK shows that UK tax breaks increase R&D, and then makes assumptions about spillovers – there is no analysis of actual UK spillovers.  

I tried to pre-empt what seems to be their main point relating to the UK, which is about absorptive capacity, because I know it is a last-ditch defence of the conventional view that national innovation correlates positively with national growth (or should do). I noted: ‘The argument that R&D is needed to absorb overseas innovations is an untested assertion. It will doubtless be true in some cases, but not in others, probably most, given the importance of multinationals and straightforward transfers. It is not supported by national-level data.’   Ben and Pedro just reassert the point I was objecting to.   Sure, as I noted, there is doubtless something to it, but how important is it?  What is the evidence that this is significant, especially for the academic research base?  Are we supposed to believe we have electric cars because of British battery R&D?  Science policy needs a proper evidence base, not the reassertion of supposed truisms. It is too important for that.

Lastly, Ben and Pedro suggest that I think that because the economy has not grown it is because of a failure of research. Again, obviously not as I am pointing to importance of overseas research.  My point was about the failure of the model they are defending, which does indeed arise from not noting that other things are involved in growth, which they seem to think is my problem, not least overseas research. Ben and Pedro don’t seem to want to see  that what I am calling for is not the end of British R&D, but for a different policy for it, one which might work.   As I say, all grist to my mill.