"Science and Policy"
A talk given at the D. Allan Bromley Memorial Symposium
New Haven, Connecticut
December 8, 2005
My interactions with Allan Bromley were limited to a few brief periods, but they were close encounters and I was grateful for his friendship and advice. The first time I met Allan was during my term as chairman of Universities Research Association. URA had the Department of Energy contract to build and operate the Superconducting Super Collider in Texas starting in 1989, the year Allan became President George H. W. Bush's science advisor. Congress voted to terminate the project in 1993, the year Allan left Washington to return to Yale. You could say that the SSC survived as long as Allan was there to defend it, but of course the history of that project is complicated, with many contending forces. My experience with the SSC actually began about a decade earlier when the proton collider ISABELLE at Brookhaven Lab was in balance. Its fate was sealed when the heavy weak interaction vector bosons were discovered in 1983 at CERN, and the particle physics community realized that President Reagan would support the construction of a much larger next generation machine. While the collider was still in play, Stony Brook colleagues had urged me to help make the case for Brookhaven, and I used Congressman Bill Carney's conservative credentials to arrange a visit with Reagan's budget director David Stockman in 1981. But it was too late. The subsequent phoenix-like emergence of the Relativistic Heavy Ion Collider from the abandoned ISABELLE infrastructure probably owes something to the regional strength in nuclear physics to which Allan Bromley, of course, was a significant contributor.
Bromley always seemed to be at gatherings where physics and politics converged, and I saw him off and on at events in Washington during the next seven years. When the news broke that President George W. Bush had nominated me as his science advisor, Allan was among the first to contact me and offer his advice and assistance. He traveled across Long Island Sound to join me for a long lunch at a restaurant near Brookhaven National Laboratory, and afterwards we continued our conversation by phone. Looking back on those conversations, I realize that some of the most useful things he told me were anecdotes (which he very much enjoyed telling) about his experiences. He appreciated that each administration is a world unto itself, but that people are people and politics is politics, and his anecdotes illuminated a corner of the Washington scene with which I had no experience at all. As it turned out, Bromley and I were drawn to different aspects of the science advisory role (with respect to style in public affairs, I am a minimalist, he was a maximalist), and when I finally arrived on the scene that role had been altered further, if only temporarily, by the terrorist attacks of September 11, 2001. But we thought alike on substantive issues, and maintained contact until his untimely and unexpected death, and exchanged views on a wide variety of topics.
My title, Science and Policy, encompasses two distinct areas to which Allan was passionately devoted, and today I would like to reflect on these concepts and their mutual relationship from my own perspective. Before the White House personnel office called to suggest that I should be a candidate for this position, my interest in science policy was very narrow and wholly selfish. I wanted tools for physics, and more generally for the science programs I was trying to build, first at the University of Southern California, then at Stony Brook, and finally at Brookhaven. I was an advocate for my institutions, and I worked for their success, not for the success of science overall, or of the larger purposes it serves. So in the summer of 2001, at the age of sixty, I began to think seriously about policy for the first time. Allan Bromley's example was an important guide.
Bromley had set forth his admirably structured views of the aims and operation of the office of the science advisor in Yale's Silliman lectures of 1993, given while his memory was still fresh and his notes intact.1 (One had the impression that Allan's notes were always intact.) Supplements appeared later in books edited by William Golden, 2 the guru of science advice to U.S. presidents, and by the Technology and Policy Program at MIT in the proceedings of a symposium celebrating the 25th anniversary of the legislation that established the current version of the Office of Science and Technology Policy. 3 These books were useful to a policy neophyte, and they served me well as I contemplated my task in late 2001. At the time three issues – you might call them "meta-issues" impressed me as important for science policy today.
The first thing that struck me, as I looked back over the history of the U.S. government involvement with science, was how reactive the pattern of support was to more or less random external events. The mother of all such events was World War II, which came serendipitously on the heels of the discovery of nuclear fission, and brought science forcibly into contact with national affairs. It created an opportunity for Vannevar Bush, Bill Golden, and others to insert science permanently into the federal establishment, at least within the Executive branch, and it set the stage for science policy through the cold war and into the period when Bromley served the first President Bush.
As the nation's discretionary budget grew during this period, the science budget initially increased exponentially after Sputnik to a peak during the Apollo program, and then settled down to a relatively predictable pattern. Non-defense science funding held to a nearly constant fraction of the domestic discretionary budget, the share of the total rising and falling slightly about a slowly rising mean in a pattern roughly coincident with the solar sunspot cycle. Big science projects have come and gone, each with its own story, while the NIH budget ascended monotonically to its present dominance. Today NIH consumes nearly half the non-military federal research budget. NASA consumes about fifteen percent. NSF, DOE, and DOD basic science share nearly all the rest. I have struggled to identify a rational basis for this distribution of funds, and failed. Many observers, including the President's Council of Advisors on Science and Technology4 and a recent panel sponsored by the National Academy of Sciences,5 point to an imbalance in federal research support to the physical sciences as compared with the life sciences. And indeed there are many more imbalances than this one, depending on what you mean by "balance".
The potential irrationality of federal funding patterns was apparent early on the leading edge of the huge increase in science budgets in the early 1960's. Alvin Weinberg, in a 1961 Science magazine article that should be better known,6 said "…it is presumptuous for me to urge that we study biology on earth rather than biology in space, or physics in the nuclear binding-energy region, with its clear practical applications and its strong bearing on the rest of science, rather than physics in the Bev region, with its absence of practical applications and its very slight bearing on the rest of science. What I am urging is that these choices have become matters of high national policy. We cannot allow our over-all science strategy, when it involves such large sums, to be settled by default, or to be pre-empted by the group with the most skillful publicity department. We should have extensive debate on these over-all questions of scientific choice: we should make a choice, explain it, and then have the courage to stick to a course arrived at rationally."
Such a debate has never occurred in the science community, although Frank Press (who had served as President Carter's science advisor) tried to get one started in 1988 when he divided science programs into three priority categories and "named names" of projects that should be in each, urging scientists to take responsibility for setting priorities. It was not a popular proposal. Quotes from the subsequent media coverage make interesting reading. Here's Al Trivelpiece, then Executive Director of the American Association for the Advancement of Science: "Nobody asks farmers whether they want price supports for wheat rather than for cotton. Why should scientists be treated any differently and be required to choose from among several worthy projects? I think the issue for scientists should be the quality of the research." And a congressional staffer: "I hope we can forget his words and move on." And an official of the American Association of Medical Colleges: "It’s a question of strategy. Why should we assume that there’s a fixed pot of dollars? I prefer the idea that support for science is not fixed, at least not until we get to a level that represents a reasonable proportion of our GNP." Whatever you may think of the wisdom of Frank's statement, it was a call for rationality in the midst of a very irrational battle for federal funds. Part of that battle ended in 1993 when Congress voted the International Space Station up and the SSC down. Neither was at the top of Frank's list. Science advisors do not have the luxury of ignoring the need for prioritization. In a time of tight budgets it can be the most important issue in science policy.
The second "meta-issue" that seemed significant to me in 2001 was the interplay between basic and applied science and technology. During most of history, technology got on without science. We should keep in mind that nearly the whole of the industrial revolution occurred while scientists still thought heat was a material fluid. That changed toward the end of the nineteenth century, and the relationship between science and technology has been changing ever since.
Much of value has been written about the relationship between basic and applied science. Congressman Vern Ehlers emphasized the value of "targeted basic research" in his important 1998 report7 sketching a new post-cold war science policy. Gerald Holton spoke of "Jeffersonian science" in a 2000 conference on Science for Society whose proceedings8 Lewis Branscomb sent me in the summer of 2001 as I was meditating on these things. Princeton's Woodrow Wilson School dean Donald Stokes wrote an entire book titled "Pasteur’s Quadrant – Basic Science and Technological Innovation"9 where Vannevar Bush's "linear model" of the continuum of basic to applied research to technology was replaced by a two dimensional space. Probably more than two dimensions are needed here. The evolving complexity of this relationship was an important theme of an excellent and influential report10 produced in 1995 by a National Research Council Committee chaired by Frank Press that I want to dwell upon for a moment. This report introduced the new category of "Federal Science and Technology," or FS&T, into the science policy lexicon. The Office of Management and Budget decided to adopt such a category for the first time in the President's FY2002 budget proposal to Congress. The authors of the NRC report stated that "The committee's definition of FS&T deliberately blurs any distinction between basic and applied science or between science and technology. A complex relationship has evolved between basic and applied science and technology. In most instances, the linear sequential view of innovation is simplistic and misleading. Basic and applied science and technology are treated here as one inter-related enterprise, as they are conducted in the science and engineering schools of our universities and in federal laboratories."
This report is one of the more important science policy documents of the past decade, and it needs to be taken even more seriously than it has been. It bears on the significance of "development" (the "D" in R&D), and of industrial research, which are being given far too little credit in today's advocacy briefs for increasing federal support for science. As Alvin Weinberg realized already in 1961, it would be possible for us to double or triple funding for the overall basic research category and still not address the need for substantial investment in the kind of basic research that most effectively addresses societal needs. I had the impression, talking with Allan Bromley, that he understood better than most the complex processes that lead to innovation or economic competitiveness.
The third "meta-issue" in science policy that caught my eye four years ago is just how weak the tools of science policy really are. I made this point earlier this year in my address to the AAAS Science Policy Forum in April11, and in a subsequent editorial in Science magazine.12 In contrast with tax policy, where economic policymakers have a substantial body of ongoing scholarship to guide them, science policymakers have very few resources that help make choices among policy options. We have more data than we have models for interpreting it, and the data definitions are weak and not keeping pace with the changing practice and content of science. I think the situation is most serious in resolving questions about science and engineering workforce policy. What are the implications of globalization of technical work, rates of graduation in engineering and science programs in other countries, and the impact of information technology on research, design, and manufacturing? Empirical and theoretical bases for policy suggestions in this area are surprisingly weak. Some of these concerns surfaced in a recent NRC study I cited in my AAAS talk13, and OSTP strongly supports the recommendations in that report.
In reading over the key policy documents of the past decade in preparation for this talk, I came across a statement by former House Science Chairman George Brown bound with Vern Ehlers' 1998 report7 as a "Supplemental View." Brown declined to sign onto the report "because it does not sufficiently probe the depth of the problems facing our scientific enterprise. Any new policy should adhere to three principles which require more study. First, it should reflect our understanding of the process of creativity and innovation. Second, it should articulate the public's interest in supporting science – the goals and values the public should expect of the scientific enterprise. Finally, a new science policy should point towards decision-making tools for better investment choices." These three principles align well with the three "meta-issues" that seemed important to me in 2001.
I will conclude with a few remarks on concluding remarks that Bromley made in "The President's Scientists," the book1 based on his 1993 Silliman Lectures. "I was asked toward the end of my stay in Washington," he wrote, "what my biggest surprises had been. The answer was surprisingly obvious. First, the people with whom I had the privilege of working, both in the Administration and in the Congress, were much more able, dedicated, and hardworking than I expected. Second, it took longer to make anything happen than I could have believed possible!"
"There are some truly fundamental problems that require attention in Washington," he continued, "Perhaps the most important is the balkanization of Congress. We have too many committees and subcommittees, and there is no rational distribution of responsibility among them. What is needed is a complete restructuring of the way Congress does its business … "
Congress did implement a minor reconfiguration of its appropriations subcommittees this year, and it is too early to say whether it will help science. But reorganizing Congress is not going to be the solution to problems of irrationality in our federally supported science and technology programs. I think many of the irrationalities we see in our national behavior are a direct result of the complexity and chaos of events to which the Administration and Congress must respond. That first surprising discovery Allan made, that the people in government are able, dedicated, and hardworking, should give us hope that with better tools for understanding the chaos of events and for mapping out effective plans, our government – even in its present form – can do a better job of focusing its resources. Who better than the science community itself to provide these tools?
Allan Bromley's willingness to dedicate much of his life to the improvement of science policy formation will make it easier for others to follow his example. I am very pleased to be able to contribute to this symposium in his memory.
Director, Office of Science and Technology Policy
Executive Office of the President