There are two things about the DOD style of laboratory management that are noteworthy: reliance on outside laboratories for sophisticated exploratory work, and the existence of a special organization, the Defense Advanced Research Projects Agency (DARPA), to serve as a kind of "venture capitalist" for DOD. Relatively little of the most advanced exploratory work is being done in house. Much of the work sponsored by the Office of Naval Research is supported by contracts to industry and grants to universities. DARPA has carried this approach to its limit. It has no research installations of its own; rather, DARPA program officers define a research agenda, arrange for one of the services to be its procurement manager, and work with contractors to bring a particular concept say, a tank autoloader or a robot with threedimensional vision to demonstration. Many of DARPA's most promising ideas are culled from unsolicited proposals. This approach has led to major advances in supercomputer technology, computer-tocomputer communications (for example, the nationwide packet-switching network known as ARPANET), electronic warfare, materials science, and lasers (ref. 206). Indeed, DOD's tendency to go outside the walls is of long standing; one thinks of the establishment of RAND and the Aerospace Corporation as contract research centers for the Air Force, and the reluctance of the Navy's Special Projects Office to use naval laboratories in developing the Fleet Ballistic Missile. Since the early 1970s, DOD in general and the Air Force in particular have moved to reduce the proportion of basic and exploratory research carried out by government employees. The autonomy of DOD laboratories has thus been further reduced.* differences in service needs. This is why joint programs, beginning with the TFX plane of the early 1960s, have proved so difficult to manage. As the General Accounting Office noted of an attempt to merge three laser-guided missile seekers into one: ". the Navy Bullpup was ready for production at merger time, the Air Force laser Maverick was beginning advanced development, and the Army Hellfire was still in the conceptual stage. All the services eventually dropped out of the joint seeker program." GAO, Joint Major System Acquisition by the Military Services: An Elusive Strategy (NSIAD-84-22, December 23, 1983), pp. 14-15. * According to the National Science Foundation, between 1976 and 1983 DOD funding for intramural basic research increased at an annual rate of 9.5 percent, compared with 3.2 percent from 1973 to 1976. However, the laboratories' share of basic research at 5 percent remained unchanged. It seems to us that increased funding for intramural basic research including that carried out at government-owned, contractor-operated facilities, which are not counted by NSF as intramural - is very desirable, even if the laboratories' share of such research remains the same. The defense industry is unique in many ways: There are formidable barriers to the entry of new firms, especially the small ones which are so fruitful a source of innovation; only the few large firms already established as defense contractors are capable of managing sophisticated programs; and the shifting of funds to routine development work means that less money is available for In the end, the ability of a laboratory to deal on an equal footing with its principal sponsor is affected by many things: the reasons that led to the laboratory's creation (which may not be the same as the reasons for keeping it in being), the existence or absence of competing organizations able to do the laboratory's work, the size of laboratory programs relative to its agency's mission, and much more. In the next two sections, we see how two quite dissimilar kinds of research installations have interpreted their missions in the light of changing environments. Strengthening Productivity: The Role of the National Bureau of Founded in 1901, the National Bureau of Standards (NBS) (fig. 49) is charged to maintain and develop a national measurement system (ref. 207). Although it has been given many subordinate responsibilities, the Bureau's primary mission remains its reason for existence today. This national measurement system is not easy to describe because the boundaries of the system shift constantly. Even if we divide the system into user and supplier sectors, it will quickly appear that many institutions are both (table 11). U.S. National Bureau of Standards, Gaithersburg, Maryland FIGURE 49.-A view of the National Bureau of Standards complex near Gaithersburg, Maryland. The large building at the left houses the administrative offices. fundamental research. Under the circumstances, DOD laboratories serve an essential purpose as sources of new ideas, as a means of making DOD a more sophisticated buyer, and as points of contact with the universities. On the structure of technology development carried out on contract to DOD, see Jacques Gansler, The Defense Industry (Cambridge, Mass.: MIT Press, 1980), Chapter 4 and p. 304, n. 17. Table 11. Supplier and User Sectors in the National Measurement System Source: Science Policy Research Division, Congressional Research Service, Library of Congress, The National Bureau of Standards: A Review of Its Organization and Operations, 1971-1980 (May 1981), p. 6. What is the role of the NBS in this diffuse and voluntary system? To simplify somewhat, we would argue that the NBS performs three closely-related functions which are vital to maintaining the system. It provides the basis for uniform and accurate measurements throughout the economy; it drafts voluntary standards for determining that a product, a process, or a service meet certain criteria; and through its Institute for Computer Sciences and Technology, it recommends uniform dataprocessing standards in the Federal Government. This mission, in turn, can be carried out because of certain special features inherent in the Bureau's organization. The first is scientific competence in many areas: in materials processing, calibration, computer-aided design, mathematical modeling, structural analysis, and instrument design. So much may appear obvious. But without competence across many disciplines, the Bureau could scarcely play a significant role in developing standards. This system is fragmented, decentralized and, above all, voluntary (ref. 208). Without regulatory authority to enforce standards, the Bureau can gain acceptance for its standards only on the strength of the quality of the work that goes into drafting them. A second distinguishing feature of the Bureau is its responsiveness to the needs of other Federal agencies and industry. NBS is unique among Federal laboratories in the amount of work for others it performs. Much of this extramural work is required by law; but in any case, the Bureau has accepted many assignments, the better to carry out its mission. During the Second World War and for several years thereafter, the Bureau received up to 85 percent of its funds from other agencies, principally the military. Other agency funding has since tapered off and, as a matter of policy, Bureau officials prefer to maintain it at about 40 to 45 percent (ref. 209). What has kept the level of support this high has been the proliferation of Congressional mandates since the late 1960s. Congressional committees which have to do with NBS have seen it as a national resource, capable of contributing to economic growth and supporting the science and technology infrastructure. These mandates led to many problems: They caused the Bureau to carry on work that went beyond its competence in the physical sciences; Congress neglected to appropriate funds to cover these programs, forcing the Bureau to divert funds and manpower from its core activities; and it led to disagreement with OMB over whether a Congressional mandate in a given area say, in studying the economic effects of metallic corrosion - sufficed to make the Bureau the lead agency in that area. The point is that the Bureau's work for others affects its mission in ways which have no parallel in other Federal laboratories. Third, the Bureau's effectiveness depends not only on technical competence but on the perception by its sponsors as an objective, neutral authority. The Bureau has shunned a regulatory role and it has been generally cautious in taking on work that might give the appearance of competing with industry. It has been NBS policy to avoid the later stages of product development. At NBS, “. . . product development is usually not taken as far as it is in other Federal laboratories, where the usual policy is to continue developing a new technology to the point of pilot demonstration. Development at NBS usually ends with an earlygeneration prototype, a few steps prior to pilot demonstration. Then other Federal agency or private sector laboratories are left to continue development." (ref. 210.) Even where the Bureau has taken an activist view of its mission and has moved further into technology development, it has avoided large demonstration projects.* In 1983, for example, the Bureau installed an *Note, however, that it is not the case that if a Federal agency avoids supporting large demonstration projects, that any work up to that point is appropriate for Federal support. The test is the judgment - by the sponsoring agency, Congress, OMB, or all of them that the results of a research and technology development activity would be appropriable by private industry. Automated Manufacturing Research Facility (AMRF) in its machine shop (ref. 211). The AMRF is to be a small, modular flexible manufacturing system that will serve to test certain measurement concepts. The novelty of the facility is not in its hardware; all of AMRF's components are off the shelf. Rather, its purpose is to determine if different components from different manufacturers can exchange information without the need to modify software or protocols. This work is being supported by industry, through the loan of equipment and personnel; and the transfer of technology in this case, standards for the interfaces between components of computer-aided manufacturing systems will occur through the usual channels: industrial and trade associations, and the American National Standards Institute. The features so far discussed technical competence, neutrality, responsiveness to other organizations suggest that the Bureau is a rather conservative organization. They also suggest that, given the structure of the national measurement system, technical competence and mission performance are linked in a peculiarly intimate way. What remains to be discussed is how the Bureau's conception of its role in stimulating industrial productivity has changed under outside pressure. Of all Federal agencies, the Department of Commerce, the Bureau's sponsoring agency, is the one most closely tied to the ups and downs of the U.S. economy. As data gatherer (Census Bureau, Bureau of Economic Analysis), as a regulator of international trade, as a registrar of patents and trademarks, and as the parent agency of the National Weather Service, the Commerce Department touches the economy at many points. To some Congressmen, it seemed reasonable that Commerce in general and NBS in particular should play their part in stimulating technological innovation. But there is not (to put it mildly) unanimity as to the best approaches to government sponsorship of industrial innovation - let alone what the Bureau's role shall be. In the past ten years, the Federal Government, through the National Science Foundation, has provided seed money for university/industry cooperative research centers; these centers are intended to work on the cutting edge of technology, in biotechnology, polymer processing, robotics, computer graphics, microelectronics, and much more (ref. 212). Again, in 1980, Congress passed the Stevenson-Wydler Act, which authorized the Secretary of Commerce to establish an Office of Industrial Technology to assess the climate for industrial innovation and to propose methods for advancing it; and to create Centers for Industrial Technology, similar to the cooperative research centers sponsored by the National Science Foundation. The Act also requires every Federal agency operating one or more laboratories to set aside 0.5 percent of its research and development budget to support |