Dr. Marburger’s address last week spent a fair amount of time discussing his notion of a ’science of science policy’ and how that could help government make more efficient policy decisions for science and engineering investments.  Reflecting his scientist credentials much more strongly than his administrator credentials, Marburger has spent most of his time on this idea focused on the experimental/research part of this challenge.  The most tangible evidence of this concept is the National Science Foundation’s research program of the same name.  There is also an interagency working group and listserv associated with the program.  The idea for all of these elements is to increase and improve the state of information on innovation in general, with an emphasis on economic factors.

On their own, these elements will do nothing to improve or affect policymaking in these areas.  There are several reasons why.  First among them is the complete absence of any mechanisms or processes for transferring this knowledge from the bench to policymakers.  This is no shame on NSF because they aren’t focused on knowledge transfer in any of their research programs.  However, this particular program, at least as Dr. Marburger seems to envision it, would connect research knowledge to policy action.  His comparisons of science policy communities to tax policy and foreign policy communities suggest a particular failure in this area.  Transfer mechanisms are necessary.  There could be the traditional mechanisms of conferences, personnel exchanges and other interactions between researchers and policymakers, but much like the phenomena investigated in the research program, it’s not clear whether these mechanisms would be the right ones to transfer this knowledge.

A worthy question is whether a research proposal to test such a mechanism could be funded by the SciSP program.  This could be done prospectively or retrospectively.  Examples of the latter would be analyzing the failure of the American Competitiveness Initiative to provide steady funding increases, or testing whether the story of Sputnik’s influence on the scientific workforce really had the influence commonly ascribed to it.  An analysis of awards made so far in the program suggests the answer is no.  Erin Moore at the University of Colorado has a rough assessment of the research program (H/T Robert Frodeman), including some revealing pie charts describing the disciplinary breakdown of awards to date.  At first blush, if your proposal is outside economics or statistics, your chances are small.

This feeds into Moore’s general criticism of the program (and of current federal science policy), which is persuasive, that SciSP is focused on outputs rather than outcomes.  From her perspective, there’s no real examination of the influence of innovation policies on society.  Perhaps that’s from a narrow framing of the overarching question – how can innovation policies be more efficient?  The focus is on the desired goal of the policy, and not about the consequences of those policies.  A related narrowing involves a strictly economic orientation on innovation, with a presumption that advances in science and technology are valued primarily, if not only, on their direct economic impact.

The question remains, however, about how knowledge transfer between researchers and policymakers could be improved, or created, should you think that the capacity for transfer does not yet exist.  There is also the related question of how more ‘rational’ (more informed by the research output of SciSP) policies can be implemented.  I will focus on these questions on Monday.