I noted in a previous post that there's been a lof of recent interest in developing a theory of technological change. The science policy community is one of the groups leading the charge, with research falling under the moniker "The Science of Science Policy". John Marburger catalyzed this research effort in 2005, during his tenure as director of the Office of Science and Technology Policy, by calling for better methods to decide how to allocate federal R&D funding, and determine what impact it has on the economy and the development of new technology. This led to a 2008 research roadmap that encouraged research toward "models to understand the production of science, qualitative and quantitative methods to estimate the impact of science, and processes for choosing from alternative science [funding] portfolios".
The Federal Government spends about $140 billion annually funding research and development. The importance of technology to economic growth and the necessity of government funding for science was established by economic research in the 1950's and 1960's. Solow found that a sizable portion of the growth in economic output per capita could not be explained by growth in inputs (e.g. labor) and accumulation of capital. The extra growth, termed the "residual", was attributed to improvements in technology. Work by Nelson and Arrow established the need for federal funding because of market failures that cause private firms to underinvest in basic science research. Underinvestment occurs because basic science is risky, and because it is difficult for a single firm to appropriate the resulting knowledge for its own profit.
However, while there is a strong rationale for federal science funding, a theory for how to best allocate the funding is not well developed. It is not clear how to decide which research programs to fund and how to value investments in science relative to other investments such as education or transportation infrastructure. John Marburger articulated these questions in an editorial (paywall) in Science Magazine:
"How much should a nation spend on science? What kind of science? How much from private versus public sectors? Does demand for funding by potential science performers imply a shortage of funding or a surfeit of performers? "
In addition to questions of funding allocation there is a need to understand how government policy affects research and development through intellectual property law, tax policy, and by setting standards to foster openness and competition among researchers.
Currently, the prevailing method for making policy decisions is the use of "expert judgment", in which panels of scientists who are expert in their field gather to peer review grant applications and serve on advisory committees. Once the panels are assembled, the decisions are made without guidance from higher level theory or empirical data regarding what makes for effective science policy. Each scientist decides based on judgement derived from his own experience working in the field. However, there is an emerging view that the system of scientific research has become too complex to manage with expert judgment alone (reference, p. 4).
The 2008 research roadmap lays out three themes and ten research questions for understanding the dynamics of the science and technology system. I've collected and summarized some of the salient points below:
1. What are the behavioral foundations of innovation?
Research and development ultimately derives from the behavior of individuals and groups of people. Social scientists have long examined the economic, social, and cognitive principles that underly human behavior, but the application to the particular context of scientific research is less well developed. The roadmap encourages research into cognitive processes at the individual and group level that support the discovery process, as well as the impact of different incentive and organizational structures on the innovation processes. It also encourages the use of social network analysis techniques to study the transmission of ideas through scientific communities, and the formation and evolution of new scientific communities.
2. What is the taxonomy of the research and development system?
There is a complicated process that leads from new scientific knowledge to new inventions that are commercialized and diffuse across society. For example, the concept of the "idea innovation network" divides research and development into six different stages from basic research to commercialization research, with feedback loops between the stages. Furthermore there are a plethora of technology adoption models to describe how technology is adopted, who adopts it, and how it spreads through society. The roadmap encourages efforts to synthesize knowledge about the taxonomy of the research and development process into a full systems approach to mapping science, technology, and innovation. It also encourages the development of new ways to measure and describe technology adoption and diffusion.
3. How can the value and impact of research be determined and maximized?
Federal agencies have developed a variety of ways to track the effectiveness of their investments including case studies, cost-benefit analyses, peer review, bibliomeric analysis, surveys, science mapping analysis, options modeling, and growth accounting. However, there is no standardization and little communication across different funding agencies. The roadmap encourages the coordination and evaluation of different techniques and the establishment of a set of best practices for measuring the impact of federal research funding.
This research roadmap was written in 2008, and efforts have since led to a handbook of essayspublished in 2011 and more than 130 grant awards in the field. I will delve into some of this research in future posts.