For nearly fifty years, U.S. military superiority has depended on American technological preeminence financed through public spending by the national security establishment. At the peak of the Cold War during the 1980s, the Pentagon, the Department of Energy, NASA, and the intelligence agencies together spent more than $60 billion a year on defense-oriented research and development-- nearly two-thirds of all government-funded R&D outlays during the Reagan years.
This public investment didn't just build weapons. It created a technical archipelago of immense value, whose crown jewels are the three national weapons laboratories at Los Alamos and Sandia in New Mexico and Lawrence Livermore in California. With the Cold War over, the national labs will either be gradually defunded as relics or reborn as civilian technology centers of development.
Today America's commercial technology base is eroding in the face of stiff, government-assisted competition from foreign firms. This erosion does not stem solely or even mostly from exchange-rate imbalances, low rates of domestic savings, wage-rate differentials, or the process of industrial followers inevitably "catching up." Most of America's competitive problems are structural. Primary among them is the country's overreliance on defense-driven technology a strategy sufficient to win the Cold War but not to prevail in the commercial arena.
To meet this competition, the U.S. needs to formulate an explicit, civilian-oriented technology policy. In one sense, the U.S. is lucky that it already has the national laboratories in place. If they didn't already exist, the United States would have to create them. Yet the traditional military rationale for the labs, as a special exception to a general dogma of laissez-faire, makes it difficult to broaden their mission.
Under a new policy committed to the support of commercial technology the weapons labs would be one instrument among many in a national competitiveness strategy (see "Investing on the Frontier," by Michael Borrus, p. 79). They would facili- tate innovations American companies would not make if left to draw on their own resources. Adding public to private resources, they would encourage investment in technologies that reflect a new definition of national security embracing environmental as well as defense concerns.
Beyond Test Tubes
Ever since J. Robert Oppenheimer led the Manhattan Project at Los Alamos, beginning in 1943, the national labs have benefited from enormous investments in human and physical capital. Today they employ approximately 27,000 scientists and technicians. They have a capacity, rarely matched in privately owned industrial laboratories, to undertake costly basic and applied research aimed at long-term payoffs. Although the weapons labs are owned and financed by the U.S. Department of Energy, they are operated by contractors. Los Alamos and Livermore are run by the University of California and Sandia (until October 1993) by
The word "laboratory" conjures up images of white-frocked scientists mixing chemical concoctions in test tubes and conducting arcane basic research. In fact, the national weapons labs are highly applied, "mission-oriented" organizations. Building a nuclear weapon is, after all, a manufacturing enterprise that requires concrete problem-solving in theoretical and applied nuclear physics, computer simulation, metallurgy, precision fabrication and testing, as well as actual production. At the weapons labs, interdisciplinary teams of scientists and engineers solve design and manufacturing problems as concrete as those at any consumer products company. The labs also deal with safety-related technical problems involving weapons manufacturing, transportation, and storage.
Though the labs have been used primarily to design and develop weapons, their use for nondefense purposes has fluctuated over the years. At the high point of the perceived energy crisis of the 1970s, the laboratories devoted nearly 50 percent of their budgets to energy research. During the Reagan years, the research priority shifted back toward weaponry. Today about 25 percent of the laboratories' budgets finance nonweapons research emphasizing energy and the environment. The labs are housed in the Department of Energy because it was the successor to the old Atomic Energy Commission, which was in charge of nuclear weapons and later nuclear energy under Eisenhower's "Atoms for Peace" program. But thanks to their practical experience with a variety of advanced technologies, the labs have also made significant contributions in fields such as diagnosis and treatment of cancer and AIDS, development of new structural materials, hardware and software advances for supercomputers, and techniques for sequencing the human genome (see "Chips With Everything," p. 94).
As the labs' primary mission nuclear weapons recedes, at issue is just how they should contribute to basic and applied commercial research. In government policy circles, three main approaches are being debated. The first, espoused by many in the Departments of Defense and Energy, asserts that the labs should continue to focus exclusively on their defense missions. A second approach concurs that the labs should focus on defense, but argues that the defense technology base itself requires the labs to promote more civilian technology. A third, more radical, proposal would convert at least one of the three laboratories, presumably Lawrence Livermore, into a research center for civilian technology.
These are not the only possibilities. Other options include the wholesale dedi cation of separate (and perhaps new) national laboratories to the solution of particular problems, such as a National Environmental Protection Laboratory or a National Energy-Efficient Transportation Laboratory. We sympathetize with these aims, but the history of the national weapons labs suggests the particular value of interdisciplinary institutions with multiple missions. The weapons labs have shown a capacity for flexibility and technological serendipity that would be lost if they were converted to single-purposes. The challenge is to broaden their franchise so that the labs can anchor a renewed national commitment to high-tech research and production generally.
Spinning a New American Technology Policy
There is nothing new under the rising sun. Well before Japan devised its famous industrial policy, the United States had industrial and technology policies for its emerging sectors. These policies were largely dictated, however, by the nation's military needs. In fact, most of America's high-technology success stories including aircraft, semiconductors, telecommunications, and computers trace their origins to government efforts to promote technologies for national security. Sometimes, the federal government even fostered civilian markets for essential military products for example, providing postal routes for aircraft during the 1920s so that defense contractors would not disappear or lose their technological edge for want of markets during peacetime.
The commercial benefits of America's defense-driven technology development strategy were pervasive if unintended. Pentagon research funding, development projects, and procurement expenditures provided four crucial functions not found sufficiently in private industry: patient capital; a mechanism for collaborative research; high-volume "launch markets" for new products and production processes such as integrated circuits and computer-controlled metal cutting tools; and rapid diffusion of the latest technologies.
The United States has spent billions of dollars on agricultural extension, but almost nothing on industrial extension. However, military technology has been the closest American equivalent to a technology-diffusion policy. Military requirements for domestic second sources on major weapons contracts, along with the extensive use of subcontracting, ensured rapid and widespread diffusion of new production equipment and materials, fostering further rounds of manufacturing and design innovation. All of these led to essentially unintended "spin-offs" that had commercial uses.
But lately, as military technologies have become more rarified, the possibility of easy spin-off has diminished. Military-oriented companies have come to exist in their own distinct subculture. Over time, narrow military performance requirements, along with byzantine procurement regulations and national security restrictions on exports, have isolated the defense sector of high-tech industry in its own ghetto.
Paradoxically, as the underlying technical base of military and civilian products have converged (mostly around microelectronics), the very flexibility of the underlying technologies made it possible for the products themselves to diverge more than ever before, driven by the very different needs of military and civilian consumers. The military can now optimize semiconductor designs, for example, for higher speed processing, whereas many commercial chip producers want to emphasize low power consumption (often a necessary trade-off). A machine tool for the Air Force can be constructed with a built-in capacity to cut along five-axes in unbounded space, whereas a commercial cutting tool just needs to be able to carve a curved or straight line along a stencil. And defense satellites can be built and programmed to confuse enemy spy equipment, whereas commercial satellites are typically made to travel in predictable geosynchronous orbits.
Recent technological advances have enabled commercial producers to match and surpass military producers in product performance while maintaining their traditional superiority with regard to price. For example, while the Pentagon's Very High Speed Integrated Circuit (or VHSIC) program ultimately produced new high-speed chips for insertion into the cruise missile, Intel Corporation, which refused an offer to participate in VHSIC, met the same technical requirements commercially more than a year earlier, and at much lower cost.
In the U.S. military technology establishment today, there is a debate that goes under the heading "spin-on versus spin-off." If spin-off means commercial applications of defense-oriented technology, spin-on suggests technologies de signed in the commercial sector that can have military uses. Of late, as advanced technologies have proliferated, most Pentagon planners have concluded that advanced technology is simply too diffuse and dynamic for the military establishment to be the sponsor of every technology that might have military uses. Hence, even narrowly military needs require a broad commercial high-tech industry.
For the foreseeable future, America's industrial technology base will be shaped by three trends: declining military
Today the same functional elements of government technology policy that once worked so well for the Pentagon
A New Mission
In light of the shifting relationship between military and commercial technology, what course for the national labs makes the most sense? Consider the three proposals that have emerged thus far.
Status Quo. The most conservative proposal, propounded in a January 1992 report of the Secretary of Energy Advisory Board (SEAB) Task Force on the Department of Energy National Laboratories, counsels adherence to the traditional defense missions of the laboratories, even while it argues that the world situation has changed radically. The task force sidesteps this contradiction, contending that "a strategic vision of the future missions of the National Laboratories cannot be developed until the Department and the Nation have developed a similarly clear vision of their own role in the future multipolar world." In fact, the report denies any role for the labs as part of a larger national economic infrastructure. The report minimizes the labs' role in technology transfer, declaring that it and "other activities" such as math and science education should be undertaken only insofar as they do not "detract from the overall Laboratory focus on their core missions." In effect, the task force's recommendation regarding diffusion of technology amounts to an endorsement of a "trickle down" theory of technology transfer that will transfer very little of practical use to industry.
Bifurcation. A second proposal has been advanced by Congressman George Brown, Jr., chairman of the House Science, Space, and Technology Committee. Responding to the SEAB report, Brown proposes that Lawrence Livermore be transformed into a wholly nondefense national laboratory, charged with the advancement of critical commercial and environmental technologies. In Brown's design, Sandia would become a technology transfer "center of excellence" and the main provider of arms-control verification technologies. A significantly downsized nuclear weapons
Brown's proposal reflects an understanding of the economic security challenges facing this nation and a seriousness of purpose that is either wholly absent or obscured in the SEAB proposal. The proposal could, however, create inefficiencies and greater costs in the long term. In separating their defense R&D from civilian R&D efforts, the labs may fail to take advantage of the increasing similarities between the generic technologies of both sectors. For example, the microelectronics and advanced materials technologies used in defense technology are not significantly different from those in civilian production.
To be sure, defense products typically cost more, take longer to develop, and lag behind their civilian counterparts in both quality and performance. But such differences are not inherent in the technologies; the example of very high speed integrated circuits demonstrates how the organizational separation of technology development projects can interrupt the flow of information between research groups that gear their applications to the diverse needs of different customers. Rigidly separating defense and commercial labs would limit possible applications and spillovers. Congressman Brown's proposal, however well intentioned, would isolate defense from commercial development. But if the nation is to continue to integrate the most advanced technology into its defense systems at an acceptable price and pace, the flow of civilian technology into those systems should not be artificially constrained.
Partnership. The third and most visionary of the proposals to emerge thus far is a bill titled the "Department of Energy Laboratory Technology Partnership Act of 1992" introduced by Senator J. Bennett Johnston, the chairman of the Senate Energy and Natural Resources Committee. This bill would establish partnerships involving the DOE national laboratories and industry, universities, and other federal agencies for the purposes of research, development, and application of technologies critical to national security and technological competitiveness.
Recognizing the gravity and breadth of nondefense challenges the nation faces and the relevant technical strengths the labs possess, the bill is a refreshing departure from most standard congressional attempts to come to grips with the post-Cold War world. The bill casts a wide net, specifically identifying the need for partnerships in areas such as energy efficiency, energy supply, high-performance computing, the environment, human health, advanced manufacturing technologies, and transportation technologies.
The bill is designed to strengthen the 1989 National Competitiveness Technology Transfer Act, which, for the first time, permitted the national weapons labora tories to enter into cooperative
Still, there are a number of problems with the bill. The first and most critical shortcoming is that the partnership activities it promotes are meant merely to supplement traditional missions of the DOE and its labs in defense and basic energy research. These activities are likely to become subservient to "missions" of the labs and will not likely command priority funding and staffing.
Moreover, while the bill's emphasis on "dual use" implicitly recognizes the convergence of generic technologies in both defense and commercial industry, it overlooks the fact that commercial technology often outpaces military technology. The Pentagon's leading dual-use programs in the 1980s in very high speed integrated circuits and strategic computing led to few significant civilian applications.
The problem is that the funding for dual-use programs that make military applications paramount passes through (and is managed by) a Pentagon or armed services office. Despite the best dual-use intentions, the technical defense requirements imposed by these offices almost always aim the technology's evolutionary trajectory toward uniquely military applications. Thus a dual-use strategy dictated primarily by military needs ignores the fact that the commercial sector can increasingly accommodate military needs for products and processes something that was not usually true in the past.
A final problem is the risk that this approach could lead the labs to be overly obedient to the short-term needs of private industry. Taken to an extreme, this approach risks turning the labs into short-term "job shops"; in short order, the resources and talents resident in the laboratories will be chewed up, burned out, and effectively dispersed.
Thus the three general proposals suffer from fundamental shortcomings, as does the resulting state of the debate. Instead, the labs should be retained as broad, multipurpose organizations, with new resources and a broader charter to work both on "market-driven" projects with industry and on basic and applied research that serves the national interest.
Labs as Economic Infrastructure
The missions of the labs must be recast to meet the post-Cold War imperatives of national and global well-being. This is not to suggest that the national weapons labs abandon defense R&D. Rather, we propose that they radically shift their efforts toward meeting nondefense needs in such areas as energy, the environment, and health care, as well as economic competitiveness essentially reversing the current 75 percent-25 percent ratio between defense and nondefense R&D efforts.
Under this conception, the labs would:
- pursue a civilian-driven R&D mission that is not merely appended to the present weapons mission but that entirely supplants the present mission in terms of importance;
- continue to serve the nation, not U.S. industry, as their primary customer;
- focus their R&D on a broad range of generic applications, with no "dual-use" bias imposed upon it;
- redirect their mission as part of a larger, comprehensive post-Cold War national science and technology strategy.
Notwithstanding the stale industrial policy debate of a decade ago, government has always played an active role in the U.S. economy not in "picking winners" but in helping to incubate technologies. To compete in global markets, American firms need a place to work together to develop and diffuse generic, "enabling" technologies that are expected to produce significant spillovers (in terms of both product applications and manufacturing know-how) to a broad swath of industrial sectors. Today no individual firm has in place the same multiprogram, interdisciplinary strengths already present at the national labs, nor can any individual firm afford the investment in both time and money that would be necessary to create, from scratch, a similar range and breadth of expertise.
The national labs are not the only available instruments to develop enabling technologies, but as industry cuts back on its own long-term
The national labs can help restore this engine of development and diffusion that the U.S. lost with the commercialization of Bell Labs (and which the Ministry for International Trade and Industry and NTT Labs still provide in Japan). Because the labs have no commercial interests of their own, they are a neutral cooperative arena for companies many of them sometime competitors to work together. At the same time, unlike other possible government arbiters, the labs are technologically competent in their own right.
Though still minuscule compared with their defense programs, successful partnerships at the weapons labs are under way in areas such as oil recovery, superconductivity, and specialty metals. These partnerships have four characteristics that define a useful model: they share costs, where government supports the laboratories and industry supports the research to which it contributes; their efforts are focused on areas of technological opportunity; they are directed substantially at the local level, with decisions made by the partners, not government bureaucrats; and they are industry-driven to ensure market relevance but not industry-dominated. This approach, which the partnership bill emphasizes, should be refined and enlarged.
Since the passage of the 1989 technology transfer legislation, the federal government has increased funding for cost-sharing projects at the labs. Energy Secretary James Watkins has requested $117 million for such efforts for 1993, up from $69 million in 1992. Yet, these efforts remain underfunded, representing a small fraction of the budgets at the three weapons laboratories. For example, out of the approximately $250 million in nondefense R&D conducted in 1991 at Los Alamos (including basic science), only about $20 million, or 2 percent of the lab's total budget, went to support cooperative, nondefense R&D with U.S. industry.
There is still room for improvement, of course. Since the 1989 legislation, however, industry-laboratory interactions have been more constructive and productive and the level of trust and confidence continues to rise on both sides. Experience has overturned the simplistic view that the national labs are "technology warehouses" containing uncommercialized commodities ripe for the taking. The evolution from technology transfer to technology collaboration an evolution accelerated by the 1989 legislation marks a major shift for the better.
The national laboratories should also be given a strengthened mission in basic and applied civilian research. Such research is a public good, indeed, critical to a new definition of national security the health, wealth, and overall well-being of American citizens. Under such a definition, research in alternative energy sources, energy-efficient transportation systems, and even mathematical algorithms, which can save millions in communication costs, becomes a top national priority. Consequently, it would make as much sense to privatize science as it would to privatize the military.
The national laboratories are the fruit of nearly five decades of public investment in state-of-the-art equipment, multiprogram research facilities, and world-class scientists and engineers. Indeed, the labs are unsurpassed in the federal sector for the quality of their research and the breadth of their achievements. The time is right to make the labs available to tackle a broader set of national purposes. For fifty years, government and industry teamed up to win the Cold War. With that battle won, and a new, peaceful, constructive competition getting under way, this is the moment to recognize the labs as the unique national asset they are.