Anything Dan Sarewitz writes is worth reading. Here (PDF) is a short essay he prepared for a recent NSF workshop on the “Science of Science Policy” in which he discusses what such a research agenda might look like. Here is an excerpt:

1. We need a conceptual framework, perhaps analogous to “national innovation systems,” that can help put some boundaries around, and illuminate structure and dynamics within, the complex institutional setting for knowledge creation and use aimed at goals other than wealth creation.

2. Given that public investments in research are usually justified in terms of particular, desirable outcomes, we need to develop generalized approaches for systems-based institutional analysis and mapping that would allow such justifications to be contextualized and tested simply for plausibility. (For example, we might want to test the idea that more fundamental knowledge about the climate system is important for catalyzing a global technological shift to a decarbonized energy system.)

3. As Toulmin realized 40 years ago, science policy discourse often focuses on trade-offs between various scientific fields, rather than between science and other approaches to a particular social need or goal. Because we don’t understand research institutions ecologically, we still lack a decent analytical basis for understanding the role of research within a broad portfolio of potential policy interventions aimed at some goal. Mapping institutional ecologies of knowledge creation, use, and value could provide a foundation for developing new decision tools that allow policy makers to confront fundamental questions, such as: When is “more research” the right prescription? When is it unlikely to make a difference? What other factors are necessary for it to make a difference? When would a different intervention offer a more efficient or plausible route to a desired outcome? (For example, California voters might have benefited from a discussion of the variety of ways that the $5 billion allocated for CIRM might be applied to improving public health in the state.)

4. We do know that “institutions that do research” are embedded in different ways in broader institutional ecologies. In particular, certain types of research settings—e.g., agriculture; private sector software development; clinical medicine—have been identified as sites where feedback between knowledge creation and use is supposed to be strong. We need many, many more institutional case studies to help map out the variety of designs that are available, and to develop comparative frameworks and metrics based on the relations among institutional design, and knowledge creation, use, and value.

5. And of course we also need to reflect on where our own efforts fit in. “Bring back OTA” is not an adequate prescription. What are the loci of decision making where better understanding of, and discussion about, the institutional ecology of knowledge creation, use and value might make a difference? What types of insights, tools and products might decision makers find useful? This workshop strikes me as a huge opportunity to begin to enhance the public value of science policy and science studies research, but we need to start, needless to say, by attending to the institutional ecologies within which we now operate.