Holdren Responds to Proposals to Limit Federal Support for Research

In remarks prepared for the AAAS Forum on Science & Technology Policy, Office of Science and Technology Policy Director John Holdren addressed recent developments on Capitol Hill regarding federal support for research. Holdren was responding to congressional concerns about federal funding for social science research and proposed legislation impacting the grant making process at the National Science Foundation. Selections from his remarks, delivered yesterday in Washington, D.C., follow:

Funding for Social Science Research:

“I want to make a further point about the kinds of research that the Federal government is and should be funding. Members of Congress have recently suggested, variously, either that the social sciences are not really science and should not be supported by the tax-payers at all; or that research in political science, at least, should only be supported if the NSF will certify to Congress, for each grant, that the research will advance either the economy or national security (a provision now actually embodied in law in the most recent Continuing Resolution governing spending for the remainder of FY13); or that all taxpayer-funded research should have to pass the test of offering a predictable benefit for some national interest.

“Let me therefore be clear about the position of this Administration, as President Obama was in his remarks on Monday at the 150th anniversary meeting of the National Academy of Sciences.

“First, the social and behavioral sciences - which of course include economics, sociology, psychology, and anthropology, as well as political science - are sciences. Researchers in these fields develop and test hypotheses; they publish results in peer-reviewed journals; and they archive data so that others can replicate their results.

“Second, while much of the work in these sciences meets the definition of basic research - expanding our understanding of ourselves and our surroundings - much work in the social and behavioral sciences is aimed at having (or ends up having without being aimed that way) practical applications to society’s direct benefit.

“Political science research helps us understand the motives and actions of nations and peoples around the world, strengthening our foreign policy, and it helps understand our own democracy and how to make it stronger. Economics research has clarified not only the economic importance of innovation, but also its determinants, which in turn have helped us craft policies that effectively promote innovation and thus economic growth.

“Social and behavioral research has helped us make hurricane warnings more effective, improve methods of instruction and training in school and in the workplace, and manage common resources efficiently without centralized regulation. And it has taught us that social-distancing strategies, like staying home from work or school, can be a crucial complement to vaccination strategies when it comes to breaking the transmission of influenza from person to person.”

Targeted Research:

“Third, whether we are talking about research in the social and behavioral sciences, or in the natural sciences, it makes no sense at all to confine taxpayer support to those projects for which a likely direct contribution to the national interest can be identified in advance. (Unless, of course, the national interest is defined to include expanding the boundaries of knowledge, which would be fine with me but is not, I think, what members of Congress proposing the criterion have in mind.)

“Imposing such a national-interest criterion in the form its sponsors seem to have in mind would throw out the basic-research baby with the bathwater, inasmuch as basic research constitutes precisely that subset of research activity that is aimed at expanding knowledge without reference to possible applications.

“My friend Professor Charlie Townes got the Nobel Prize in physics for his work in quantum electronics that led to the maser (a term now nearly forgotten, standing for ‘microwave amplification by stimulated emission of radiation’) and the laser, where the ‘m’ for ‘microwave’ is replaced by ‘l’ for ‘light’. Neither Charlie nor anybody else at the time he and his colleagues were doing this pioneering basic research, in the 1950s, could have imagined that some decades later the laser would be the basis of how we photocopy, play music, cut metals, and perform eye surgery, among a host of other practical uses.

“Of course there is a fuzzy middle ground of fundamental research that is use-inspired rather than just curiosity inspired, for which beneficial applications can be imagined but not narrowly specified or confidently predicted. But this category, too, would be imperiled by some of the formulations circulating on what assurances funding agencies would need to provide to Congress.

“Should the NSF Director of the time have had to predict, for Congress, that the NSF grant to Stanford graduate students Larry Page and Sergey Brin for an investigation of search algorithms would lead to a global revolution in how people find information? Or to the founding of a company whose market capitalization stood yesterday at $271 billion?

“The fact is that nobody can predict where new understandings developed in fundamental research will ultimately lead -- and what benefits to society will ultimately result. Even in applied research, it is rarely possible to predict with confidence whether the work will achieve its intended goal or not, never mind what ultimate benefit might follow from achieving that goal, or some other outcome of the research that was not even envisioned at the outset of the work. A corollary of this reality is that some well conceived research projects will fail utterly. That is the nature of research. If you do not end up funding some failures, you are not funding anything very interesting. Certainly you are not funding the sort of cutting-edge, high risk research that, if successful, can be transformative.

“No system of deciding what research the Federal Government should fund will succeed perfectly, whatever the standard of perfection. But the overall degree of success of the competitive, peer-reviewed grant process that is employed by the NSF, the NIH, and in much of the rest of the government’s R&D funding - success measured by the pace of advance in basic science and the pace of the applied breakthroughs - has made that peer-review-based process the gold-standard, recognized around the world.

“The President made it very clear on Monday that this Administration will do everything it can to protect that gold standard -- to ensure: ‘that our scientific research does not fall victim to political maneuvers or agendas that in some ways would impact on the integrity of the scientific process.’ I couldn’t agree more.

“Of course, this doesn’t mean that the way this peer-review process is implemented in different agencies shouldn’t itself be reviewed from time to time to make sure it is as good as it can be. But fiddling in any fundamental way with the model of judging research proposals via review by scientific experts in the relevant fields would place at risk the world-leading quality of this Nation’s scientific and engineering enterprises.”