Call for a Department of Science not new

By Fred W. Weingarten
CRA Staff

One of the unusual characteristics of research funding by the US government, when compared with that of other countries, is its fragmented nature. Research programs exist in literally dozens of agencies. The structure of support that has evolved since World War II has mission agencies concentrating their funding in particular areas and styles of research tailored to their individual needs. These programs often support long-term research in university laboratories. But at some level it needs to be tied to the agencies' needs.

Complementing the focused support of the mission agencies is the National Science Foundation. A major goal of NSF is the broad support of fundamental research efforts.

This structure seems to have worked well, and the pattern has been supported by most people in the science establishment. However, it has not pleased everyone. There have always been critics who think the structure of government science is chaotic and nearly unmanageable. Hence, over the years, some science policy experts and politicians periodically have suggested creating a larger, cabinet-level Department of Science.

This proposal nearly always surfaces every time a new administration comes into office. Looking for ways to reorganize government is a favorite way for new presidents to show that they have come to town to change things. The particular structure of the new department differs, but most proposals would combine NSF with NASA and several other civilian research agencies. (Most proposals leave defense research programs and the National Institutes of Health, the two largest sources of government research funding, out of the new department.) In the past, these proposals have quickly sunk without a trace.

With this history, it should not be surprising that another proposal is being circulated, this time by Rep. Bob Walker (R-PA), who is chair of the House Committee on Science. What is surprising are the tremors that went through the science establishment and agency officials when Walker raised the idea. With all the emphasis on change and the eagerness of the Republican Congress to re-examine the fundamentals of government, some observers think it more likely that such a proposal could pass. Another possibility is that, with all the new voices being heard in the science policy arena, the leadership has no effective filter with which to separate the serious proposals from the perfunctory.

The boundary conditions for the debate have changed.

• Growth of cross-agency programs. In the late 1980s, the Bush administration began to develop interagency research programs coordinated through the Office of Science and Technology Policy. The High-Performance Computing and Communications Initiative was an early example of such a program, and it is still the largest one. The Clinton administration created an even more elaborate coordination structure and, in its reinventing government report, referred to creating "Virtual Science Agencies."
• Tight budgets. Research budgets have stopped growing and may be on the verge of dropping precipitously if the long-term budget proposals in Congress come to pass. In such an environment, there is much greater pressure to prioritize and to avoid duplication.
• Orphan R&D programs and facilities. In addition to cutting research budgets, the congressional budget resolutions, particularly the House's, propose the elimination of whole departments. Some of those departments, in particular the departments of Commerce and Energy, encompass some large research facilities and science offices. DOE has its national laboratories. Commerce has organizations such as the National Institute of Standards and Technology and the National Oceanic and Atmospheric Administration. Assuming there may not be a political desire to close those particular operations, they need to be housed somewhere.
• The growing importance of science and technology. It's an argument that may sound a bit self-serving, but, clearly, a strong science and technology base has become central to the economic and social well-being of the nation. Cabinet-level representation in the government would reflect that importance and move S&T to a more central point in government decision making.

In Walker's view, these reasons all point to the need for a Science Department, although, as he admits, the arguments for forming a new agency fly in the face of the strong desire in his own party to reduce government.

The historical arguments against a Science Department still seem strong, the strongest being the argument for diversity in research funding. Differences among agencies in perspectives, priorities and funding strategies can increase the opportunities for new ideas or even new fields to be developed.

Computing research is a good example. In the late 1960s, while NSF officials were trying to convince a conservative and traditional management that computer science was a potentially new and exciting field for research investments, the Advanced Research Projects Agency was pushing ahead aggressively and pouring large grants into a few research programs at Stanford University, Carnegie Mellon University, the Massachusetts Institute of Technology and other major universities. This was a strong stimulus to the growth of US computer science and engineering, in terms of creating a discipline and educating the first generation of computer researchers actually trained in that discipline.

Meanwhile, NSF continued to expand its programs of computing support in its traditional model of a single-investigator, competitive grant, which opened up support to the broader group of researchers not located at ARPA-funded centers. NSF and ARPA, with comparable funding levels in computing research, still operate in different but clearly complementary modes of support to the field.

DOE's national laboratories were a third major influence on computing research. They stimulated the development of supercomputer technology, particularly through the 1960s and 1970s. Needing substantial computing horsepower, mainly for their weapons design missions, the labs worked with several computer manufacturers to develop new architectures for high-speed computers and the software tools for using them. This work took place at the companies and at the labs.

Each agency supported the discipline in its own way, tailored to its own needs and style of operation. And it was probably to the advantage of the discipline.

The arguments are basic ones: Diversity versus efficiency. Top-down versus bottom-up management of the scientific enterprise. Is it best to have a science czar controlling all expenditures, setting priorities, determining the nature of support? Or are we better off with a diversified system, with diffused authority and loose coordination through the science adviser's office, with occasional crosscutting programs such as HPCC?

Some of the resistance to change comes from senior people, scientists, politicians and bureaucrats whose funding or authority is based on the current system. But looking at the success of the present system, it is hard to argue for change.

Nor is there likely to be such change, despite Walker's support. Forming a new agency in the face of strong opposition within the administration and the community the agency is supposed to serve would be an enormous challenge. That is not to say, however, that there will not be some significant restructuring of some agencies, particularly if the parent department is eliminated. Chances are that the DOE labs, NIST and NOAA would not be totally dismantled; they would have to go somewhere. But a Department of Science is not likely to be their destination.