A Broader Vision for Government Research

Too many federal agencies lack the research funds they need to stimulate innovative solutions to national problems.

In recent years, the question of “balance” has been a hot topic in science policy circles. Although federal support for biomedical research has increased significantly, support for the physical sciences, engineering, and social sciences has been flat or has even declined in real terms. This is an important issue and clearly needs to be addressed.

However, there is another sense in which the government’s research portfolio lacks balance. R&D plays an important role in meeting some national goals, such as national security, health, and space exploration. However, for other national goals, such as improving student performance, promoting sustainable development, and empowering more Americans to lift themselves out of poverty, it plays little or no role.

Noted science fiction author William Gibson once observed that “The future is here, it’s just not evenly distributed.” A similar comment could be made about the ability of the U.S. government and the research community to help create the future. Some agencies, such as the Departments of Defense and Energy, the National Institutes of Health (NIH), and the National Aeronautics and Space Administration, have the budget and the capacity to support research and innovation on problems that are related to their mission. Others, including the Environmental Protection Agency (EPA), the U.S. Agency for International Development, and the Departments of Education, Labor, State, and Housing and Urban Development, have little or no such capability.

This imbalance may limit our ability to support research in what Donald Stokes referred to as “Pasteur’s Quadrant” and what Gerald Holton and Gerhard Sonnert have called Jeffersonian science: research that pursues fundamental understanding but is also motivated by consideration of some practical problem. If the agency charged with advancing a particular set of national goals has little or no ability to support research, and if such research is not an attractive investment for firms, there may be important and systemic gaps in the nation’s research portfolio.

Moreover, these gaps are likely to persist over time. Federal budgets tend to be incremental. If an agency has little or no money for research, this is not likely to change from one year to the next. Furthermore, agencies don’t know what they don’t know. If they don’t have a strong relationship with the research enterprise, they are unlikely to appreciate how research funding might help them achieve their objectives or what scientific and technological advances they might be able to exploit. Finally, the lack of a vibrant community of leading researchers interested in a particular problem deters agencies from trying to create one. There may be lengthy startup costs associated with creating such a community, during which time there may be little or no payoff.

To evaluate the opportunity costs of these gaps, it is necessary to understand the benefits that can flow from creating a high-quality, well-supported, multidisciplinary community of researchers that are interested in helping to meet a particular national policy objective.

How research helps

First, research can advance the state of the art in an area of science and technology that will make it easier or less expensive to meet a given national goal, or even reframe the way that a policy issue is debated or discussed. For example, EPA currently pursues the goal of a cleaner environment primarily through command and control regulation, as opposed to supporting the creation and diffusion of technologies that minimize pollution in the first place. Greater emphasis on the latter approach might allow the United States to achieve its environmental objectives while reducing the economic costs imposed by regulations. Increased support for research in experimental economics and mechanism design would allow policymakers to understand when and how to use market-oriented mechanisms such as the EPA’s “cap-and-trade” program for acid rain. The old adage that “if all you have is a hammer, the whole world looks like a nail” is certainly true for federal agencies.

Second, research can help create a more rigorous basis for making decisions or setting public policy. For example, some areas of policy (such as welfare policy and adult training) have benefited significantly from randomized field trials analogous to the clinical trials conducted by medical researchers. The researcher randomly assigns some individuals to a control group and others to an experimental group that receives the “treatment” that is being evaluated. Researchers generally have greater confidence in conclusions reached by randomized field trials than those produced by nonexperimental research. Although randomized field trials are not always feasible, and they cannot shed light on all policy questions of interest, they are clearly underutilized in some important policy areas such as education. A recent analysis of 144 contracts for program evaluation awarded by the Department of Education between 1995 and 1997 found that only 5 used a randomized controlled design to measure the impact of federal programs. Social science research can also help shed light on the broader economic and social context that will shape and in turn be shaped by technological advances.

Third, government support for research, particularly university-based research, helps create or expand a workforce with specialized skills. Creating such a workforce may be critical to achieving a particular policy objective. Recently, for example, the federal government has acted to increase the number of undergraduate and graduate students with a background in cybersecurity, since the government was unable to recruit enough people with the necessary skills. Similarly, NIH has recognized that exploiting the revolution in genomics will require addressing the shortage of researchers in bioinformatics. Agencies that lack the ability to support university-based research, fellowships, and traineeships will not be able to help create this kind of specialized workforce.

Fourth, government support for research can lead to innovation in the development and use of new technologies. Researchers can start new firms and transfer their technologies to existing firms. They can suggest “figures of merit” that create new metrics for measuring technological progress or develop open technical standards that serve as the basis for entirely new industries. They can create test beds that offer insights into the impact of novel combinations of technologies in real-world settings. They can help dramatically lower the cost of a given activity, such as sequencing genomes and storing, transmitting, and processing information. Finally, once a research community with an interest in tackling a given problem has been created, its members will be able to identify future advances in fundamental understanding or technological capability that are feasible and relevant. Although these predictions can trigger a backlash if researchers overpromise and underdeliver, they can help policymakers understand the potential benefits of investments in science and technology.

One useful thought experiment is to imagine what might happen if a federal department with a limited capacity to support research suddenly grew a research agency with the resources, reputation, and expertise equivalent to those of the Defense Advanced Research Projects Agency (DARPA). Imagine that EPA or Education or Labor now has a research arm with a budget of several billion dollars. It has no “entitled constituencies” and can support research in industry, academia, and national labs. Commercial companies and industrial researchers are more likely to interact with it, because it has been given special exemptions from some burdensome procurement and personnel regulations. It can support activities ranging from fundamental research to technology demonstrations and can also establish cash prizes to stimulate technological advances in areas of interest. It recruits entrepreneurial program managers that are the peers of the leading researchers in the country and encourages them to take risks. It also empowers them to create portfolios of research projects that will lead to outcomes that are greater than the sum of the parts, as opposed to automatically funding those proposals that received a favorable “priority score” from a peer review process. Its solicitations command instant attention from the university research community, because instead of offering the standard grant that can support only one or two graduate student, it is willing to support teams of faculty and their graduate students. It is willing to “make the peaks higher” and does not feel compelled to support researchers in all 50 states. Finally, it has a unique ability to support multidisciplinary research, because it is not organized around scientific disciplines.

Although this is an admittedly idealized portrait of DARPA, it is useful for stimulating our thinking about what could result from expanding the research capacity of a given agency. What medium- and long-term goals might this new agency set? What would be the societal payoff if these goals were achieved? Would the new agency fund research that would be significantly different from that done by the current federal research agencies? Do the research communities needed to perform the research exist, and if not, how might they be created? What existing and future technological waves–information technology, microsystems, biotechnology, nanotechnology–would this new agency seek to harness?

Funding gaps

I do not mean to suggest that more R&D will solve all national and global problems. In almost all instances, R&D is likely to be just one component of a broader response to a given public policy challenge. But I do believe that R&D is underutilized as an instrument of national policy. One of the reasons for this is that the capacity to support research varies widely across the federal government. Policymakers and researchers take the current distribution of science and technology capabilities across the federal government as a given, and it rarely undergoes any serious consideration or discussion. Below are a few examples where we are not spending enough as a nation, although in all cases there is at least some level of effort.

Improving education and lifelong learning. A 1997 report by the President’s Council of Advisors on Science and Technology observed that we were investing less than 0.1 percent of total K-12 expenditures on R&D, as compared to the 23 percent R&D-to-sales ratio in the pharmaceutical industry. This is an area where the United States is clearly underinvesting, given the importance of education and lifelong learning to a well-functioning democracy and our long-run standard of living. As mentioned above, there has been little experimental research on K-12 education. A 2002 report by the Coalition for Evidence-Based Policy concluded that “existing practices have rarely or never been tested”; these include standards setting, whole-school reform, charter schools, math and science curricula, teacher training, and language instruction for students with limited English proficiency.

In addition, although the United States has made a large investment in providing schools with computers and Internet access, very little research has been done to advance the state of the art of educational technology and to understand how it might best be used to support teaching and learning. Possible goals identified by the researchers include developing software that approaches the effectiveness of a one-on-one tutor; creating interactive simulations that allow students to engage in learning by doing; embedding assessment in learning environments that can be used to continuously guide instruction; enabling students and subject-matter experts (as opposed to programmers) to rapidly create high-quality, reusable educational content; and constructing collaboration tools that allow for the sharing of expertise between peers, tutors, and experts. Many experts believe that the federal government has not supported grants with the necessary scale and scope. Educational technology research projects conducted in a lab or a single classroom may not shed any light on what it will take to move these innovations into everyday use in a large urban school district.

Although the National Science Foundation (NSF) has made some investments in these areas, the Department of Education has made almost no investment in educational technology R&D and has supported very few assessments of educational practices using random assignment. The administration’s proposed FY 2004 R&D budget for the Department of Education is $275 million, or only 0.2 percent of total federal R&D expenditures. The Department of Labor has no R&D budget. As such, it has done very little to explore the role that technology might play to allow adult learners to acquire skills more rapidly and conveniently.

Environmental technologies. There are many interesting ideas for technologies that could contribute to a cleaner environment and sustainable development through better monitoring and assessment, pollution minimization, and environmental remediation. Examples include “green chemistry,” designing for reuse and remanufacturing, and engineered microorganisms for remediation. Unfortunately, EPA, the agency that might be expected to take the lead on such an initiative, has almost no money for research in this area. Most of EPA’s modest research budget supports regulatory decisionmaking and the assessment of environmental and human health risks; little is left to support the development of technologies that would minimize pollution to begin with.

This clearly leads to some missed opportunities. EPA invests only $2 million in the $2.18 billion multiagency information technology initiative and $5 million in the $847 million nanotechnology initiative. At these levels of funding, EPA is unable to adequately support the many potential applications of nanotechnology that it has identified, such as filtration, remediation, environmentally benign manufacturing, and low-cost, highly selective sensors. For example, an EPA that was given an expanded R&D budget and charter could develop an ambitious strategy to put environmental monitoring on the Moore’s Law curve, leveraging recent advances in microfluidics, nanotechnology-based sensors, and wireless networks. This might allow individual communities to have continuous detailed information about air and water quality in their neighborhoods.

Science and technology for developing countries. Many believe that science and technology could play a much greater role in fostering economic and human development in developing countries. For example, according to the Global Forum for Health Research, only 10 percent of global medical research is devoted to diseases that cause 90 percent of the health burden in the world. Of the 1,233 drugs that reached the global market between 1975 and 1997, only 13 were for the treatment of tropical infectious diseases that primarily affect the poor in developing countries. More broadly, researchers could develop technologies that could help address a variety of challenges faced by developing countries and expand access to clean water, cleaner sources of energy, and affordable information and communications services. To date, however, developing countries have had a limited ability to support R&D, and development agencies such as USAID and the World Bank have not viewed supporting innovation as an important part of their mission.

In short, many potentially important areas of science and technology are not adequately funded. We have a “revolution in military affairs” but no “revolution in diplomatic affairs,” because the Defense Department has a $62.8 billion R&D budget and the State Department has none. We are not thinking systematically about the nonhealth applications of biology or “synthetic biology” because this is not the responsibility of NIH, the major government sponsor of biology. Agencies are embracing e-government to improve customer service and become more efficient, but no agency is sponsoring research and experimentation on “e-democracy”: the role of the Internet in expanding opportunities for informed participation and deliberative discourse.

What can be done

If the research community, foundations, and policymakers are interested in addressing this problem, there are several different possibilities worth exploring.

Universities and foundations could demonstrate what is possible. Although the federal government plays the dominant role in funding university-based research, universities and faculty members could identify areas of research that they believe are being underfunded; develop a compelling research agenda; and seek support from foundations, state governments, industry, and individual donors. With preliminary results in hand, they would be in a better position to demonstrate to Congress and the administration the potential benefits that might flow from increased investment in a particular area. For example, the University of California at Berkeley and other University of California campuses, with initial funding from the state of California, have launched the Center for Information Technology Research in the Interest of Society (CITRIS). This center is encouraging faculty to explore the use of advanced technologies to address a wide range of economic and societal challenges, many of which have not traditionally received much attention from the major federal research agencies.

Leadership might also come from foundations. Consider the critical role that the Rockefeller Foundation played in molecular biology beginning in the 1930s, the Whitaker Foundation’s critical investments in biomedical engineering, and the leadership that the Bill and Melinda Gates Foundation has demonstrated in global public health. Foundations looking for areas where they can make a unique contribution might choose to help “seed” an area of research that the federal government is neglecting. This is a potentially promising approach to explore a new area, but is probably not a substitute for federal investment over the long term.

NSF could be given the resources to significantly expand funding in important “gap” areas. NSF is the only agency with the mission to support science and engineering in all disciplines. In addition to evaluating proposals on the basis of their intellectual merit, it also considers the potential benefits to society of the proposed activity. NSF has a successful track record of working with other mission agencies to fund research that is motivated by specific national problems. The NSF/EPA Partnership for Environmental Research is a good example of the kind of collaboration that could be expanded with additional funding. There has been a great deal of support in Congress for doubling the NSF budget, but the current administration has rejected the proposal as being “formula-driven” and arbitrary. Identifying broad areas of use-inspired fundamental research that are currently being underfunded could help establish a more compelling case to increase the NSF budget.

There are several risks to this approach. The first is that NSF needs more money to do its core mission. The average size and duration of NSF grants are inadequate, particularly for multi-investigator awards, and NSF could easily absorb a doubling or tripling of its budget by addressing this problem. The second is that focusing the additional resources at NSF will not significantly change the culture of the relevant mission agency and may reduce opportunities for coordination between R&D and other public policies.

New research institutions could be created or the capability of an existing agency to support research could gradually be expanded. Policymakers could also consider creating new research agencies. A National Institute of Learning, for example, could fund research on cognitive science, educational psychology, advanced learning technologies, and large-scale rigorous evaluations of different proposals for K-12 reform. Another approach would be to increase the funding and capacity of an existing agency to support research. This might involve appointing or elevating the position of the chief scientist; using the National Academies to identify potentially relevant research issues; aggressively using the Intergovernmental Personnel Act to recruit researchers from universities and national labs; establishing external advisory committees; determining where an incremental investment might allow the agency to leverage science and technology funded by industry and other federal research agencies; and gradually increasing the agency’s R&D budget as it demonstrates the ability to successfully manage a high-quality extramural research program.

Science, technology, and innovation could be making a bigger contribution to a broader range of national and global goals. We should stop treating the current funding structure for research as a given and begin experimenting with different ways in which science and technology could help address our most pressing challenges.

Your participation enriches the conversation

Respond to the ideas raised in this essay by writing to [email protected]. And read what others are saying in our lively Forum section.

Cite this Article

Kalil, Thomas. “A Broader Vision for Government Research.” Issues in Science and Technology 19, no. 3 (Spring 2003).

Vol. XIX, No. 3, Spring 2003