David Guston has a policy commentary in the latest issue of Nature – 454:7207, pp. 940-941, (21 August 2008) where he discusses the capability of policy to influence innovative research.  It’s worth a read (as most of the Nature articles on innovation are) in it’s own right, but I want to take a couple of his points as the start of a discussion on what should ground policy.  From Guston’s article:

“Policies are, and should rightly be, about articulating public values.” (940, referencing Bozeman’s Public Values and Public Interest)

The second premise [of three that Guston considers to underly contradictions within innovation policy] is that policy is supposed to be grounded on a clear understanding of the natural world.

Guston goes on to note that science provides an incomplete and changing picture of the world.  I agree, and certainly think that grounding policy in a clear understanding of the natural world – in good natural science – is insufficient.  Even adding the work of the social scientists and humanities scholars will not ground policy as well as it might be grounded.

What I think is also missing here is a clear understanding of the policy world.  In other words, it’s not enough to know what the science (or technology) says about specific policy outcomes or particular research portfolios.  The linkage between them and the values articulated in policies is tenuous at best, and often absent.  That I know the scientific or technical consensus behind a particular issue or concern is fine, but that’s not going to tell me what the best policy instruments will be to make sure that the articulated public values are achieved.  This is true even when considering those policies that Harvey Brooks called “policy for science.”  I do not mean to say that knowledge and values are incommensurate, but that different processes and questions are needed to produce knowledge than are needed to make sure values are properly articulated through policy.

Policy research (or science and technology policy research if we want to be specific) must complement other scholarly research in order to best inform policymakers in making sure that their policies properly articulate the values that they want.  While the National Science Foundation’s (NSF) Science of Science Policy (SciSP) program will likely produce new knowledge about innovation, metrics, and research, it remains to be seen if that program can say anything about what policy instruments are most effective in articulating the values that policymakers want.  I’ve been a skeptic of this program for a while, and this ‘missing half’ of SciSP is a big reason why.  Now maybe NSF is not well suited to handle that second half.  But there is a big unanswered question about how the knowledge produced through this program will be used by policymakers, if by anyone at all.