Science Policy Tools: Time for an Update

All of us involved in science and technology (S&T) policy are fond of commenting on the increasing pace of change, the upheavals caused by novel technologies and expanded scientific understanding, and the unprecedented challenges to Earth’s resources and natural systems. Yet we typically find ourselves responding to these developments within the conceptual framework established by Vannevar Bush more than 60 years ago. From time to time, we need to step back from the specific challenges we face to reflect on the effectiveness of the assumptions, the strategies, and the institutions that shape our responses. Does the policy framework that underlies the U.S. S&T enterprise need to be updated?

My short answer is that many things need to change if the United States is to continue to be a leader in S&T and ensure that the American people are the beneficiaries. Government agencies and many other institutions are of a different era and ill-equipped to function well in today’s world. But to that I would add several caveats: Positive change is very hard to bring about in this system and usually comes slowly; there are fundamental elements of the Vannevar Bush philosophy that should be protected; and in the U.S. political system, especially at this moment in time, we should be careful what we ask for! With those provisos in mind, I will offer a few suggestions for policy changes that are feasible and would help move the country forward.

First, a word about Vannevar Bush and his 1945 report Science—the Endless Frontier. So much has been written and said about Vannevar Bush and his report that I have to be reminded occasionally to actually read it again. It really is an amazing document, for its content and foresight as well as its brevity (about 33 pages plus appendices in the National Science Foundation’s 1990 reprinted version).

Bush argued that science and the federal R&D system that proved to be so successful during World War II would be important to the nation’s progress in peacetime, which turned out to be dominated by the Cold War and arms race with the Soviet Union. Bush made three main points:

“Scientific progress is essential.” It is needed to meet the nation’s needs: the war against disease, national security, and public welfare. To accomplish this, he recommended federal support of basic research in universities and medical schools; strengthening applied research in federal agencies, guided by a Science Advisory Board reporting to both the executive and legislative branches; and creating incentives for industry to fund research.

“We must renew our scientific talent.” He recommended a program of federal support for scholarships and research fellowships, with special immediate attention given to those returning from the war.

“The lid must be lifted.” He recommended the formation of a board of civilian scientists and military officials to review all secret government scientific information and release, as quickly as possible, everything that did not have to be kept secret.

And to implement these recommendations, Bush put forward a plan of action that included the creation of a new civilian federal agency, the National Research Foundation (NRF), to take on the task of funding (basic) research and education in universities in all fields, including health, medicine, and long-range military research.

These three points and plan of action made up Bush’s vision and strategy to ensure that the federal government would continue its investment in science in the postwar years.

One further comment should be made about Bush’s report. He has often been criticized for oversimplifying his arguments for a robust federal research investment by accepting a linear model of progress: Basic research should be carried out without an application in mind; basic and applied research are at opposite poles; all technological advances are the result of research; and the nation that does the research will reap most of the benefits. To some extent, these notions are as much a reflection of how the public and policymakers thought about the role of science in World War II as they were statements of fact by Bush. It can be argued that none of these is entirely correct. Indeed, Bush himself would have agreed. But I offer a word of caution. Although scientists are comfortable engaging in “nonlinear thinking,” the same is not true for the general public and most policymakers. So although it is useful to revisit Bush’s assumptions in an effort to craft the most effective means to argue for the importance (perhaps even unique importance) of S&T, we should proceed with caution, lest we find that the message that is received is not the message intended.

The Bush effect

Much of Vannevar Bush’s vision has come to pass, even if not entirely as he intended. But today’s world is very different from that at the end of World War II, and Bush could not have been expected to foresee developments such as globalization and the rise of multinational corporations, the collapse of the Soviet Union and the rise of terrorism, Moore’s Law and the information revolution, erratic swings in U.S. politics, and other factors that have placed S&T in a precarious place in 21st-century U.S. society.

Bush’s notion that “scientific progress is essential” to meet the nation’s health, national security, and public welfare needs has become accepted by policymakers of all political stripes and by the majority of voters. That said, the genuine need to address immediate national issues, such as unemployment, the lack of affordable health care, inadequate K-12 education for most Americans, and many others, tends to crowd out important long-range goals, including investments in basic research. But even for short-term objectives, the uncoordinated federal support structure for R&D is ineffective in aligning more-applied R&D with urgent national needs.

The public tends to support the investment of taxpayer dollars in scientific research but has little understanding of science, particularly the nature of research or how results are translated into things people need. Aside from medicine, it is not easy for the public to see the connections between research and the things that are most important in their lives. Moreover, deep partisan divides on almost all issues and a media focused on entertainment rather than news make it almost impossible for the public to actually know what is going on in the country, let alone the rest of the world, in S&T or anything else. This disconnect with the public is perhaps the greatest threat to the future of the country’s research system.

Bush’s advice “to renew our science, engineering and technical talent” remains a priority(at least it is a subject of much study and political rhetoric), but the nation’s efforts to attract homegrown boys and girls to these careers, as well as to improve science, technology, engineering, and mathematics education have been disappointing. The United States has been fortunate in attracting many of the brightest young women and men from other parts of the world to study and establish their careers here. But current U.S. policies and practices on visas and export controls are making the country less attractive as a place to study and work. Increasingly, bright young people are finding attractive opportunities elsewhere.

Bush’s recommendation that the “lid be lifted” on classified information was influential, at least initially. But there remain issues of overclassification and ambiguous categories such as “sensitive but unclassified.” In spite of laws designed to shine light on government, some federal agencies are inclined to hold back information that might be inconvenient or embarrassing. In addition, the imperative to make all data resulting from federally supported research available to researchers who want to confirm or refute various scientific claims has become especially challenging, in part because of the enormous volume of data involved, the cost of making it available to others, the need for software to interpret the data, and other factors. Yet the integrity of the scientific process depends on this kind of openness. The National Academies have focused on this issue and made recommendations in the 2009 National Research Council report Ensuring the Integrity, Accessibility, and Stewardship of Research Data in the Digital Age.

Bush’s “plan of action” to establish the NRF did not happen, at least not as he proposed. The National Science Foundation (NSF), with its presidentially appointed director and National Science Board, was established in 1950, but it was a very different agency with a narrower mission and a much smaller budget than Bush envisioned. It is doubtful that Bush’s model of the NRF would have been successful. Had it been established, pieces would probably have spun off as fields evolved and generated their own interested constituencies.

Bush’s so-called linear arguments still, I believe, continue to underpin the continued political support for federal funding, even if the processes leading to discovery, engineering design, innovation, and application are recognized to be more complex than Bush had indicated. Finding the right language to explain to the public how scientific discoveries make their way to applications and markets remains a work in progress.

Finally, Bush’s government-university (GU) partnership, in which federal agencies provide funding for academic research and for the construction and operation of experimental facilities at national and international laboratories that are open to university researchers, remains in place, at least for now. This partnership is, perhaps, the most important outcome of Bush’s report. With the federal agencies doing their best to select the “best people with the best ideas” as determined by expert peer review, the standards have been kept high, and the political manipulation of the research through congressional earmarking has been minimized. The resulting integration of research and education in the classrooms and laboratories of universities across the country has enabled the United States to build the most highly respected system of higher education in the world.

Another strength of the GU partnership, on the government side, is the plurality and diversity of federal agencies, each with a different mission and structure, that support academic research. It’s not what Bush intended, but it has benefits. Academic researchers can propose their ideas to several agencies, looking for the best fit and timing as well as sympathetic reviewers. Also, the agencies can focus their support on areas of science and engineering that are most relevant to their missions.

In recent decades, especially after passage of the Bayh-Dole Act in 1980, universities have established partnerships with companies, both as a means of providing their students with better access to industry and future jobs, and as a possible generator of revenues from the intellectual property created by faculty researchers. Thus, the two-way GU partnership has evolved into a more complex government-university-industry (GUI) partnership, and that trend is likely to continue. Industry’s support has grown steadily but still constitutes only 6% of total academic research funding, compared with 60% from the federal government, 19% from university funds, 7% from state and local governments, and 8% from other sources.

Thus, although much has changed at home and around the world, Bush’s GU (now GUI) partnership remains in place and is perhaps the most important outcome of Bush’s report. But now all three sectors are under stress, pressed to meet rising expectations with limited resources, and in the case of industry, faced with intense competition from abroad. The partnership is in trouble.

A troubled partnership

Universities are going through a difficult period of transition as the costs of higher education and, consequently, tuition continue to rise, and governors and legislators demand greater accountability while cutting their states’ contributions to public universities. It is not clear that the term “public university” should be applied to institutions that receive only 20% or less of their budgets from the state government. Private universities are less affected by state politics, but all universities face the burden of complying with a growing body of uncoordinated federal regulations and other reporting requirements related to faculty research that add more to the institutions’ costs than the 26% maximum overhead reimbursement for administrative costs that comes with government grants.

Universities with major investments in biomedical research, especially those with medical schools, face special challenges in planning and operations. When National Institutes of Health (NIH) budgets are going up, institutions hire more researchers (non–tenure-track faculty and postdoctoral researchers) and borrow money to build new buildings, with the expectation that the overhead on future grants will support the salaries and pay off the loans. When NIH budgets fail to match expectations, the institutions are left to cover the costs. The model makes sense only if NIH budgets continue to rise at a predictable rate, indefinitely. Large swings in NIH funding since the 1990s have exacerbated the situation, and thousands of bright young biomedical researchers have ended up in long-term postdoctoral positions or have left the field. As former National Academy of Sciences President Bruce Alberts has observed, it appears that the GU partnership in this important field is not sustainable.

One more point about the university side of the GU partnership. The nation is producing more Ph.D.s in some fields than there are jobs, or at least jobs that graduates want and are being trained for, and the imbalance is likely to get worse. This may be especially true for biomedical research, but it is a problem for some other fields as well. The nation may need more scientists and engineers, even Ph.D.s, but not in all areas. There are policy options that could be employed, but they are not easy. Given the rapid pace of change in this age of technology, it is simply not possible to predict what the specific needs will be 20 years from now. The ability of the United States to continue to be a world leader in technological innovation will depend not only on having the necessary technical talent, homegrown and from abroad, but also being able to retrain and redirect that talent in response to new developments. At the very least, universities need to ensure that their graduate programs include curricula and mentoring that will adequately prepare their students for careers very different from those of their professors. Some of the professional master’s programs do a good job and should be expanded so that all students pursuing graduate study have the option to earn a professional master’s degree, even if the Ph.D. is their ultimate objective.

Federal agencies have their own problems. Because the agencies’ budgets have not grown at a rate that matches the expansion of research opportunities, NSF, NIH, NASA, the Department of Energy (DOE), and other agencies are unable to provide adequate support for even the most meritorious research proposals and the necessary experimental facilities. Most researchers must apply for and manage multiple research grants to support a viable program and thus end up spending much of their time dealing with administration rather than science. The multiple grant applications and reviews also add to the administrative costs of the agencies, cutting into the efficiency of their operations. The agencies must try to plan with little certainty about their budgets for the next fiscal year or even the current fiscal year.

Every president has coherent goals and priorities when the budget request is put together, but Congress often does not share the president’s priorities, yet has no plan to offer as an alternative. It becomes painfully clear in the subsequent budget sausage-making that there is no consensus on national priorities or goals, no process to decide on optimum funding allocations, and no mechanism to provide stability in funding. It’s up to each agency to fight for the resources it needs to do its job, as required by law, regardless of how politically unpopular that job may be for a particular administration or Congress or, more accurately, the House and Senate appropriation subcommittees. The agencies complain that their operations are micromanaged by the subcommittees, and their decisions are often attacked by members of Congress who object to a grant with “sex” in the title or to a whole field such as climate science because they are unhappy with what the research reveals.

Congress has no mechanism to have a serious discussion about S&T. The congressional committee structure, at least as regards S&T, makes little sense. Neither the House nor the Senate has an authorization or an appropriations committee that takes a broad overview of the entire federal S&T portfolio. The House Committee on Science, Space and Technology is the closest thing, but its jurisdiction does not include NIH. In addition, Congress does not have any S&T advisory committees, at least not any that are visible. The authorization legislation for the defunded Office of Technology Assessment is still in place, and it would be a step in the right direction for Congress to again appropriate funds for it. One research funding agency that has enjoyed favorable treatment by Congress over several decades, at least through 2003, is NIH. As a result, the NIH budget is roughly half of all federal research funding. But NIH has had a history of boom-bust budget fluctuations. The budget doubled between 1998 and 2003, remained flat through 2008, received a $10 billion infusion as part of the 2009 stimulus package, and has remained flat since. NSF and the DOE also have had to manage the stimulus bump. Managing such rapid ups and downs is difficult. The impact on universities, as has already been noted, can be severe. NIH Director Francis Collins has been fairly candid with his observations and cautions. There is also the larger question of balance. Should the nation be devoting 50% of its research funding to biomedical research? It was just above 30% in the early 1990s. That might be the proper share, and voters are not complaining. But progress has been slow in many areas, and medical costs have been taking a steadily rising share of the gross national product. The problem is that the U.S. political system lacks a mechanism to even discuss balance or priorities in research funding.

Several agencies have to deal with other issues of balance in the federal R&D portfolio and the GU partnership. I’ll use the DOE national laboratories as an example. Clearly, the DOE labs have important functions. Because universities must provide an open environment for study and research, they are not appropriate sites for the type of classified weapons work being conducted at Los Alamos, Lawrence Livermore, and Sandia. Nor can universities afford to build and maintain the large experimental facilities of Fermilab, Brookhaven, Argonne, Jefferson, and others. The national labs also have the capability, at least in principle, of responding quickly to national needs. But there are some troubling issues with regard to all national labs: The roles of the labs were clear during World War II and the early Cold War years, but that is no longer the case. The labs cope with mixed signals from Washington and the ever-shifting political winds as agency heads come and go and White House and congressional priorities change. The nation probably does not need so many national labs with overlapping missions competing with one another for resources. But closing a lab can cause great hardship for states and communities. The process would require a research lab closing commission, and it could get very ugly. Science could become be even further politicized. What could add substantial value to the federal R&D investment would be a much stronger research collaboration between university researchers and federal laboratories, not only those that harbor large experimental facilities but the other general-purpose laboratories as well. Accomplishing that would require significant changes in how the agencies fund R&D and how they manage their national labs. It might be worth running a few pilot programs to explore the possibilities.

Beyond the matters of balance, a number of other issues are troublesome for the agencies, including trends toward short-term focus, demands for deliverables, and increased accountability (assessment, milestones, roadmaps, etc.); a conservative peer-review system that is risk-averse; contentious issues of cost-sharing and overhead; challenges of planning, cost, and management of large research facilities; and political barriers to international collaboration. Some of these matters have been discussed by the National Academies, the National Science Board, and more recently by the American Academy of Arts and Sciences.

One further comment about the federal agencies. Just as the academic researchers supported by the government are expected to hold to the highest standards of performance in carrying out research and disseminating the results for the public good, the federal government, in turn, should be expected to operate in a manner that is open, transparent, fair, and honest. In other words it should manifest integrity. The abuses can be in both the executive and legislative branches. We have seen, not too long ago, that any science that seems to violate someone’s special interests—religious, ideological, or financial—is fair game for attacks, including amendments offered on the floor of the House of Representatives to kill specific NIH grants that are judged by some members to be offensive or wasteful of money. The integrity guidelines laid out by President Obama will help ensure that federal agencies do their part. There is no corresponding commitment on the part of Congress. The integrity of the GU partnership requires responsible behavior on all sides. To the extent that the partnership lacks integrity, the American people are denied the benefits and can, in some cases, be harmed.

With both sides of the GU partnership having problems, it should be no surprise to find that the partnership is in trouble. The risk of doing nothing to address the problems is substantial. The National Academies report Rising Above the Gathering Storm and its recent update point out that clouds are gathering that threaten the nation’s S&T enterprise and its standing in the world. And although the Gathering Storm stresses the threat to the competitiveness of U.S. industry and the related matter of quality jobs for Americans, the arguments also apply to other national needs such as national security, health and safety, environmental protection, energy, and many others that also depend on the nation’s strength in S&T and its science and engineering workforce. It is likely that Vannevar Bush would see the need for another path-breaking report to address the question of whether science in the United States is still “the endless frontier.”

A way forward

I will pass on the option of trying to be the next Vannevar Bush by proposing a new government science policy structure, but I will suggest three areas of possible policy reform that do not require reorganizing the federal government or challenging congressional authority. None of these are fully developed proposals, but they could be useful as a stimulus to discussion.

First, the federal government is in dire need of an enhanced interagency mechanism to coordinate S&T- related activities, share information, and work with Congress to obtain more flexibility in funding interagency activities. The whole of the federal S&T effort should be significantly greater than the parts. The National Science and Technology Council (NSTC) and its coordinating committees have done good work, for example, in helping to organize the National Nanotechnology Initiative in the Clinton administration, but the NSTC needs more clout. The White House and Congress should consider authorizing the NSTC and providing a line of funding in the White House Office of Science and Technology Policy budget for NSTC staffing and activities such as reports, workshops, and seed funding for interagency cooperative R&D efforts.

Second, the federal R&D agencies should be encouraged to experiment with new structures modeled on the Department of Defense’s Advanced Research Projects Agency (DARPA or ARPA at different times) that can invest in higher-risk, potentially transformative R&D and respond quickly to new opportunities. The DOE is trying such an experiment with ARPA-Energy, which was launched with funds from the stimulus package and is included in the president’s fiscal year 2012 budget request. Examples of other new initiatives are DOE’s Energy Innovation Hubs and the National Oceanic and Atmospheric Administration’s Climate Service. Political inertia is difficult to overcome, so initiatives of this kind will gain traction only with leadership by the president and S&T champions in Congress.

The third is more of a stretch. The nation may have arrived at a time in its history when it needs a new kind of policy-oriented, nonpartisan organization: a GUI policy partnership among the federal government, universities, and industry, with funding from all three, that could address important areas of U.S. S&T policy such as the conduct of research and mechanisms for the translation of research into applications. This organization would be a place where knowledgeable individuals who have experience in the relevant GUI sectors and who have a stake in the health of the nation’s S&T enterprise would have access to the relevant data and policy analysis and could engage in serious discussions about a range of policy issues related to S&T.

Such a GUI policy organization would support policy research in areas of strategic importance, collect and analyze relevant information about the state of S&T in the nation and the world, and perhaps develop policy options for consideration by decisionmakers. This organization might be able to fill the pivotal role that Roger Pielke Jr. calls the “honest broker.” It would not go beyond defining policy options, leaving the final choice of direction to elected officials or whoever is responsible. Were it to make recommendations for a specific course of action, it would soon find its independence and integrity challenged as competing interests sought to influence its decisions. But even without advocating specific actions, an organization that is respected for the integrity of its data and analysis and the transparency of its operations would be of enormous value. Its credibility and political clout would derive from its grounding in the three critical sectors. There are many excellent nongovernment policy centers and other organizations that carry out policy research and issue reports, and their important work should continue. But there is no mechanism to follow through, to make sure someone is paying attention, to ask if any of the recommendations are being considered, and to explain to a largely uninformed public the implications of various policy options and report on subsequent decisions in a way that the public can understand.

The new GUI policy organization could take on many of the issues mentioned above, especially the problems the federal funding agencies are facing. Which of those are the most serious and might lend themselves to solutions short of reorganizing government? What are the most important policy barriers to cooperation between universities and industry and how might those be resolved? Are there ways to make a rational judgment about the various balance issues with regard to research funding? One task that such a new organization might take on is an analysis of trends in the respective roles of the federal government, research universities, and industry in the process of innovation. And here I mean innovation in both commercial products and processes, as well as how the federal government addresses various national needs.

Commercial innovation has been advanced as one of the prime rationales for increasing the federal investment in R&D and science education. Certainly, commercial innovation is vital to the nation’s future, but so is innovation in applying discoveries and inventions to national security, human health and food safety, energy security and environmental protection, transportation, and the many other societal needs that require new ideas and new technologies. In particular, the federal regulatory agencies have the task of complying with federal law by issuing rules that are consistent with the best scientific evidence, even when the evidence is not clear. Too often, the process, at least as portrayed by the media, looks more like a shootout between the affected industries and activists of various kinds than it does an evidence-based deliberative process. There are many policy issues relating to the regulatory process that could benefit by having the attention of an unbiased organization that is respected by all interested parties.

The new GUI policy organization should not attempt to duplicate the important work of the National Academies and National Research Council or the American Academy of Arts and Sciences or any other organizations. Nor would it replace the many outstanding policy institutes and centers around the country. NSF is authorized by Congress to collect and disseminate information about S&T, and the National Science Board publishes updated summaries in its Science and Engineering Indicators. The American Association for the Advancement of Science (AAAS) also is an invaluable source of S&T policy information, particularly R&D funding data. These efforts should continue. Indeed, one could imagine an alliance of organizations, governmental and nongovernmental, with common goals in support of a more rational national S&T policy. The kind of organization I am suggesting is not a Department of Science and Technology or a reinvention of Bush’s NRF. The present federal R&D funding agencies will remain in place, hopefully making improvements in their structures and operations. The latter is more likely if the agencies have a source of sound analysis and advice and, at least as important, public support for the changes they need to make.

As for the option of reorganizing the federal government, we could each devise an “ideal” structure that would do all the things we think need to be done. The only problem is that we would have to ignore the political realities. In the U.S. system of governance, structural change is very difficult, and no matter how elegant the proposal, what emerges from the political process is likely to be a disappointing, if not disastrous, caricature of what was proposed. Those of us who have struggled in the labyrinth of S&T policymaking have often dreamed that some wise reorganization will come along, but waiting for that to happen is not a likely path to progress. However, the failure of that dream solution is not a reason to abandon hope for more targeted incremental reform. There are many paths that could move the country in the direction of a more rational and inclusive approach to S&T policymaking.

I see the potential to take a few initial steps by generating synergy among some existing efforts. The National Academies have the Government-University-Industry Research Roundtable that meets regularly to discuss issues at the GUI interface. The Council on Competitiveness is an important forum for discussions about S&T’s role in commercial innovation. The American Association of Universities also focuses on the partnership, including federal research funding. All the major disciplinary societies have policy committees that deal with matters relevant to their memberships. AAAS has enormous convening power. Many other organizations pay close attention to policy matters that affect the GUI partnership. Perhaps some of these would entertain discussions about such a GUI policy initiative. It might even be a good agenda item for the President’s Council of Advisors on Science and Technology. And although I recognize that Congress is locked in an ideological battle that makes coherent action on any topic seem unlikely, perhaps one or more of the relevant congressional committees might give the idea some thought and perhaps discover some elusive common ground.