The Dual-Use Dilemma
Political pressures make it difficult to prevent defense needs from squeezing out commercial considerations.
The Clinton administration began with high hopes for its plan to forge a stronger link between military and commercial technologies. Observers inside as well as outside the administration argued that the Technology Reinvestment Project (TRP), designed to propel the development of commercial technologies critical to the military, could spur innovation throughout industry, give U.S. firms a new edge in international competition, and help ease the post-Cold War conversion of the defense industrial base to commercial production. These high expectations paved the way for rapid congressional approval of the effort. At the same time, they saddled the program with the burden of meeting objectives that it was never designed to achieve.
Approved by Congress near the end of the Bush administration (which resisted its implementation), TRP was launched by the Clinton administration in early 1993 as an experiment to improve the outcome of dual-use technology programs. (It was replaced by a new program at the beginning of fiscal 1997.) The Pentagon had begun to pursue a dual-use investment strategy in the early 1980s, funding innovative R&D for technologies with both commercial and military applications. The strategy was based on the recognition that maintaining a separate defense industrial base is expensive and unwieldy and fails to capitalize on innovation in the commercial sector, where cutting-edge technologies most often originate. In addition, policymakers were concerned that if the United States let foreign firms take the lead in the commercial development of dual-use technologies, the Pentagon could ultimately become dependent on foreign suppliers for key military components.
The first round of dual-use programs, however, produced technologies that served only military needs. The emphasis in TRP was on developing technologies that could result in viable commercial products as well as military uses. Although its life was short, TRP’s innovative programmatic features-including government-industry cost sharing and the awarding of competitive grants to industry-led teams-have earned a positive assessment. It is too soon to judge whether it has contributed in a consistently positive way to the development of viable commercial technologies.
Moreover, in satisfying the expectations of some of its proponents, the program has frustrated the hopes of others. Although TRP was created as a military-sponsored R&D program, it came to be considered the centerpiece of both the administration’s technology policy and its defense conversion efforts; a burden that is simply too heavy for this program to bear. A review of TRP illustrates the political dynamics underlying the dual-use strategy and illuminates the need to disentangle the goals of technology policy and military procurement.
TRP was designed to escape the fate suffered by earlier dual-use initiatives, such as the Very High Speed Integrated Circuits (VHSIC) project and the Strategic Computing Program (SCP). These programs were moderately successful in developing military applications for commercial technologies that were already under development, but they did so by gravitating toward dedicated military production, even prompting manufacturers to build entirely separate production lines for the military versions of the technology. The result was to create separate trajectories for the development of commercial and military technologies.
To avoid creating new defense-dedicated industrial niches, TRP’s designers set out to fund projects that would advance the general technological state of the art. Where a technology development trajectory had already been established commercially and confirmed by a pattern of private-sector investment, TRP projects were supposed to accept the market-driven trajectory, not reshape the line of technological development to achieve a particular military objective.
TRP sought to harness the power of the commercial marketplace for defense needs without distorting market forces. But it was designed to serve the military, not the commercial sector. Although the program was intended to accelerate the development of commercial technologies, its purpose in doing so was to make them available to the military more quickly and cheaply, not necessarily to improve the competitive position or performance of commercial firms.
TRP’s management structure reflected its priorities. Competitions for TRP grants were conducted by the Defense Technology Coordinating Council, made up of nonmilitary people from the National Science Foundation, the Departments of Energy and Transportation, the National Aeronautics and Space Administration (NASA), and the Commerce Department’s National Institute of Standards and Technology. The diversity of the council’s membership was meant to enhance the program’s technical expertise and to ensure that TRP managers were cognizant of recent technological developments in commercial markets and civilian federal agencies. Management and control of the program, however, was assigned to the Pentagon’s Advanced Research Projects Agency (ARPA). Putting TRP under ARPA created a fundamental political imperative: TRP projects had to be justified primarily on the basis of their anticipated value for national defense. This mission was much narrower than the goals of many of TRP’s early backers.
TRP was plagued from the start by the demands of disparate political factions with sometimes contradictory goals. The coalition that mobilized to support the program combined elements of U.S. business bruised by foreign competition, members of the defense research establishment concerned about growing dependence on foreign sources for military technology, and neoliberal Democrats who believed that traditional Democratic economic policies wrongly emphasized consumption and distribution over investment and growth. In addition, the labor and peace movements focused on TRP as a way to stem the loss of domestic manufacturing jobs and win a share of the evanescent “peace dividend” that the United States was supposed to reap after the end of the Cold War. Finally, various agency officials and White House staff who were charged with designing, publicizing, and implementing the program also wanted to demonstrate that a federal dual-use R&D program could actually work.
To cement support for the program, the Clinton administration packaged it as the centerpiece of the national defense conversion effort. The young administration was still reeling from the unexpected February 1993 defeat of President Clinton’s modest economic stimulus package. TRP was intended to be part of the administration’s strategy for addressing long-term structural problems in the economy, whereas most of the administration’s other conversion initiatives emphasized short-term economic adjustment assistance of the type included in the failed stimulus package.
In the chaos of the Clinton administration’s early months, however, the message was seriously muddled. Most Americans who were aware of TRP were convinced that it was primarily aimed at solving the short-term adjustment problems created by the end of the Cold War. Although officials directly responsible for the design and implementation of TRP never saw it as a defense conversion or jobs program, administration officials reinforced this impression in speeches and public documents describing the program.
In a White House press release dated April 12, 1993, President Clinton called TRP “a key component of my conversion plan . . . [that] will play a vital role in helping defense companies adjust and compete.” He added: “I’ve given it another name-Operation Restore Jobs-to signify its ultimate mission, to expand employment opportunities and enhance demonstrably our nation’s competitiveness.”
Ironically, the broad support and high hopes that TRP generated became the source of its greatest political vulnerability. Despite the program’s genuine accomplishments, various constituencies became disillusioned with its failure to live up to the administration’s rhetoric. Moreover, the very features that have protected the program against replicating past failures have prevented it from attracting stronger political support.
Learning from experience
Let’s look briefly at how TRP was supposed to work.
Closing the pork barrel. To ensure that TRP awards would provide government support without dampening market signals, TRP’s designers built in several valuable programmatic features:
- TRP invited industry input in designing its research agenda. This ensured that TRP awards would target militarily relevant technologies that were also likely to be commercially viable. Investments were made in an array of fields to ensure that the champions of any particular technology or industry did not exercise undue influence.
- Grant applicants were required to compete in teams combining defense and commercial companies or universities and national labs. The competitions were judged according to technical and economic criteria by a panel of government experts or other independent peer reviewers. Applicants were also required to provide evidence that the technology could be commercially sustained within five years without further federal funding.
- To ensure that program participants did not become lazy due to an overreliance on the new government subsidy, TRP required grantees to risk their own money as well as the taxpayers’: They had to cover at least 50 percent of the project’s costs. (This commitment was also designed to encourage them to abandon technological approaches if it became clear that they were not working.) Over time, each TRP dollar was matched by an average of $1.33 of nonfederal funds.
- To prevent the creation of pork-barrel projects that the government could not shed, TRP’s managers made only time-limited grants. Technical milestones and other benchmarks were established up front.
Promoting spin-offs. TRP’s managers looked for technology development projects where military performance requirements will clearly complement commercial market requirements. In other words, the same technology could be used in different products that would meet military and civilian needs.
For example, TRP funded a project to develop a turbo alternator for electric hybrid vehicles. The turbo alternator would serve a military need for tank engines that will be harder to detect with infrared sensors, because the hybrid engines emit less heat. It would simultaneously serve a clear commercial need for small, energy-efficient, low-emission engines that can be used in hybrid (diesel/electric) city buses.
Promoting spin-ons. The purpose of TRP spin-on development projects was to provide the military with leading-edge technology that would be expected to become affordable over time as the evolution of the technology led to the creation of a self-sustaining commercial industry. By adapting a new technology for military use, the project could provide an important market, contributing to the industry’s growth and the rapid advancement of the technology. The trick was to do this without disrupting the market signals that would guide the trajectory of the technology’s commercial development. This was thought to be possible only if the military-sponsored project generates equipment and tools that in turn would expand the commercial market.
For example, one TRP project was aimed at enabling military surgeons to learn surgical procedures on a computer simulator. The project will spin on to military use technologies already developed for commercial purposes. The goal is to ensure that commercial producers develop these systems further with military needs in mind; for instance, treating injuries such as shrapnel wounds that are found frequently on the battlefield but rarely in civilian life. The software and hardware packages that will be developed for these specific military purposes can then be adapted for use in civilian medical training, disaster response, emergency room medicine, and commercial telemedicine. Indeed, the project is expected to facilitate the commercial development of many minimally invasive surgical techniques for which there is now a growing demand among civilian medical professionals.
Civilian-military integration. TRP’s most important long-term goal was to promote commercial-military integration-the creation of a single unified industrial base for commercial and military technology development. Most TRP officials believed that this would be a more viable strategy than direct defense conversion. They argued that even if traditional defense contractors converted to new products, they would be unlikely to compete successfully in commercial markets, given their high-overhead operations and cost-plus-contract culture. Instead, they stressed the need to help specialized defense companies find commercial partners to teach them what they need to learn about marketing and high-quality low-cost manufacturing and thus overcome the barriers that separate the defense industrial base from the rest of the economy.
An early assessment
After less than four years, TRP made significant progress toward meeting its major goals. It fostered industry-led dual-use R&D in a number of fields, and it encouraged defense and commercial firms to work together on the commercial development of a number of militarily relevant technologies. The features that TRP incorporated to avoid government failure proved to even more successful. It is too early to judge the effect of time-limited grants, but other traits-government-industry cost sharing, competitive selection, and the requirement that applicants be made up of industry-led teams-combined with the vigilance of TRP’s directors to make the program nearly free of political pork.
To judge whether TRP will actually prove to have accelerated the development of commercial technology, one will have to examine the results of the program project by project after several more years. However, it appears that by playing midwife to research consortia, TRP has likely facilitated more rapid technology transfer and innovation. With TRP funds, for example, the defense contractor Aerojet partnered with commercial companies such as General Motors, Admiral, and Boeing that were potential users of a class of new materials called aerogels, which have superb heat-insulating properties.
Many TRP projects also boosted or accelerated the efforts of organizations that were already working together. For example, TRP made awards to California’s CALSTART electric vehicle consortium, originally funded by the state government as well as corporate funders, including Lockheed, Allied Signal, and Hughes. In addition, several teams that were created specifically to apply for TRP grants but did not win any TRP money in the first round of competition recognized the potential leverage that they had acquired as a group and decided to stick together and competed successfully for TRP funds in the second and third rounds.
By encouraging this kind of teamwork, TRP also contributed to the cause of defense conversion even though conversion was never a specific objective of the program. It sponsored technology development projects that sought new civilian uses for military technology, as well as technology deployment projects that attempted to modernize engineering education and promote the diffusion of best-practice manufacturing throughout the economy. One example was the somewhat mystically named “realization consortium,” an effort by several leading U.S. engineering schools (including MIT, Cornell, and Tuskegee) to revolutionize undergraduate engineering education by forming a virtual design studio of the future over the Internet. Finally, TRP helped to promote economic adjustment in defense-dependent regions by supporting teams of dual-use manufacturers who were inevitably and disproportionately clustered in heavily defense-dependent areas of the country.
Despite its successes, TRP clearly suffered from some of the dilemmas that had plagued other recent dual-use initiatives. Because the program received all of its money and most of its technical expertise from the Pentagon, the panels that awarded TRP grants were compelled to justify the projects first and foremost in terms of their military value. This raised a number of key issues that, if unresolved, will continue to keep dual-use R&D programs from reaching their full potential.
Defending the program. It was always easiest to justify the value of TRP’s investments to the Department of Defense (DOD) on the grounds that they met the “but for” test-that is, but for federal funding, industry would not have undertaken these investments on its own. But this rationale put pressure on the program to invest in projects that emphasized military-specific attributes of emerging technologies that commercial manufacturers would not otherwise have developed-the same trap that VHSIC and SCP fell into.
Moreover, justifying broader commercial investments on this basis left the program vulnerable to the strongest argument of the dual-use strategy’s detractors: that the R&D in question would be conducted by private industry regardless of TRP’s involvement. Thus, critics have argued, TRP was merely substituting public funds for private; it was not increasing the total amount of militarily useful research done by commercial firms.
It is almost impossible to defend against this argument, for one cannot prove that a company would not have made the same investment in the absence of the TRP subsidy. In fact, many recipients of TRP awards volunteer that the TRP grant simply accelerated an investment they were planning to make anyway.
A stronger justification, then, may be that the government’s investment speeds the development of new technology, enabling U.S. systems to incorporate it sooner and at lower cost. By establishing an early, dominant position in markets for an advanced technology, this argument goes, U.S. companies can lock in control of a long-term stream of follow-on product and process innovations, making market entry much harder for companies in other countries. Thus a temporary market advantage can turn into a more enduring technological and perhaps economic and military advantage.
In some cases, however, it might actually be to the economic advantage of U.S. manufacturers to be second rather than first-to reap the windfall of investments that foreign governments and companies have made and to start production further along the technological learning curve. According to this more conventional line of economic reasoning, an economy that is prepared to absorb and capitalize on innovation, whatever the source, will be better able to establish a dominant position internationally. This argument implies that DOD should have encouraged TRP to make more grants for technology deployment-upgrading contractors’ manufacturing capabilities so that they could adapt innovations more easily-and fewer for technology development.
Either justification, whether stressing the need to accelerate innovation or improve our ability to absorb innovation, have provided a politically more defensible rationale for the program.
Supply-side or demand-side? TRP took a supply-side approach to the spin-off of new technologies: It facilitated cooperation among companies in order to help them take advantage of technological and economic opportunities. This may have been sufficient in cases where the commercial market for a new application was already well established. In the past, however, demand-side strategies have achieved the greatest success in speeding the commercialization of military technologies. The classic example is the spin-off of integrated circuits, a process that DOD and NASA launched by agreeing to provide a guaranteed market for the new technology at premium prices.
The political justification for this type of strategy is that the social benefit (in this case, a national security benefit) justifies the cost. In fact, many TRP spin-off projects promote social goals that the Clinton administration views favorably: energy efficiency, job retention, improved public health, environmental remediation, and pollution prevention. A demand-side strategy could still give a real boost to many TRP technologies. But politically it is much easier to justify demand-side intervention on the grounds of national defense than to mobilize support for a civilian mission. Thus, for TRP-backed technologies, there has been no publicly subsidized civilian demand analogous to that provided by DOD, which is committed to purchase TRP-developed technologies that are expected to strengthen the nation’s defenses. DOD will surely buy military versions of hybrid electric vehicles, but there is no plan to require or even encourage the procurement of civilian versions by federal, state, or local government agencies.
The Clinton administration has never attempted to construct a full-blown demand-side strategy in which nonmilitary government agencies-say, the Department of Transportation or the Department of Energy-guarantee procurement of emerging dual-use technologies at premium prices in order to promote a clearly articulated social goal. To be sure, any attempt to do this would run counter to the administration’s procurement-reform efforts, which are aimed at removing government-mandated restrictions and inducements and replacing them with market signals. But this may be a case in which policies that harness the power of government to create a healthy market should take precedence over policies aimed at unleashing the power of the marketplace.
Pressure to protect investments. Finally, it is not clear whether TRP and its successors will be able to avoid becoming trapped by their own technological choices as a result of political pressures from the defense establishment. Such pressures have already threatened to channel TRP projects into emphasizing military-specific rather than dual-use technologies.
Learning from past experience, TRP’s managers consciously attempted to fund multiple technological approaches to meeting critical military needs. For example, TRP funded two approaches for developing rechargeable lithium ion batteries (which would be widely deployed on the battlefield as well as in commercial portable phones and laptop computers); three approaches for developing uncooled infrared sensors (for military and police and firefighter night vision systems); and three approaches for developing low-cost manufacturing processes for advanced display technologies (for a host of defense and civilian products). The rationale for this strategy was that making a commitment to a single approach in the first round of TRP competitions might give it an unfair advantage in subsequent rounds. After all, TRP’s managers were investing taxpayer dollars in dynamic areas of technology that could follow any one of a number of paths of evolution; they understood that it was critical that they have the flexibility to transfer money into the paths that become more promising as they evolve.
Because the money is coming from the defense budget, however, the awards made have to be justified to Congress in terms of their supposedly unique value for national defense. As a consequence, DOD planners are likely to find that it is politically difficult to abandon the paths that they establish in the initial funding rounds. In fact, many DARPA managers were wary of becoming involved in the high-profile TRP because they feared–correctly in my opinion–that it would subject them to unwelcome congressional scrutiny and threaten their flexibility and independence.
The Clinton administration’s single most controversial effort to develop dual-use technology-the National Flat Panel Display Initiative-illustrates this problem. Although it was not part of TRP, some of its development projects were being run through TRP and its rationale was fundamentally the same as TRP’s. The goal of the flat-panel initiative is to generate a globally competitive U.S. industry in a sector already dominated by a globally competitive Japanese industry.
The initiative makes a portion of DOD’s R&D investment in future display technologies available to U.S. companies that commit to producing current-generation products domestically in high volumes and to meeting DOD’s specialized display requirements. However, because the effort initially targets military-specific markets, the TRP-backed flat panel projects emphasize attributes that have no obvious commercial value. The displays must be able to operate in desert and Arctic temperatures; be readable in sunlight as well as in night combat; have extremely high resolution; integrate specialized information-processing capabilities; and come in nonstandard sizes. In addition, each TRP-backed flat panel project is specifically designed so that the demonstration phase of new displays is done using military hardware produced for different branches of the military, in order to ensure broad military support for the technology.
The danger of this approach is that U.S. suppliers could end up with flat panel displays that are of interest only to military users. If this happens, civilian firms will continue to buy their displays from foreign producers, and the initiative will have created exactly what it was designed to avoid: a specialized, government-subsidized arsenal for military flat panel displays.
Designers of the initiative contend that it should be viewed as an insurance policy that can be canceled before the premiums get too high. But if the three or four production plants that they envision can provide the military with gadgets that cannot be obtained anywhere else, the Pentagon’s temptation to protect the investment will be strong, particularly because the investments have been justified to Congress as uniquely necessary to promote the nation’s defenses.
The political pressures to protect investments justified on the basis of military need are relentless, and it is not clear that TRPand its successors TRP will be able to resist them in every case. So far, the bulk of TRP’s projects appear to be adhering to their dual-use objective. In the long run, however, all the Pentagon’s dual-use projects would be better able to resist the pressure to emphasize military applications if they could make a strong public policy case for the nondefense applications of the technologies they sponsor.
In addition, the success of TRP’s projects will depend ultimately on the extent to which DOD aggressively implements federal acquisition reform. Spurred by Secretary William Perry, the Department of Defense recently adopted new internal policies intended to reduce the use of military specifications in procurement. In addition, the Clinton administration worked successfully with Congress to pass the far-reaching Federal Acquisition Streamlining Act of 1994. If the Pentagon bureaucracy nevertheless continues to prevent military customers from buying commercially created technologies off the shelf, commercial and military applications of dual-use technology will continue to follow irreparably divergent paths of development.
Policy, politics, and priorities
Pressure on TRP’s program managers to place even more emphasis on military needs has increased significantly since the congressional elections of 1994. Even before the elections, Congress mandated that fiscal 1995 TRP funds could not be obligated until the secretary of defense ensured that representatives of the military services were full members of the Defense Technology Coordinating Council. The new Republican Congress added further restrictions aimed at ensuring military relevance: The undersecretary of defense for acquisition and technology must now certify to Congress that representatives of the military services constitute a majority of the membership on TRP project-selection panels. And before obligating any funds to a new project, Congress must now receive a report describing its military objective.
In response, ARPA has made a number of changes. TRP was discontinued at the end of fiscal 1996 and replaced with a similar dual-use technology development program more clearly dwesigned to fulfill military needs. The new program will not solicit technology deployment projects or regional alliance or manufacturing education and training projects, because these efforts are considered less relevant to defense. Technology development efforts driven primarily by competitiveness concerns will be left to the Commerce Department’s Advanced Technology Program. ARPA officials have been paired with representatives of the military services to jointly manage previously funded TRP projects. In the end, TRP projects-even if they succeed in generating commercial as well as military applications-will be judged almost entirely on their value for national defense.
The Clinton administration never successfully cemented a political coalition in support of TRP that could counter these pressures from the defense establishment. Because TRP focused mainly on technology development, it was a poor vehicle for quickly replacing lost defense jobs, a fact that all but eliminated support for it among labor and peace groups. At the same time, business support for the program remained weak. The requirement that companies compete for awards in teams limited its appeal among defense contractors, who did not necessarily welcome the new competition from commercially oriented firms. Cost-sharing requirements, meanwhile, limited the size of the program’s cash grants to individual companies. The relatively small size of the awards-they ranged from $130,000 to $80 million and averaged about $5.8 million-have made the much larger and easier-to-obtain tax incentives that the Republicans now offer as an alternative for promoting technology development more attractive to many companies than TRP’s matching grants, which were awarded competitively and remain subject to regular, rigorous review. In addition, congressionally mandated restrictions on foreign participation in TRP continue to buck the market trend toward international strategic alliances for high-risk technology development.
Finally, the program’s success in making sure that its projects were market-driven has reduced opportunities for congressional earmarking. This in turn has limited the ability of TRP’s proponents to cultivate stable political support for continuing the program. Compare the political fortunes of TRP with that of two other Clinton-era technology initiatives: the National Flat Panel Display Initiative, which has survived congressional scrutiny downsized but intact because of the efforts of a specific, and therefore readily organized, industrial constituency; and the Commerce Department’s Manufacturing Extension Partnership, which has built broad support by placing more than 40 manufacturing extension centers in more than 30 states and which the Republican Congress has actually chosen to expand.
TRP has retained a good deal of support among “competitiveness” advocates, but this issue has lost political currency since the program was launched. In the mid-1990s, as changing economic conditions at home and abroad conspired to strengthen the U.S. position in global markets, the focus of the average American’s economic anxieties shifted away from declining international competitiveness and toward the continued stagnation and unequal distribution of U.S. wages and family incomes. Advocates of competitiveness policies remain convinced that efforts to help companies raise their long-term productivity are essential to addressing both sets of concerns. The Clinton administration’s technology policies, however, have often appeared to reward only those groups that are already benefiting from the new high-tech economy, and so have failed to gain support from a larger potential constituency.
To the extent that Americans are wary of spending taxpayers’ money to boost the profits of private firms, they are likely to remain ambivalent about government’s proper role in technology development. Programs such as TRP will attract support from a broader coalition of political forces only if they are embedded in a broader set of public purposes, such as widely perceived threats to health, safety, or the environment, and linked to a set of more rapidly demonstrable results. In the absence of broader public support for federal investments in technology and science, the utility of programs such as TRP for accelerating the commercialization of innovations, whether for defense or “competitiveness” objectives, is limited.
In the brief period since its inception, TRP proved to be a useful and innovative tool for advancing the development of commercial technologies critical to defense, encouraging the more rapid adoption of commercial technologies by the military, and promoting the integration of defense and commercial production within firms. But for believers who thought that DOD, or at least ARPA, would altruistically promote competitiveness or the technical competence of U.S. industry, the lesson of the past three years is clear. Commercial technology spin-offs or the creation of viable commercial industries can never be anything more than byproducts of a defense-sponsored technology program, because its ultimate goals are not economic or scientific but political. The capacity of a program such as TRP to promote defense conversion, or commercial-military integration, or national industrial performance will remain hostage to the program’s overriding military goals.