GENERAL CONCLUSIONS: BALANCING THE COSTS AND BENEFITS OF CONTROLS
The Panel attempted to identify and evaluate alternative approaches that simultaneously attain national security goals and cause the least damage to the capacity of the research community to make its many contributions to American life. Such a task required a review of the overall advantages and disadvantages—the benefits and costs—of the U.S. technology transfer control effort. This chapter lays out the general framework and principal results of the Panel’s review.
The Panel has attempted to examine the relations between controls and three facets of the national interest: deterring advances in Soviet military strength that come about through the use of American research results; safeguarding continued progress in U.S. military and economic capabilities, which also depends in part on American research results; and protecting long-standing educational and cultural values. The Panel’s observations about these relations in each of these areas are presented, together with observations on the feasibility of controls.
Although the Panel’s mission was to investigate the effects of restrictions on scientific communication generally, it found in reaching its recommendations that the component of the American research community that requires separate consideration is the university. Restrictions on open communication have categorically different implications for universities than they do for industrial, governmental, and other components of the American research community. This is so for two primary reasons: first, universities alone integrate the research functions and degree educational programs, so that any adverse effects on research also adversely affect the quality of the next generation of scientists and engineers. Second, unlike other research institutions, universities have never established broad controls on access to ensure that sensitive proprietary or classified information is protected. Restrictions on communications thus present an unfamiliar and distinctly unwelcome challenge for university practices. Because the potential national security concerns are most likely to arise in work that is funded by the government, the Panel’s conclusions concentrate on government-supported research.
Based on a review of costs and benefits, the Panel concludes (1) that most university research should be unrestricted; (2) that in rare
cases such research should be classified for reasons of national security; and (3) that in a few specific cases, limited control measures short of outright classification may be warranted.
PREVENTING SOVIET MILITARY ADVANCES BASED ON U.S. RESEARCH
The Relation to Controls
The fundamental justification for controls is that they retard the rate of advance of Soviet military capacity by preventing Soviet access to relevant American science. Specific questions that must be evaluated in order to assess the merits of this justification are the extent to which Soviet military strength depends on U.S. technology in general, the extent to which Soviet military advances—either immediate or long-term—benefit from U.S. academic research, and the relative contribution to leakage accounted for by the different channels of scientific communication (e.g., visits by foreign scientists scientific visitors to the United States, published papers, oral presentations, and espionage).
The Panel’s Assessment
The evidence reviewed by the Panel on the overall problem of leakage from all sources suggests that a substantial and serious technology transfer problem exists. A net flow of products, processes, and ideas is continually moving from the United States and its allies to the Soviet Union through both overt and covert means. A substantial portion of this unwanted transfer has been of little consequence to U.S. security, either because the United States did not enjoy a monopoly on a particular technology or because the technology in question had little or no military application. The Panel has also found, however, that a significant portion of the transfer has been damaging to national security.
The ease of global communications and the constant expansion of overseas sales of American products have increased greatly the number of points at which U.S. science and technology can be acquired. The loss of technology through non-U.S. sources continues to be a problem, despite efforts to reduce leakage from the West by tightening COCOM restrictions. In fact, as channels of leakage from the United States are closed or constricted, third countries may become much more attractive targets for acquisition. Stemming the overall flow of technology thus presents a very difficult challenge for the United States and its allies.
The losses of concern are not restricted to transfers to the Soviet Union. Technology transfers to Third World countries will permit them to modernize their military establishments faster and more efficiently; however, knowledge about the extent of this aspect of the leakage problem is fragmentary.
Leakage and the Research Community
The Panel took special care to assess the research community’s contribution to the overall leakage problem. In its close examination of the cases that have involved a significant loss of technology that has been critical to national security, the Panel was shown no documented examples that were the direct result of open scientific communication. However, the absence of concrete evidence linking the research community with specific losses of information critical to national security does not imply a lack of Soviet intent to exploit this source. Moreover, a substantial volume of information has no doubt been transferred as a result of open scientific communication; information on basic research has contributed to the scientific base of the Soviet Union as well as to that of other nations.
Nonetheless, we are confident that fewer significant losses have occurred as a result of normal scientific communication than have occurred by other means. These other means include legal equipment purchases, outright espionage (particularly outside the United States), illegal conduct by some individuals and corporations in international trade, and secondary transfers through legal or illegal recipients abroad to the hands of U.S. adversaries.
The Panel observes that information acquired through open communication or by means of espionage activities on U.S. campuses may not often add substantially to the Soviet military capacity in the near term. The designers of Soviet military systems are conservative, and thus new scientific advances, whatever their origin, may not be readily adopted in military systems. Moreover, such information is probably best understood by Soviet researchers, and it may not flow readily to Soviet military designers because of the highly secretive and compartmentalized nature of the Soviet military R&D and procurement process.1
The Panel, therefore, concludes, based on all the above considerations, that, in comparison with other channels of technology transfer, open scientific communication involving the research community does not present a material danger from near-term military implications.
However, the present situation is dynamic and may call for future reevaluation. There has been a dramatic expansion during the past decade in the size and scope of the Soviet effort to obtain scientific and technological information, and this effort is much better organized and more carefully targeted today. The presence of Soviet intelligence agents in the United States—both on campus and elsewhere—is substantial. Also, some university scientists may continue to extend their research beyond basic scientific investigation into applications of technology with military relevance. This raises the distinct
possibility that the university campus will come to be viewed as a much better target of opportunity for the illegal acquisition of technology. There could be a convergence between the Soviet Union’s enhanced acquisition capabilities and the possibilities of significant loss of technology through university campuses if certain types of research activities expand there in the future. An example of such research is found in some areas of microelectronics.
These areas of research are important because they develop know-how—practical techniques for solving technical problems—whose transfer to adversary nations could provide them with near-term military gains. Although published basic scientific results rarely have direct military application, technical know-how can be of substantial value in military design and production. In the Panel’s view, it is at times as important to safeguard technical know-how in areas of rapid advance as it is to protect military systems themselves.
Special concern about the loss of know-how is justified in areas where process development, or “recipe” specification, is part of a research project. In these cases, research involves the development of a series of practical steps that makes it possible to use scientific principles to manufacture a product to satisfy production requirements. The recipe is therefore the critical know-how that must be protected, because the scientific concepts have little practical use without the manufacturing recipe.
Unlike the results of scientific research, know-how is rarely communicated effectively in written form or in brief conversations. The acquisition of know-how requires a long-term, hands-on working acquaintance with a scientific or technical area. Know-how is transmitted through on-campus apprenticeships—e.g., a visitor’s continuous relations with a mentor and a research team in a research project—or through industrial training agreements. The importance of know-how thus raises special concern about the role of foreign visitors, including those in scientific exchange programs, who spend time working directly on high-technology projects likely to have near-term military applications.
Research that involves know-how is becoming more common in university laboratories for two reasons: (1) the development of equipment and processes for the manufacture of various items is often only an extension of the equipment and processes developed to conduct the basic research, and (2) as funding patterns and institutional relationships have changed, universities have sought more industrial support and have moved into areas closer to engineering design and product development than was generally the case in the recent past.
FOSTERING U.S. MILITARY AND ECONOMIC STRENGTH
The Relation to Controls
The fundamental question is the extent to which technology transfer controls intended to stem international leakage will also harm domestic communication and thus impede the contribution of American science to military and industrial advances. A second concern is whether controls
on exchanges could reduce the ability of the U.S. government to understand Soviet capabilities and intentions.
In order to evaluate the potentially adverse effects of controls, it is necessary to assess the extent to which research contributes to improved U.S. military and economic performance and the extent to which controls that diminish the traditional openness of the scientific enterprise—thus limiting informal feedback, delaying the discovery of errors, narrowing critical evaluation, and complicating the scientist’s search for predecessors’ results—will diminish its overall contribution. In other terms, the question is whether controls intended to bar international communication will inevitably impede scientific communication within the United States. In considering the impact of controls on economic progress, the key question is whether restrictions increase firms’ costs so as to affect the competitive position of American firms on the world market.
The Panel’s Assessment
The Panel believes that scientific research and technological development are best nurtured in an environment where such efforts are dispersed but interdependent. Openness and a free flow of information are essential aspects of such an environment. The technological leadership of the United States is based in no small part on a scientific foundation whose vitality in turn depends on effective communication among scientists and between scientists and engineers. Thus, the short-term security achieved by restricting the flow of information is purchased at a price.
The contributions of basic research may be limited by any bars that would prevent researchers from learning from each other’s work. If the research environment is altered in such a way as to discourage scientists from participating in certain areas of science, there will be fewer new ideas, and the pace of scientific advance will decrease. Openness, on the other hand, assures that new ideas are exposed to critical review by the best experts in the world. Only with such review can standards of U.S. research remain at high levels, and only such open review can ensure that logical errors, wrong paths, and unsupportable interpretations of data are avoided. Openness also fosters creativity, for it gives researchers assurance that they are building on the best and the newest ideas that exist worldwide. Prompt dissemination is also an enormous stimulus, since scientists are keenly competitive with one another. Moreover, openness leads with some regularity to the serendipitous answers to problems in one field through the use of research methods used for completely different purposes in other fields.
Openness and Military Strength
Restrictions on scientific communications may be costly because they could make work in those areas most relevant to U.S. military strength
much less attractive to the best researchers and the best students. Both the rate of innovation in science-based technologies and the supply of young new technical talent trained at the frontier of knowledge in these fields could decline.
World War II demonstrated that the nation’s top scientists and engineers—particularly those at the universities—constitute a major technical resource that can be of great value to our national security. The war showed decisively that military victory can depend on the outcome of the race between the scientists and engineers of adversary nations. Radar, for example, was developed just in time to play a critical role in the defense of Great Britain, and the subsequent development of microwave radar kept the allies ahead of other nations’ countermeasures.
On a smaller scale, scientific communication through exchange programs can contribute to U.S. national security in several ways. There have been several cases in which Soviet scientists have made significant contributions to American research efforts. Furthermore, visits in both directions allow U.S. scientists to bring back informed appraisals of Soviet capabilities in science and technology (although not all scientific institutes in the Soviet Union are open to U.S. visitors). Exchange visits to the Soviet Union by U.S. scholars in fields other than science have great value in the development of Sovietology in this country. This has led to a better understanding of how Soviet society functions, which is essential to the formulation of well-informed U.S. policies.
Openness and Economic Strength
Restraints on scientific communication can limit the efficiency with which important information is transmitted to and within industrial firms.2 The applications of new results from the research
The Panel found that there is a general acceptance in private industry of restrictions on the technologies that are formally controlled under Defense Department contracts or by classification. However, leaders of U.S. technology-based industries believe that more stringent export controls, or even the continuation of current controls, raise the risk of restricting the flow of information within their firms and between Europe, Asia, and North America in a way that may ultimately be costly to American society. They believe that openness in scientific communication is vital to the economic vigor of the United States. These leaders also value the freedom to employ any personnel well trained in high-technology areas from the best universities. Many call for clearer identification of the technologies that should be restricted; clearer enunciation of the rationale for their control and of the damage that would occur without control; and means for ensuring prompt removal of controls when a technology is no longer at the state of the art or when the U.S. monopoly on information ceases through the normal process of international diffusion (see Appendix C).
community outside the universities will thus be impeded, possibly leading to higher internal research costs or to the loss of world market shares as the U.S. product performance falls behind that of foreign competitors. Particularly vulnerable are high-technology, dual-use technologies such as high-speed microelectronics, because these rapidly evolving areas could fall under national security controls that would inhibit the best researchers and advanced students from entering—and advancing—the field. Many high-technology firms depend on outside research scientists for contract research and consulting, and rely on universities to supply young talent trained at the frontier of technology.
A striking example of the economic benefit of openness is provided by the history of the transistor. The free dissemination of information about transistors in the early 1950s gave an extraordinary stimulus to the electronic and computer industries and led to the great U.S. superiority in these fields. This superiority has meant not only superior capacities in particular types of military and civilian equipment but also a strengthening of the economy as a whole.
Security by Accomplishment
To summarize, current proponents of stricter controls advocate a strategy of security through secrecy. In the view of the Panel security by accomplishment may have more to offer as a general national strategy. The long-term security of the United States depends in large part on its economic, technical, scientific, and intellectual vitality, which in turn depends on the vigorous research and development effort that openness helps to nurture.
A strategy of security by accomplishment has several institutional components. First, universities have the tasks of training new scientists and engineers and conducting basic research, the source of long-term progress. Second, government laboratories undertake research directed to particular national interests in defense, medicine, space energy, and agriculture. Third, industry translates the results of research into new commercial and defense technology. It is important that all these institutions attain their full potential, for economic as well as for military reasons. Open scientific communication plays an important part in keeping scientists and engineers in government, industry, and universities aware of each others’ needs and findings.
PROTECTING EDUCATIONAL AND CULTURAL VALUES
The Relation to Controls
Controls on scientific communication could adversely affect U.S. research institutions and could be inconsistent with both the utilitarian and philosophical values of an open society. In order to investigate this potential effect, it is necessary to assess the extent to which controls on scientific communication disrupt the educational
process—a process that has all but merged with the research function in major American research universities. With respect to cultural implications, it is necessary to assess the conflict between such restrictions and the health of our political system, the First Amendment’s guarantees of free speech and a free press, and the need for an informed electorate.
The Panel’s Assessment
American universities are particularly vulnerable to restrictions on scientific communication because of two special characteristics, namely, the critical role of research at many universities and the intimate relationship between university-based research and educational programs, particularly advanced training in research. Over the coming decades the nation’s research performance will be heavily dependent on the continued health of university research programs, especially those in basic science.
Most university communities strive to maintain an “open society” where faculty and students freely share both the results of their research and the ideas that may develop into new areas of inquiry. In general, universities have seen such openness as an essential element of a scholarly environment. Controls on communication thus present a significant threat to a central tenet of university life and as a result are likely to discourage university-based scientists from participating in certain areas of science. Not only is the advance of science thereby slowed, but also the breadth of knowledge in the university community is thereby gradually diminished. Thus, the university can no longer fulfill its role as a central repository of knowledge.
With respect to cultural factors, the Panel believes that the costs of even a small advance toward government censorship in American society are high. The First Amendment’s guarantee of free speech and a free press help account for the resiliency of the nation. If political authority is to be exercised effectively, there must be trust in government on the part of those affected—a trust that is promoted by openness and eroded by secrecy. Openness also makes possible the flow of information that is indispensable to the well-informed electorate essential for a healthy democracy. Openness also strengthens U.S. institutions by allowing comparison with the performance of others and nurturing adaptation to changed circumstances.
THE FEASIBILITY OF CONTROLS
The Panel believes it is important to keep a realistic perspective on the feasibility of controls. There is a danger that proponents of control measures may make the mistake of equating the existence of controls with the lasting denial of technologies to the Soviet military. Because the United States has a monopoly on only a fraction of all technology, and because the Soviet acquisition effort is
carefully targeted, even totally effective U.S. controls would close but one channel among the many available to the Eastern bloc intelligence services. Efforts to restrict leakage by member nations of COCOM vary widely, and some industrially advanced nations, notably Sweden and Switzerland, are not members. Thus, in many cases losses can occur outside the United States and beyond the reach of its control efforts.
In addition, experience shows that it is difficult—if not impossible—to maintain total secrecy for long periods of time. Information leaks are inevitable—even when information is highly guarded—and others will, in time, make the same discoveries. The development of the atomic bomb provides an example of the latter assertion. Even without the the espionage attributed to Fuchs, Greenglass, and the Rosenbergs, the very use of the bomb revealed its basic character (through the production of intense radioactivity) and meant that trained scientists and engineers could easily make informed guesses about the bomb’s nature from the known body of scientific theory. The reduction of theory to actual practice is by no means as easy, but this example does illustrate that secrecy can never replace the need to develop new ideas. Indeed, the most important fact about a technology is probably its very existence—not its design details. Once feasibility is proved in one country, other governments can confidently launch development efforts of their own. Concealing the existence of American military technologies is usually impractical.
The Panel also notes the inherent limitations in technology transfer controls as they might apply to research campuses. It is unclear how controls can be successfully applied to many activities that commonly take place on university campuses, including lectures, seminars, and discussions of faculty and students with visiting scientists. All of these are potentially subject to export regulations, yet all are, by long tradition, open activities. Moreover, any attempt to apply controls to such activities in a research environment (e.g., monitoring the movement of foreign scientists or students on campus) are logistically impractical, and therefore the chances of successful implementation are slight.
BALANCING COMPETING OBJECTIVES: THE PANEL’S JUDGMENT
After listening to the testimony, weighing the evidence, and pondering alternatives, the Panel concludes that the best way to ensure long-term national security lies in a strategy of security by accomplishment, and that an essential ingredient of technological accomplishment is open and free scientific communication. Such a policy involves risk because new scientific findings will inevitably be conveyed to U.S. adversaries. The Panel believes the risk is acceptable, however, because American industrial and military institutions have the capacity to develop new technology with a speed that will continue to give the United States a differential advantage over its military adversaries.3
In any event, more than national security is at issue. Basic research investigations undertaken today may lead to applications in the long term (perhaps 10 to 20 years from now), often in unexpected ways. To attempt to restrict access to basic research would require casting a net of controls over wide areas of science that could be extremely damaging to overall scientific and economic advance as well as to military progress. The limited and uncertain benefits of such controls are, except in isolated areas, outweighed by the importance of scientific progress, which open communication accelerates, to the overall welfare of the nation. Security by accomplishment is a strategy that has served the nation well.
Principles for University Research
The Panel concludes that the vast majority of university research, whether basic or applied, should be subject to no limitations on access or communications.
Undoubtedly, some things must, by their very nature, be kept secret. It is clearly important, for example, to keep secret those properties of actual weapons systems that would enable a potential enemy to develop effective countermeasures. Where specific information must perforce be kept secret, it should be classified strictly and guarded carefully. The decision to accept or reject classified research projects or to establish off-campus classified facilities is a matter to be decided by universities.
The Panel concludes that there are a few gray areas of research that are sensitive from a security standpoint, but where classification is not appropriate. These research areas are at the ill-defined boundary between basic research and application and are characteristic of fields where the time from discovery to application is short. At present, a portion of the field of microelectronics is the most visible among the small handful of such new technologies.
providing the United States with a superior infrastructure in support of military technology. An important consideration here is the length of the military procurement cycle—the time from technological development to manufacture. Some have argued that the Soviet Union may be able to compensate to some degree, claiming it can introduce new technologies faster because it redesigns its military systems more frequently than the United States, providing an earlier opportunity to translate design into hardware. The Panel has not been able to explore this point, but it merits study.
Guidelines for Classified and Gray-Area Research
While it is impossible to specify classified and gray-area research with precision, there are some broad criteria that help to define the few areas in question.
The Panel recommends that no restrictions of any kind limiting access or communication should be applied to any area of university research, be it basic or applied, unless it involves a technology meeting all of the following criteria:
The technology is developing rapidly, and the time from basic science to application is short;
The technology has identifiable direct military applications; or it is dual-use and involves process or production-related techniques;
Transfer of the technology would give the U.S.S.R. a significant near-term military advantage; and
The United States is the only source of information about the technology, or other friendly nations that could also be the source have control systems as secure as ours.
In order to specify the areas where greater control would be appropriate, it may be useful to look at some examples of research that do not meet all of the above four criteria. Monoclonal antibody research is developing rapidly, and the interval from basic discovery to application may be short; but there appears to be no way in which this research could result in a significant military advance. Hence, there should be no need to impose controls in this field. Similarly, the science underlying aerodynamic design, even though it possesses obvious military significance, is a mature, slowly evolving field that is unlikely to provide any significant near-term military advantage to the Soviets. Thus, it too should be free of controls.
The Panel recommends that if government-supported research demonstrably will lead to military products in a short time, classification should be considered. It should be noted that most universities will not undertake classified work, and some will undertake it only in off-campus facilities.
In those few cases of government-sponsored research where national security considerations may require restrictions on publication, limitations on foreign access to facilities, or security classification, the Panel believes that certain guiding principles and procedures should be followed. The provisions of EAR and ITAR should not be invoked to deal with gray areas in government-funded university research. Rather, in the Panel’s view, appropriate procedures should be incorporated in research contracts or other written agreements in those rare cases where some measure of control is required. The advantages of such provisions are that they give prior notice to the researcher that the funded research may turn out to have national security significance and foster a spirit of negotiated accommodation that helps prevent future misunderstandings about the researcher’s obligations and recourse.
The Panel recommends that in the limited number of instances in which all of the above criteria are met but classification is unwarranted, the values of open science can be preserved and the needs of government can met by written agreements no more restrictive than the following:
a) Prohibition of direct participation in government-supported research projects by nationals of designated foreign countries, with no attempt made to limit physical access to university space or facilities or enrollment in any classroom course of study. Moreover, where such prohibition has been imposed by visa or contractually agreed upon, it is not inappropriate for government-university contracts to permit the government to ask a university to report those instances coming to the university’s attention in which the stipulated foreign nationals seek participation in any such activities, however supported. It is recognized that some universities will regard such reporting requests as objectionable. Such requests, however, should not require surveillance or monitoring of foreign nationals by the universities.
Restrictions on access to nonclassified research, whether to research results or to physical facilities, are outside the normal operating procedures of research universities. It is, of course, within the power of the government to deny or issue conditional visas to foreign nationals who are believed to be seeking skills or technical data that will significantly damage our national security. In extraordinary circumstances, the government may seek to ensure that government-provided resources are not used to support nationals of specified countries who seek to work in specified programs. Access to program resources by nationals of designated foreign countries may be limited either through research contract terms or through other agreements negotiated with particular universities. Such contracts or agreements should not attempt to deny physical access to any university space or facility to any person accepted by the university into its community. The danger to national security lies in the immersion of a suspect visitor in a research program over an extended period, not in casual observation of equipment or research data.
b) Submission of stipulated manuscripts simultaneously to the publisher and to the federal agency contract officer, with the federal agency then having 60 days to seek modifications in the manuscript. The review period is not intended to give the government the power to order changes: The right and freedom to publish remain with the university, as they do with all unclassified research. This does not, of course, detract from the government’s ultimate power to classify in accordance with law any research it has supported.
In some cases, a contractual agreement providing for simultaneous review of manuscripts at the time of their submission to scientific journals may be appropriate. A requirement for government comment within 60 days of submission of the manuscripts should provide adequate time for the government to assess the potential near-term military significance of the dissemination and to reach accommodation with the
researcher before public release. Experience suggests that disagreements about publication can almost always be resolved by discussion between the principal investigator and the technical contract manager. The Panel emphasizes that its support for a review period is not intended to support any government effort to veto publication, or to limit the government’s power to classify, in accordance with law, any research it has supported.
To help government policy officials to supervise the application of the gray-area research criteria and to gain perspective on the longer-term effects of the restrictions imposed on such research, there is a need to ensure that an accurate accounting of such restrictions is kept.
The Panel recommends that in cases where the government places such restrictions on scientific communication through contracts or other written agreements, it should be obligated to record and tabulate the instances of those restrictions on a regular basis.