Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
The Japanese Challenge in High Technology DANEL I. OKIMOTO Japan may have to embark on a crash program to expand and upgrade its infrastructure in science and technology so that z~ can innovate. It can no longer exploit the advantages of latecomer status. It may not be able to follow a low-cost, low-risk strategy of second- to-market by capitalizing on low production costs. As comparative advantage shifts to the newly industrializing states, Japan will have to compete head to head with the United States in what has been the traditional U.S. stronghold, high technology. BACKGROUND Since the end of World War lI, He United States has dominated the area of high technology. Virtually all the major high technology industries, from nuclear energy to microelectronics, were started in the United States, most emerging out of the structure of the so-called militaIy-industnal complex and reaching, maturity in the commercial marketplace. Even today, U.S. pro- ducers hold the lion's share of world markets in most high technology prod- ucts. Indeed, individual American companies like IBM (computers), AT&T (telecommunications), and Boeing (commercial jet airliners) set the pace of competition, define standards for their industries, and demonstrate a capacity to dominate product markets even when competitors enter first. Less than 25 years ago, however, the same could be said of American manufacturers in heavy industnes. U.S. producers had a comer on We largest share of world markets, stood at the cutting edge of technology, and dictated Me pace of commercial competition. TwenW-five years ago (1960), U.S. companies accounted for more Man a quarter of world steel production, over half of total auto assembly, and (in the early- to mid- 1960s) nearly 90 percent of all color television sets produced. Famous American brand names U.S. Steel, General Motors, and Zenithwere synonymous win leadership in these industries. 541
542 DANIEL I. OKIMOTO It took less than two decades, however, for Japanese manufacturers to overtake U.S. front-runners. By 1980, Amenca's share of world production in steel had fallen from 26 percent to 14 percent, from over 50 percent to 21 percent in automobiles, and from 90 percent to less than 30 percent in color televisions. Japan's share of world production rose from 6.4 percent to 15.5 percent in steel, from less than 5 percent to 30 percent in automobiles, and from almost zero to over 50 percent in color televisions. The speed with which the Japanese overtook American pacesetters surmised everyone, in- cluding the Japanese themselves. Against the background of this experience In the "smokestack" sectors, the question arises: Is America's current leadership in high technology safe from the Japanese challenge? The Ministry of International Trade and In- dustry (MITI) has "targeted" virtually all areas of high technology as national pnonties. Does this mean that the Japanese will be able to replicate their smokestack successes? The Japanese may already be ahead in certain areas of high technology, such as robotics; as of 1982, Japan had installed more than Tree times the number of robots as the United States (President's Commission on Industrial Competitiveness, 1985, vol. 1:221. In other areas in which Americans still hold a lead, such as semiconductors and optoelec- tronics, U.S. companies are hearing the footsteps of the Japanese, who are moving speedily to close the gap. In the most lucrative commercial markets, such as computers, telecommunications, home and office automation, and medical instrumentation, Japanese manufacturers have already emerged as America's most fonnidable competitors, combining some state-of-~e-art technology with traditional strengths in manufacturing, pncing, and mar- keting. Although the Europeans possess strengths in certain market niches, none of the European states appears to be mounting a serious challenge across all areas. The race in high technology is shaping up as largely a bilateral competition between Japan and America, with Europe straining not to fall too far behind. There are not many fields in which Me Japanese can be counted out. Even in areas in which American preeminence once appeared relatively unassail- ablesuch as software, CAD/CAM (computer-aided design and computer- aided manufactunng), and laser technology the Japanese appear to be mak- ing substantial headway. Convinced Mat these technologies are of crucial importance for their capacity to compete effectively over Me long run, Me Japanese are mobilizing human and capital resources to close Me gap. Only in areas in which Japan is at a decisive disadvantage- as in military-related endeavors or in products for which the costs of energy or raw material inputs are prohibitively highcart Me Japanese be judged completely out of Me running Falling into this category are avionics, military hardware (due in part to Japan's self-imposed ban on weapons exports), commercial jet aircraft, space and satellites, and petrochemicals. In most other areas, Me high tech- nology sweepstakes appear to be wide open.
THE JAPANESE CHALl FNGE IN HIGH TECHNOLOGY 1 955- 1960 1960 1965 1 965- 1970 1970- 1975 1975- 1980 ~1 1 6.2 543 Technology contribution to Economic growth rate {%) FIGURE 1 Technology's contribution to Japan's economic growth, according to analysis by Dr. Hisao Kanarnon, president of Japan Economic Research Center. (Figures for 1975- 1980 include estimates.) Japan's swift emergence in high technology is unsettling, because Amer- icans used to take comfort in the belief that, for whatever reason, the Japanese were inferior when it came to technological innovation. Of 500 seminal breakthroughs in technology identified between 1953 and 1973, only 34, roughly 5 percent, were achieved in Japan, compared win 63 percent in the United States (Moritani, 1982:159-1731. Indeed, lapan's relative back- wardness in technology has been reflected in its heavy dependence on foreign technology. It has run chronic deficits in its technological balance of trade throughout the postwar period (Ozawa, 1974:801. According to one estimate, Japan's `'economic miracle" can be ascribed largely to the contributions made by technology, much of which was imported from We United States and adapted or incrementally improved (see Figure 1~. Importing the most advanced foreign technology had the effect of setting in motion a virtuous cycle, fostering heavy investments in new plant facilities (which embodied new production technology), stimulating economic grown, and generating greater demand for products manufactured under foreign li- cense. In addition to its import-substitution effects, Me assimilation of foreign technology also had Me effect of making Japanese manufactured goods in such industries as electrical machinery, chemicals, and iron and steel more competitive in foreign markets. The mutually reinforcing nature of import substitution and export expansion, made possible by the assimilation of for- eign technology, thus undergirds Japan's record-setting economic grown (Ozawa, 1974:4~5 11. In a very real sense, therefore, Japan's so-called economic miracle owed at least as much to the availability of foreign know-how as it did to indigenous
544 DANIEL 1. OKIMOTO technology The importance of imported know-how is underscored by MITI's survey of Japanese business leaders in 1978 concerning the relative con~m- butions made by domestic and foreign technologies to product quality and production processes. According to MITI's survey, purely indigenous tech- nology accounted for only 5 percent of the improvements in Japanese product quality and 17 percent of the advances made in production processes (Ministry of International Trade and Industry, 1982:361. Although such survey data should not be taken too literally, they tend to conf~rrn what appears to be a fairly widespread impression: namely, that Japan has not been a seedbed of scientific ferment or technological originality. This image of Japan as a technological follower and the contrasting, image of Japan as a technological innovator are elaborated below. Image 1: Technological Follower The impression that Japan is a technological follower gains credence by looking at a variety of indirect indicators. Despite its large population base, for example, Japan has won fewer Nobel Prizes in the basic sciences than much smaller countries, for example, Sweden, the Netherlands, Switzerland, and Belgium. In 1964 Japan spent only 1.4 percent of GNP on R&D, com- pared with America's 3.4 percent and the United Kingdom's 2.3 percent. In 1971 the number of researchers per 10,000 population base was only 18 in Japan, compared with 25 in the United States (Ministry of International Trade and Industry, 1982:42, 441. Against this background, it is not hard to un- derstand how the stigma of imitator and technological free rider came to be attached to Japan. A variety of reasons have been cited for Japan's relative lack of techno- logical originality. They range from historical circumstances (especially Ja- pan's status as an industrial latecomer) to sociocultural impediments (e.g., the tendency toward group conformity), inadequacies of Japan's educational system (e.g. deficiencies in university-based research), and institutional factors (e.g., the lack of a venture capital market). Other obstacles to in- novation sometimes cited include Me practice of seniority-based, lifetime employment (said to limit the diffusion of innovation) and Conservative, risk- averse attitudes on the part of highly levered Japanese corporations (alleged to inhibit bold R&D investments). Whether the Japanese are simply incapable of technological innovation, however, is far from clear. Those refusing to believe so can point out that the same kind of doubts used to be expressed about Japan's alleged inability to ensure high standards of quality control. Look at what the Japanese have done to shatter that myth! Even those who believe that Japan is a follower tend to view Me problem as essentially correctable, provided certain insti- n~tions, practices, and policies can be overhauled. Unless one assumes that
THE JAPANESE CHALLENGE IN HIGH TECHNOLOGY 545 the problem is genetic, therefore, Japan's undistinguished technological record can be traced back to structural impediments in Japan's catch-up R&D system. Very much aware of these impediments, the Japanese are trying hard to overcome them through structural change. Some of these changes are being explicitly engineered for the purpose of building, an environment more con- ducive to innovation; others are the by-product of evolutionary development. The essential point is that most of the impediments are not insurmountable. From the standpoint of short-term results, probably the only obstacle that may not be immediately changeable is the deeply ingrained, sociocultural values and patterns of socialization. To the extent that Japan's problems are entangled with sociocultural impediments, the prospects for rapid correction do not appear bright. On the other hand, since culture consists of a "collage'' of diverse, in- consistent, and incompatible ingredients not a single, coherent blend its impact on something as complex as innovation is exceedingly hard to mea- sure. Out of this collage the specific elements that emerge to affect human behavior usually depend on the structure of institutions and Me set of policies and practices that give culture its concrete shape within a given, time-bound context. One indirect way of altering what are otherwise deeply embedded cultural parameters, therefore, is to restructure the institutional framework within which they exist. Japan appears to be doing just that. It is trying to upgrade the quality of research at leading Japanese universities. The embryo of a venture capital market seems to be taking shape. Japanese labor markets are adapting to the functional requirements of high technology. The government is doing all it can to push Japan beyond the frontiers of technology by organizing a variety of ambitious national research projects in such seminal areas as new materials and optoelectronics. And with the deregulation and internationalization of financial markets, many Japanese corporations have reduced their levels of dependence on debt financing in order to cope with the loss of insulation against wide interest rate fluctuations. All these changes could significantly affect Japan's capacity to innovate. Public policies are also undergoing change. Having graduated from being a latecomer, Japan finally finds itself in a position to take on the challenge of trying to innovate at We frontiers of technology. Prior to this time, Japan had been wholly absorbed in the task of industrial catch-up. In this sense, We question of technological originality is only meaningful now that the era of playing catch-up is over. Because frontier innovation has become more important than ever, Japan seems bent on mobilizing to upgrade its tech- nological capabilities. It is now spending 2.4 percent of an ever-expanding GNP on R&D, and the projections are that allocations will eventually reach 3 percent in the 1990s. The number of researchers per 10,000 population
546 DANIEL 1. OKIMOTO has risen from 18 to 27. This growing investment of capital and manpower in R&D is beginning to pay dividends. Although Japan still runs an overall deficit in its balance of technological trade, the ratio of Japanese technological exports to imports is climbing. What used to be a 1:47 ratio has become a 2:3 ratio; in 1980, exports of Japanese technology amounted to $638 million and imports of foreign technology came to $958 million (Ministry of Inter- national Trade and Industry, 1982:1731. Areas in which Japanese technology is so advanced that it can be sold overseas include agricultural chemicals, construction equipment, transportation machinery, electrical machinery, iron and steel, and ceramics areas in which Japan also happens to be successful at exporting finished goods. Image 11: Technological Leader Nonvithstanding its imitator stereotype, Japan has managed to achieve state-of-the-art technology in certain fields of telecommunications (e.g., fiber optics), microelectronic components (e.g., gallium arsenide memory chips), robotics (e.g., numerically controlled devices), office automation (e.g., com- puter peripherals), nuclear energy (e.g., fast breeder reactors), and biotech- nology (e.g., fermentation). It is even making big strides in technologies in which Amenca's lead once seemed insu~ountable, such as in artificial intelligence (e.g., expert systems). Japan's rapid technological progress and its long-term commitment to compete across the board suggest an image that stands in sharp contrast to that of technological imitator and free rider: namely, that of a technological leader or possibly even a pacesetter. Indeed, against the background of lapan's astounding accomplishments in heavy manufacturing and the un- expected speed of its advance in high technology, a new image has emerged outside of Japan: Mat of an almost invincible Goliath, capable of vanquishing any rival, whatever the field of endeavor. Yesterday, it won in Me smokestack sectors; tomorrow, it might be high technology; and thereafter, the service sectors. Japan is thought to possess superior strength in such areas as political stability, economic policies, government-business cooperation, labor-man- agement relations, corporate financing, business-banking ties, and so on. It is almost as if "victory" is assured when tile government "targets" an industry as central to its economic future and proceeds to mobilize massive resources to ensure eventual domination. Which Is the "Real" Japan? Is Japan a technological imitator and industrial overachiever? Or an astute learner and unbeatable colossus? Is America up against a David or a Goliath? Will Japan dislodge the United States from its current position of dominance
THE JAPANESE CHALLENGE IN HIGH TECHNOLOGY 547 in high technology as convincingly as it did in He smokestack sectors? Or has Japan reached the limits of its phenomenal postwar growth? The answer to these questions is bound to have a crucial bearing on America's industrial future, not to mention the fundamental shape of the world system. In order to provide an adequate answer, one must evaluate lapan's strengths and weaknesses, review where it has advanced arid where it has not, understand some of the underlying reasons, identify where significant changes seem to be taking place, and assess what it all means. Of course, no analysis- no matter how detailed can offer a definitive assessment on which accurate predictions can be based. The variables are much too complex. The most that can be realistically achieved is to provide a crude overview of some of the major factors at work in Japan's transition from a war-ravaged economy to one based increasingly on high technology. JAPAN'S TECHNOLOGICAL STRENGTHS AND WEAKNESSES Perhaps the best way of approaching the questions just raised is to examine in what fields of high technology He Japanese have fared comparatively well and in what fields they have been less successful. Such a review should help sort out He characteristics Hat underlie He patterns of successes and short- comings. Consumer electronics and semiconductors stand out as perhaps the best-known success stones. CAD/CAM technology, avionics, and space and commercial aircraft represent technologies in which lapan's progress to date has been notably less impressive. The review begins win the common fea- tures of Japan's successful forays into high technology. Successful Industries Consumer Electronics Despite the fact that transistors and integrated circuits were invented in the United States and despite Amenca's pioneenug work in solid-state physics, Japan was He first country to succeed in com- mercializing transistor technology for radios and televisions; miniaturization revolutionized the entire consumer electronics industry. Japan's commercial successes, especially in the early stages, can be attributed largely to several factors: the availability of basic patents from the United States, the ability of Japanese companies to modify imported technology and bring it to very high levels of refinement, sustained and growing levels of capital investments in R&D, low production costs, aggressive pncing, and mass marketing at home and abroad. A typical example of Japanese product improvement based on foreign patents is the video tape recorder, one of Japan's biggest export items (accounting for nearly $6 billion in export revenues), which happened to be invented in the United States but was adapted and commercially ex- ploited by He Japanese.
548 DAN EL I. OKIMOTO - The Japanese have also excelled in consumer electronics because of heavy and sustained investments in both R&D and in new plant facilities. To survive in Japan's fiercely competitive environment, Japanese corporations must allocate a significant portion of revenues for investment in ongoing R&D and manufacturing facilities. Indeed, low-cost, highly reliable manufacturing is a hallmark of the country's industrial success. Part of Japan's mar~ufac- turing prowess can be attributed to the quality of its labor force, techniques of quality control, and extensive subcontracting networks; but part of the explanation can also be traced back to the high priority placed on investing in world-class process and production technology, which are thought in Japan to constitute the keys to commercial success. The emphasis placed on process and production technology is accentuated by such distinctive characteristics of Japanese industrial organization as the practice of permanent employment. Company secrets on process and pro- duction technology can be safeguarded in Japan, a task far more difficult in the United States where labor is more mobile. Investing in process and production technology thus makes sense because it can give companies the competitive edge they seek. In sum, lapan's experience in consumer elec- tronics demonstrates that if latecomers cannot compete at the cutting edges of new-product design, they can still compete effectively as second-to-market entrants by concentrating on manufacturing and process know-how, incre- mental product improvements, and mass marketing. Japan's success at mass marketing has grown out of a large and rapidly expanding domestic market (protected from the 1960s to the mid-1970s by "infant-industry" measures), access to big overseas markets, and aggressive pricing (leading in some instances to charges of dumping abroad). Gaining large shares of Me world consumer electronics market has provided a spnng- board on which diversified Japanese electronics giants have been able to expand into technologically related industries such as semiconductors, computers, and telecommunications. Hand-held and desk calculators, for example, created brisk demand for mass-memory chips and generated mo- mentum for the development of Japan's semiconductor industry (Okimoto et al., 1984:1791. One noteworthy aspect of the consumer electronics experience is that Japan's development in this field took place outside the scope of industrial policy "targeting." Consumer electronics did not grow out of a long-te~ M1TI blueprint or plan. It was not weaned on preferential credit allocations, ambitious national research projects (except for related industries like semi- conductors and information processing), research subsidies, extensive ad- ministrative guidance, or government intervention. Only in such areas as export facilitation and protection of infant industries did MITI extend a visible hand In most over fundamental respects, market competition supplied Me kinetic energy behind the development of the consumer electronics industry.
THE JAPANESE CHALLENGE lN HIGH TECHNOLOGY 549 This illustrates that technological innovation in Japancontrary to popular misperception- does not depend necessarily on industrial policy targeting (Okimoto, 19831. The govemment's power to foster innovation is not nearly as great as is sometimes assumed (Semiconductor Industry Association, 19831. It can iden- tify the high-pnority technologies, to be sure, but it cannot guarantee that private corporations will succeed in commercializing them. It can throw up a cordon of infant-indus~y protection around domestic markets in order to keep foreign competitors out, but that often has the perverse effect of dulling incentives to innovate. It can channel subsidies into "targeted" R&D, but that can lead to distortions and waste in aggregate R&D investments. What lies within the government's effective power is largely limited to the creation of a healthy environment for business growth. MITI officials believe that a favorable overall environment is best achieved through sound macroeconomic policies, generous tax provisions, and compensation for deficiencies in the market mechanism. It is no accident that in high technology, the two most dynamic countries the United States and Japan are also the ones where market competition is keenest. Semiconductors Most of the factors cited as an explanation of Japan's success in consumer electronics the importation of basic foreign patents, high and sustained investments in R&D and state-of-the-art manufacturing facilities, superior process and production technology, economies of scale (achieved in part through aggressive export efforts), severe price competition, and so on also account for its progress in semiconductors. Japan's fast- paced development in semiconductor technology since the early 1970s is reflected in the fact that it has come from nowhere to capture more Man a 20 percent share of world production of integrated circuits. Japanese man- ufacturers are especially adept at producing very large scale integrated (VLSI) circuits, most notably random access memory (RAM) chips, which are based on straightforward, fairly predictable technologies. Japan is also advancing rapidly in the production of semiconductor equipment, an area that used to be dominated by U.S. merchant houses but which is now witnessing notable market inroads by specialized Japanese companies owned by the Japanese electronics giants. These companies produce the gamut of production equip- ment, including mask aligners, ion implanters, die-sorting machines, pro- cessing treatment machines, and testing equipment. As with consumer electronics and heavy manufacturing, Japanese man- ufacturers of semiconductors place tremendous importance on process tech- nology, an emphasis that seeks to take full advantage of Japan's outstanding strengths in process innovation. Examples of state-of-the-art Japanese semi- conductor process technology include very fast automated bonding machines, uniform film epitaxial processes, highly pressurized oxidation, low-temper-
550 DANIEL I. O~MOTO ature passivation, anodization processes, and sophisticated ion-beam ma- chines (Okimoto et al., 1984:511. Here again, nearly all He basic breakthroughs in semiconductor process innovation, including electron beam lithography, ion implantation, and plasma etching, first occurred in the United States. Making use of these breakthroughs, the Japanese adapted and incrementally improved on them in ways that eventually enabled Japanese equipment man- nfacturers to compete in their own market against American imports. Looking at the area of seminal new-product design, however, Japan's contributions have not been very impressive relative to its share of world semiconductor production. Except for Esaki's discovery of the tunnel diode, practically all seminal breakthroughs have been made outside Japan. Until the l980s, Japan gave very little indication of being able to compete in complex, state-of-the-art semiconductor technology. Indeed, owing to Heir limitations, Japanese companies felt compelled to enter into licensing or second-sourcing agreements with leading American designers of micropro- cessors, logic devices, and semicustomized and customized chips. The government has played a larger role in the development of the semi- conductor industry than it did in consumer electronics. As a '`targeted" technology, semiconductors received industrial policy supports never ex- tended to consumer electronics early tariff protection, national R&D proj- ects, R&D subsidies, and the like. However, precisely how much of a difference industrial targeting made is hard to measure. It was probably not decisive. National research projects, like He VLST, helped to mobilize resources and to close the technological gap with the United States, but in all likelihood, Japanese companies would have narrowed the gap anyway, albeit at a slower pace. The amount of government funding for R&D in high technology (not just electronics) is su~pnsingly modest. From 1980 to 1990, a conservative estimate places it at roughly $2.3 billion, or about $230 million per year, an amount substantially smaller than Hat of the U.S. or French government (Okimoto, forthcoming, Ch. 21. If government funding as a percentage of total R&D expenditures for all sectors is compared for Japan and over leading counmes, Japan comes out win the lowest share, even controlling for defense-related expenditures (Table 11. What about government funding across sectors as a percentage of R&D and R&D as percentage of total sales (research intensity)? The comparative data in Table 2 show a significantly higher level of research intensifier in the United States, but note how much higher the percentage of government funding is in most categories. Since He categories are broad, He data should be interpreted with care. Nevertheless, the contrast in government funding for the category of electrical machinery (into which semiconductors fall) is spiking. As Tables 1 and 2 reveal, the notion that technological progress in Japan is the by-product of heavy state subsidization simply fails to accord win the facts. The private sector in Japan carries a heavier share of the
THE JAPANESE CHALLENGE IN HIGH TECHNOLOGY 551 TABT E 1 Government Funding As a Percentage of Total R&D Expenditures, Selected Countnes Government Funding Country Total Non-defense-related Japan (1980) 25.85'o 25.4% United States (1980) 47.9 33.2 England (19~78) 48.1 31.6 West Ge~`any (1979) 46.8 43.6 France (1979) 51 .1 31.9 SOURCE: Kagaku Gijutsu-cho, KagaJcu Gijutsu Hakusho (Science and Technology White Paper) (Japanese Government Printing Office, 1982). national R&D burden than in other major non-Communist countnes. The Japanese government's role in semiconductors, as in most "targeted" tech- nologies, should be understood in this context. While government assistance should not be ur~dereshmated, particularly with respect to coordination and consensus building, We main driving force behind the semiconductor indus- tIy's development has been We energy, manpower, and capital resources invested by the private sector. Of all possible government actions having an impact on We growth of high technology industries, the most ~mponant has not been industrial policy targeting; rather, it has been effective macroeconomic management. Mon- TABLE 2 Government Funding and Research Intensity, 1981 (percentages) Category - Chemicals Petroleum refining and extraction Rubber products Ferrous materials Nonferrous metals Machinery Electncal machinery Telecommunications Transportation equipment Precision instruments SOURCES: Sonfu, Pagan Gijutsu K:enkyu Chosa Poke, 1982 (Survey Report on Science and Tech- nology, 1982) (Japanese Government Printing Office, 1982); National Science Foundation, Research and Development in Irld~stry, 1981 (Washington, D.C., 1982); Gary R. Saxonhouse and Daniel I. Okimoto, Technology and the Future of the Japanese Economy. Paper prepared for Japanese Political Economy Research Conference, Honolulu, Hawaii, January 1985. Japan Gov't Research Funding Intensity 0.82 United States . Gov't Research Funding Intensity 7.19 3.05 4.48 0.18 2.32 1.44 1.37 2.18 4.52 4.72 2.69 3.73 4.49 2.82 1.63 1.69 1.63 3.88 0.46 7.29 23.8 25.0 37.6 10.9 37.9 33.9 59.8 17.3 3.83 0.72 2.56 0.81 1.21 2.57 6.82 8.90 8.37 8.38
552 DANIEL 1. OKlMOTO etaTy and fiscal policies cannot help but be important because they affect interest rates, inflation, and growth rates, the basic dete~Tninants of economic vigor or stagnancy. At the early stage of its industrial life cycle, the semi- conductor industry was especially responsive to macroeconomic conditions, owing to the high elasticity of demand for semiconductor products. The higher the growth rate, the faster its development. Its rapid growth since the mid-1970s, in fact, can be attributed in significant measure to the soundness of macroeconomic policies (far more than industrial policy). Simply put, Japan's macroeconomic perfonnance dunug the period provided the sem~- conductor industry with a healthier climate for vigorous growth than did that of the United States and the European countries. This, in turn, made the implementation of industrial policy easier and more effective. Looked at in comparative perspective, the pace of Japan's development in microelectronics has been remarkable. Japan may already have passed the United States in complicated, superconductivity technologies, like gallium arsenide and Josephson junction. Unlike IBM, which abandoned its work on Josephson junction after many years of expenmentation, Japanese companies appear bent on carrying their R&D through to commercial fruition; Japanese executives obviously believe Mat IBM's decision may have been premature. If their forecast proves accurate, the Japanese might find themselves in a position to make Weir forest major contribution to microelectronics through the technology of superconductivity, a technology that could alter the whole field of silicon-based engineering. Regardless of whether they do, the point is that the Japanese are dem- onstrating Mat they can innovate at the frontiers of technology, not just in the areas of incremental product improvement and process adaptations. They have done very well in fields other than semiconductors and consumer elec- tronics in optoelectronics, telecommunications (e.g., power transmission cables), pharrnaceui~cals, fine ceramics, and new composite matenals. The evidence is sufficiently plentiful, indeed, that the myth of Japan's techno- logical inferiority can be put to rest. Innovation in Non-E'cport-Dependent Industries As pointed out earlier, Japan tends to excel at technologies that are closely tied to commodities with huge export markets- continuous casting in steel, emission-control tech- nology for automobiles, and so forth. The powerful pull of overseas demand has helped Japanese companies move down the learning curve and advance technologically. What Is not as well known, however, is that Japan has also done well in some industries Mat are not dependent on exports for a significant share of production, like pharmaceuticals (e.g., Interleukin-lI), b~otechnol- ogy (e.g., fermentation, gamma interferon from synthetic genes), and tele- cornmunications (e.g., digital switching equipment, power transmission cables). This suggests ~at, even without the boost of overseas demand, Me Japanese
THE JAPANESE CHALLENGE IN HIGH TECHNOLOGY 553 are capable of innovating on a par with front-running foreign competitors. Of course, Me availability of foreign know-how and the size of Japan's domestic market have helped make innovation in non-export-onented sectors possible. The smaller size of markets in individual European countnes cer- tainly makes it harder for local manufacturers there to compete. From this standpoint, Japan and America hold a major advantage over European en- trants in the high technology sweepstakes. It is also worm noun" Nat some of these domestically preoccupied but technologically vital industries used to be shielded (and still are to a significant extent) from the full force of international competition. Usually, protection- ism diminishes incentives to innovate, since profits can be madewith survival assured without having to face up to the threat of foreign com- petit~on. Normally, this is a formula for commercial complacency and tech- nolog~cal stagnation. That We state of technology has not stagnated is a reflection of Me fact that, even behind closed doors, Japanese producers have had to operate in a domestic crucible of fierce inteFfirm competition. (This unusual, seemingly incongruous combination of domestic protectionism and technological dynamism helps to explain why key Japanese industries have been able to graduate from infant-industry status to world<lass competitors while protected industries in over countries fail.) Characteristics of Technological Achievements By way of concluding this portion of the analysis, the following list summarizes some of Me com- mon charactensucs of Japan's best-knoun technological achievements: Technological Features · Predictable technological trajectories (RAMs) · Known theoretical parameters (solid-state physics) · Emphasis on process and production technologies (ion-beam etching) · Steep learning curves · Concentration on applied research and development · Importation and modification of foreign know-how Industrial Organization · Dominance of large corporations · Fierce interfirrn competition · Close, cooperative relations with networks of small subcontractors · Lifetime employment: relatively limited labor mobility between fins · Reduction and diffusion of risk though organizational mechanisms Commercial Factors · High and sustained R&D investments · Large and growing domestic demand · Usually, a powerful export thrust · Difficulties of penetration by foreign competitors · Success even in sectors not ''targeted'' by government (consumer electronics)
554 Government Policy DANIEL 1. OKIMOTO · Overwhelming concentration on commercially valuable technology · Targeting of `' strategic" sectors and technologies · Consensus building across public and private sectors · Cooperative national research projects in areas of high risk and heavy cost but of potentially great commercial benefit · Modest R&D subsidies · Use of positive incentives (taxes, pragmatic antitrust enforcement) · Creation of favorable macroeconomic environment "Targeted" Technologies: Slower Progress Tunning now to areas in which the pace of technological progress in Japan has lagged that of the United States, space, commercial jet aircraft, software and CAD/CAM, and lasers represent "targeted" technologies In which He Japanese have progressed more slowly than expected. The characteristics Cat have made progress in these technologies difficult for Japan can be summarized as follows: Technological Features · Heavy dependence on basic scientific research (space, lasers) · Complex parameters for problem solution (aircraft design) · Technological trajectories not clearly predictable (CAD/CAM) Complicated systems integration of a multitude of components (aerospace, aircraft) · Large-scale, multidisciplinary R&D of long gestation (aircraft, space) · Commercial uncertainties or the prospect of limited spillovers for commercial markets · Unavailability of classified technology from the United States Industrial Organization _ . . · Limited experience in organizing international consortiums for subcontracting of R&D and production (aircraft) · Exceedingly high bamers to new entry (aircraft, space) · Lack of a venture capital market · Labor and capital market constraints on the exploitation of new technological on pornlIiities by independent, small-scale companies · Inhibitions on the diffusion of technology by limited labor mobility Commercial Factors · Exceedingly high costs of R&D and manufacturing (aircraft, space) Comparatively small volumes for co~runercial sales (aircraft, space) · Very high per unit prices (aircraft, space) · Limited domestic demand (space, aircraft) · Dim prospects for large-volume exports (except for lasers) Government Policy Need to tie large budgetary outlays to national defense or security justifications · Difficulties of building very costly support infrastructure for testing, flight simll- lation, and so on (aircraft, space) · Limited powers of procurement
THE JAPANESE CHALLENGE IN HIGH TECHNOLOGY 555 As indicated above, the Japanese have had difficulty dealing with tech- nologies that are highly complex, not very predictable, and heavily dependent on basic research. Japan's strength and overwhelming emphasis have been applied research, dealing with known parameters and predictable technolog- ical trajectories (as in the geometric progression of density in integrated circuits). The Japanese are superb at problem solving, particularly when it comes to applied eng~neenng problems. Neither their college nor on-thejob training has prepared them to excel at less straightforward problems involv- ing, say, systems software or CAD/CAM technology. Problems in the Educational System Japan's educational system, though good at turning out students whose performance on standardized mathematics and science tests places them at or near the top of all national groups, does not appear to bring out creativity in the most gifted students. It stresses rote memorization over creative synthesis or critical analysis. Japanese students, even at the doctoral level, shy away from tackling research topics that require building new theories (Interview with a professor of computer science at the University of Tokyo, May 10, 19841. They prefer problems with known parameters and for which solutions can be found using extant theones. Such training suggests strengths at solving discrete problems but weaknesses at research tasks that call for creative conceptualization, like the design of advanced systems software and large jet aircraft. In certain technologies that are closely tied to the basic sciences, like space, aircraft, or biotechnology, Japan may also be at a disadvantage because of the lower level of scientific research at Japanese universities compared with leading universities in the United States, England, and other of He European countnes. Moreover, researchers at Japanese corporations do not interact as synergistically with university faculty as their counterparts in the United States. As civil servants, faculty members at leading national uni- versities cannot spend substantial portions of their time consulting for private industry. The value of close university-indusby ties can be seen in the history of interactions between Stanford University and Silicon Valley firms. It is no accident that industrial centers of high technology in He United States have sprung up around major research universities. So long as Japan's R&D needs could be met without reliance on basic research, the inadequacies of its universities did not hamper its industrial progress. However, as Japan has moved into high technology, He shortcom- ings are becoming more senous. Indeed, Japan's whole R&D system- oriented overwhelmingly toward applied research and its rapid commercial- izaiion, with scant attention paid to basic research may require extensive revamping. How far and fast Japan is able to progress in, say, pharmaceuticals and biotechnology may come to hinge increasingly on the quality of its
556 DANIEL I. OKIMOTO university-based research in biology, biochemistry, chemistry, and other natural science fields. Japan laser industry has failed to keep pace with its U.S. counterpart largely because university-based research in applied physics has lagged. University-based research in lasers has not received anywhere near the level of support that it has in the United States. Moreover, effective allocation of research funds in Japan is seriously hampered by the fact that Me country's leading universities are public, not private, and therefore dependent on bud- getary allocations from the conservative, bureaucratically rigid Ministry of Education. There is no system of competitive peer review, comparable to that used in the United States, by which research support is allocated to the most deserving scientists and projects at whatever university they happen to be. Funds tend to be funneled through the same hierarchy of universities, almost as if inertially following the grooves of past channels. Although the system is not devoid of meritocratic considerations, it is not nearly as mer- itocratic, flexible, or cost-effective as Amenca's competitive peer review system. In a system like Japan's, exciting new fields of scientific inquiry can be neglected. The bureaucratic rigidities built into Japan's educational system also make it difficult for the country's leading public universities to adjust graduate enrollments in specific fields to changing student interests and needs. At the University of Tokyo, for example, the number of graduate students in com- puter science is very smalldespite considerable student interest and Me obvious commercial importance of the field. Why are enrollments so small? The main reason is administrative inflexibility. To expand the number of graduate students in, say, applied physics or computer science, the university must reduce the number in other fields, such as agriculture; the aggregate number of students must stay the same. This means that Me departments of applied physics or computer science must persuade the agriculture department to give up part of its enrollment allocation, something any department is unlikely to do. Hence, Japan's capacity to adapt to the changing needs of university-based research and manpower training is circumscribed by the ri- g~dities of its educational bureaucracy. National Secur~ty-related R&D aM Government Procurements Besides the deficiencies in higher education and basic research, Japan's R&D system is also not geared to underwrite very expensive but militarily important projects, such as aerospace and aircraft. These industries obviously suffer from the government's inability to sponsor very costly programs involving multidisciplinary research projects of high risk, great uncertainty, and long gestation. Comparing government R&D expenditures by broad categories (Table 3) reveals a striking contrast between Japan and the United States and France with respect to three categones~defense and aerospace, agriculture,
THE JAPANESE CHALLENGE lN HIGH TECHNOLOGY TABLE. 3 Government R&D Expenditures, by Fields, 1980 (percentages) 557 Field United States Japan Price Defense and aerospace 47.3 16.3 49.3 Industry 0.3 12.2 7.9 Agnculn~re 2.7 25.4 4.3 Energy and infrastructure 14.2 34.4 16.0 Health and welfare 15.2 11.2 7.5 SOURCE: Adapted from Gary R. Saxonhouse and Daniel I. OkilTloto, Technology and the Future of He Japanese Economy. Paper prepared for Japanese Political Economy Research Conference, Honolulu, Hawaii, January 1985. and industry. Within the context of these spending pnonties, it is hardly surprising that Japan's space and aircraft industries have lagged far behind American and French front-runners, or that Japan is advanced in technologies related to agnculture (e.g., agricultural chemicals), industrial applications (e.g., numerically controlled robots), and energy utilization (e.g., supercon- ductive toroidal magnets for nuclear fusion reactors). A feature of Japan's R&D system that stands in stark contrast to the American system is the Japanese govemment's relative lack of procurement powers for high technology products. No ministry possesses anything like the vast sums of money at the disposal of the U.S. Department of Defense or the National Aeronautics and Space Administration (NASA). Only 25 percent of the Japanese Defense Agency's small budget, which is only 1 percent of GNP, is earmarked for hardware acquisition. MITI has almost no budget for procurements. Only Nippon Telegraph and Telephone (NTI), a public corporation (but pnvatized in April 1985) had the budgetary means to purchase large quantities of high technology products. This situation, quite unusual among advanced industrial states, has had a number of important i~nplicabons for Japan's R&D process: (1) macroeconomic policies aimed at expanding aggregate demand have played a greater role Man targeted industnal policy in promoting the grown of high technology products; (2) MIT] has had to rely predominantly on supply-related incen- tives, not demand-pull measures; (3) inefficient resource allocation, waste, and politicization have been kept under comparative control; (4) Japanese producers have had to recoup up-front R&D investments quickly through profits earned from consumer markets, and this has prompted Japanese man- agement to stress applied research and development instead of basic or pro- totype-development research; (~) with no assurances of government demand for new products, Japanese companies have followed a fairly consecrative approach to R&D, emphasizing reasonably high prospects of commercial feasibilitythis may be one reason Me Japanese have not been noted for creating whole new industries or major new product designs; and (6) Japanese engineers and scientists have not been diverted from commercially oriented
558 DA11lEL I. OKIMOTO R&D to carry on highly specialized research for military and space appli- cations. Like practically all other characteristics of Japan's R&D system, the implications have been neither all positive nor all negative, but a mixture of the two. Aerospace Space research in Japan was begun much later than in the United States or the Soviet Union, the two world leaders, which account for more Man 95 percent of the 2,800 artificial satellites launched to date. Japan's motivation for entering into space research had less to do with national security or international prestige Han with long-range commercial opportu- nities in communications and broadcasting, scientific observation, and, more recently, the construction of space factories for bioengineenng and new matenals. Although Japan takes pride In the fact Hat it became the fours country in the world to launch an artificial satellite (after the Soviet Union, the United States, and France), and although it produces about one-third of the earth station equipment for SAT (~e international telecommuni- cations consortium), its technology in rockets and artificial satellites is ~n- fenor to that of the United States. On He early phases of its space research, Japan had no choice but to borrow rocket technology from the United States; however, win technological ~n- dependence an explicit goal, it did manage to develop its own small N-] and N-E rockets. In 1981 it began development for the H-T rocket, using do- mestically developed technology for second- and third-stage propulsion and for the induction control system. Once tested, He H-l rocket will be capable of launching an artificial satellite of around 550 kg into geostationary orbit. In artificial satellite technology, Japan has followed a familiar pattern of development, moving from overwhelming dependence on foreign licensing to increasing technological independence. To build its CS Sakura satellite (1977), Japan was forced to rely heavily on U.S. technology and U.S. satellite components. However, in building the CS-2 communications satellite (1983), Japan was able to draw on more domestically developed technology; over 60 percent of the components installed were manufactured, assembled, and tested in Japan (Ministry of International Trade and Industry, 1982:6~77~. While Japan is making headway, it is still a long way from complete self- suff~ciency in aerospace technology, and an even longer distance behind He United States. The National Space Development Agency (NASDA) continues to depend on Amencan technology. Several American satellite manufacturers are involved in NASDA projects (e.g., Hughes Aircraft-NEC, General Elec- mic-Toshiba, RCA-NEC), and TRW Inc. has helped with down-range track- ing, software, satellite parts, and systems integration (Davis, 1985:21-28). U.S. manufacturers (e.g., Hughes Aircraft and Ford Aerospace) would like to sell satellites directly to He Japanese market, which would make sense from the standpoint of cost-effectiveness and technological cntena, but so
THE JAPANESE CHAlLENGE IN HIGH TECHNOLOGY 559 far at least, U.S. companies have been unable to do so, owing in part to Japan's long-term goal of achieving self-sufficiency. Although some of He R&D work in consumer electronics conducted by such big, diversified firms as NEC, Toshiba, and other diversified electronics firms can be applied to space technology, the spillovers from consumer electronics to aerospace are not that large. Moreover, it would be unrealistic to think that such limited spiliovers could overcome the serious shortcomings in Japan's overall space effort smog government expenditures, a short history of experience (and early position on the learning curve), an inadequate base of highly skilled R&D manpower in aerospace, bureaucratic conflicts, deficiencies in basic scientific research, and weaknesses in software and systems integration. Here is a field, in short, in which He United States can expect to maintain its dominant position into the foreseeable future. Commercial Jet Aircraft For some of the same reasons as stated above, the same conclusion can be reached win respect to the commercial jet aircraft industry. As Mowery and Rosenberg (1985) point out ~ eir excellent study, Japan's commercial jet aircraft industry seems to be mired in a ping predicament: neither advancing along a fast track toward technological self- suff~ciency as a world-class manufacturer of jet aircraft nor taking He route of establishing specialized niches in world markets. It is following policies that combine contradictory elements of both. The Japanese government has "targeted" aircraft as a national pnority, critical as an end-user industry bringing together many over high technology endeavorsmicroelectronics, new materials, CAD/CAM, and so form. Yet, MA has not sunk sufficient resources into aircraft or followed a sufficiently consistent strategy to turn it into a world-class unduly. Moreover, the aircraft industry is not especially well suited for Japan. Airspace over He Japanese archipelago is narrow, mass transportation by land (especially railways) is extensive, sales volumes are very small, the soaring costs of R&D make the organization of international consortiums for R&D and production ~ncreas- ingly attractive but Japan has had limited experience at organizing such consortiums, Japan maintains a comparatively modest military aircraft ca- pability, and He government has not installed costly testing equipment for manufacturers. Japan also seems ill-prepared to meet the demanding tech- nological requirements, including highly complex designs, CAD/CAM, avionics, and systems integration. For all these reasons, He barriers to new entry have to be considered prohibitively high. Here, in short, is another area in whichdespite government "targeting" Japan is not likely to mount a serious challenge to American dominance. Indeed, if ever Here was an illustration of fallibility in Japanese industrial policy targeting, He commer- cial jet aircraft industry would be He clear-cut example. Regardless of He criticisms that can be leveled at America's so-called m~litary-indus~ial com- --c7 r~
560 DANIEL I. OKIMOTO plex such as cost-effectiveness, limited civilian spillover effects, costly trade-offs in teas of commercial opportunitiesthe system has succeeded in creating dominant aerospace and aircraft industries. Strictly Commercial Orientation It can be argued, on We over hand, that the aerospace and commercial aircraft industries are atypical and Mat Me dominance of military considerations in America's R&D system is, on balance, more of a liability than an asset. In industries in which highly skilled research manpower is finite and in which the technological and commercial spillovers from m~lita~y-oriented R&D are almost nonexistent, as in lasers, the opportunity costs of diverting manpower and resources can be substantial. There is already some concern being expressed in the U.S. laser industry that the "Star Wars" space defense concept could divert such large resources from commercially promising endeavors Mat Japanese companies could slip past the United States unnoticed, concentrating solely on commercial appli- cations and benefiting from MrrI's organization of national research projects in industrial lasers (Conversation with Professor Robert Lo. Byer, Department of Applied Physics, Stanford University, and executives from a leading laser manufacturing company, February 2B, 19851. In having the leeway to pursue purely commercial objectives, Japan may possess an advantage over the United States. It is not easy for the U.S. government to underwrite R&D programs designed to accelerate the devel- opment of key commercial technologies of high cost and uncertainty and of long gestation, no matter how essential they might be deemed for the failure competitiveness of American industry. To secure substantial government funding, key technologies, such as artificial intelligence, usually have to fulfill some kind of military or national security need. The Defense Advanced Research Projects Agency's (DARPA's) project on artificial intelligence is a recent example; compare the fits and starts of DARPA's project with MITI's Fifth Generation Computer project, a cooperative undertaking that seems to be moving Japan's capabilities in artificial intelligence ahead at impressive speeds (Conversation with Professor Edward A. Feigenbaum, Department of Computer Science, Stanford University, January 18, 1985; see also Fei- genbaum and McCorduck (1983~. From the standpoint of R&D cost-effeciiveness and Japanese competition, what may be as much of a problem for Me United States as the domination of military priorities is the "public goods" nature of basic research conducted in Me United States. The outflow of basic knowledge cannot be regulated, even if Japan pays for none of its costs. This suggests that Japan's low rate of investment in basic research and He govemment's relatively modest R&D funding may not really hamper Me coun~y's technological advance (though it obviously hurts some industries, like aerospace and aircraft, more than others).
THE JAPANESE CHALLENGE IN HIGH TECHNOLOGY 561 Japanese National Research Projects To what extent, if at all, do gov- ernment-sponsored R&D projects give Japan a competitive advantage over the United States? No doubt projects like the VLSI have facilitated tech- nological catch-up. But the value of national research projects like that of industrial `'targeting" is often exaggerated. The historical record to date is mixed; there have been some notable successes and several equally note- worthy failures. In the latter category can be counted the 3.75 Computer project (1972-1976), which failed to come up with the operating system it sought, and the Software Development project (197~1981), which produced only a small number of computer-written, applications software packages suitable for commercial sale. Even the heralded VLSI project (197~1980), hailed as an unprecedented model of cooperative research, failed to push Japanese semiconductor tech- nology beyond the frontiers of knowledge (except perhaps for liquid crystal displays). While the VLS! project did advance the state of Japanese semi- conductor knowledge, especially In the area of production technology (e.g., silicon crystal growth and processing), Japanese companies probably would have made such advances eventually anyway. If so, the project's main ac- complishment may have been to hasten the timetable of development, a nontrivial but hardly revolutionary accomplishment. Organizing national research projects is no easy task. Even if all the organizational wrinkles can be ironed out, success is by no means automatic. To obtain useful results, Me technological capabilities of participating firms must be relatively even. One or two fins cannot be too far ahead of the others, or they will not be willing to divulge proprietary information or cooperate in ways that help their competitors close the gap. Over things being equal, the smaller the number of Grins, and the higher the market concentration, We greater the leeway for effective organization. The large number of fimns and wide technological disparities between them help to explain why Japan has had trouble organizing major cooperative projects in biotechnology, pha~maceu~acals, and applications software symbolically and substantively important areas of high technology. If government-sponsored, cooperative research is not as easily organized, nor as uniformly effective, as Americans assume, why has Japan continued relying on it? Indeed, why have national research projects expanded in num- ber and scope? One intriguing answer is that they serve to compensate for structural shortcomings in Japan's capital and labor markets. Structural im- perfecuons, such as the underdevelopment of Japan's equities market, Sax- onhouse (1982) argues, have prompted the government to encourage capital investments through the christening of seminal technologies and industnes. At the company level, high debt-to-equity financing has perhaps caused Japanese management to be more nsk-averse and conservative with respect to R&D decisions (both in terms of money amounts and the uncertainty
562 DANIEL 1. OKIMOTO factor) Man is optimal from We aggregate standpoint of national R&D in- vesunents. Similarly, because the high walls of lifetime employment impede Me diffusion of technology across fins, the Japanese government is forced to step in and facilitate diffusion Trough intercom participation in national research projects. Instead of viewing Japanese national research projects as decisive advantages, therefore, Saxonhouse sees them as necessary ~nstrll- ments of compensation for market imperfections in Japan. If Saxonhouse's assessment is valid, one can infer Mat America's decen- tralized, market-driven system is clearly more efficient (in terms of capital allocation) and arguably more effective (in terms of stimulating technological innovation) than Japan's centralized, state "targeted" system. ~deed, one may question the effectiveness and suitability of Japan's system of industrial targeting for the swiftly changing requirements of high technology. Is public policy better suited to keep pace with the rapidity of commercial and tech- nological change than the invisible hand of the free market? Is the state better at picking winners and losers than Me decentralized marketplace? Can Me state channel capital as neutrally? Japan's system of industrial targeting may have been appropriate for the needs of an earlier era of latecomer catch-up, but is it as effective, now that Japan has reached the frontiers of technology? Lack of a Venture Capital Market A swing difference in Japanese and American patterns of capital allocation for high technology is the lack of a venture capital market in Japan. In the United States, the cumulative total of venture investments, as of 1983, exceeded $7.5 billion, In Japan, by contrast, Me total fell short of $90 million. The availability of venture funds in the United States has had a profound impact not only on the pace of technological progress but also on the evolving structure of high technology industries. It has created enticing incentives for energetic entrepreneurs to convert technological know-how into small start-up companies that offer new or differentiated products on the market. Looking at Me positive effects from an aggregate perspective, the steady stream of new start-ups serves to keep competition brisk, fosters technological ferment, and promotes efficiency in capital allocation for the high-growth sectors. From Me perspective of established firms, however, venture capital can also lead to such dysfunctional side effects as unpredictable personnel turnovers, costly disruptions in R&D plans, escalating salaries for research personnel, diff~culhes in protecting proprietary inflation, and deepening entanglements In litigation concerning intellectual property rights. The actual effectswhether positive or neganve~epend on the stage of an indusmy's life-cycle (~e earlier the stage, Me more positive Me effect) and Me type of companies involved. It can be argued Mat Me lack of a venture capital market, in combination with Me characteristics of Japanese financial and labor markets, has hindered the creation of independent new start-ups. Even without a venture capital
THE JAPANESE CHALLENGE IN HIGH TECHNOLOGY 563 tradition, however, Me small business sector has made important con~ibu- tions to innovation in Japan, particularly in process and production technol- ogy. As Ken'ichi Imai (1984) points out, tile flow of information, degree of R&D cooperation, and synergistic interaction between large parent firms and their many small subcontractors and subsidiaries constitute one of Me great strengths of lapan's industrial organization. A number of innovations- mostly in process and production technology but also in some new-product designs have emerged out of the structure of such vertical relationships. Compared with the dynamism of independent, small firms in, say, Silicon Valley, however, lapan's small enterprise sector has not functioned as a fertile seedbed for technological iMovanon. A National Science Foundation (1976) study found that the number of innovations made by small and me- dium-sized Grins In five counmes was lowest In Japan. Research activity in small companies tends, on the whole, to be very limited. In 1981, only ~ percent of the small firms reported Mat they even engaged in R&D, compared with 56 percent for large companies (Organisation for Economic Cooperation and Development, 1982:1991. Small films in Japan labor under some sig- nificant handicaps, including a systemic bias in favor of big business, higher costs of capital, weaker drawing power for recruitment of topnotch re- searchers, lower prestige, and higher bankruptcy rates. It is no wonder they have not contributed as much to innovation as Weir counterparts in other countnes. The lack of a venture capital market is attributable to a variety of factors. There appears to be more than enough money for the creation of a large venture capital market; Hambrecht and Quist, for example, recently created a venture fund In Japan with Sanwa Bank and Oriental Leasing. But the problem is Mat lapan's stock market is not designed to handle small start- up companies. Without an entry vehicle into Me stock market, venture in- vesunents cannot be liquidated, and venture capitalists cannot cash in on early-round financing. More ~mportan~dy, the preference of the top college graduates for permanent employment at established corporations and Me near impossibility of lateral reentry once an employee has left make it difficult for new start-ups to attract the best R&D talent from big corporations; most appear to prefer job security to Me allure of making persona] fortunes. Unless labor patterns change, therefore, the availability of venture capital will not have Me same far-reaching impact In Japan Mat it has had in America. Small, independent firms in Japan will not have Me luxury of operating in an environment as conducive to innovative dynamism as small companies in He United States. CONCLUSIONS: MUST JAPAN INNOVATE? Whatever one's view of Japan's technological future, there seems to be general agreement about He type of technologies in which He Japanese have
~ is O~~ excelled ~ dam These ~ bs~ below, gong Bib He condoning U.S. singes: Jesse S=n~s opted ese~b ~ deve~pmem . . Comet 1~=veme~ Co_i~ Prices Recess ad pan Cola Comae 6 Apologies Quit coat On Sot, As Was O.S. S=n~s B=ic mse~b B=^ughs ad invendoDs Milan ~plic~ons New-~cl design Syrups i--on Sat Less percale Recedes New ~nchon~ides New Ibis posits Cusm~on ad act ~ Be misaim notion be conveyed ~~ obese technological ch~dsUcs eve out of idle noons Abuts, it should be pointed out ~m Me Was of coffee sand ~ condnu~ly evolving, as high ecology Asks ~~ ~oc_s_. ~= ~ ~ _ss~ Was as sag, while America seems ~ be paint ma abandon ~ process and pnoducdoD technology. One should n~ Beard cunanl sawn ~ s and weaknesses, George' as em fixed or accurate indicate of abed Be go coaxes seem hewed. Technology is moving ~ avidly ~ d~- bound genendizadons lo be valid far verylong. Even Japan's instludon~ struck is teeing a~nsfbnned. As ibis happens, Be cou~'s Craig ~ innovate c_ot help but be mad For ex~e, e~c~oD~ =~ Is new Be cop of Be list of ~~ Divisor N~=one's Dada of Clay phoddes. Despite b~=c~c Bead inEgbOng, ad h- gid~es, J~='s educing system is undoing chase. ~~ resoles berg tunneled ~m basic Web, lids Bonn unive~ides ad inmost ~ beg expanded, ales governing personnel ~ becoming ma Hexible, ad He calculi is bear mvised. Jewess lever maize ~~ ~ ^- c~on~ system is bay in need of In. Me Anacin sector, Aver Fact ~a, is also change Aids, gig -among inflicts ~ J~ese inmost ~=i~. As Mined om emit, He demon ~ who Bed lo be one of He mod dray egged ~ insulted Anacin systems in He wodd is band ~ ~= He seasoned bring s~c~, He "b~Dg-indus~ complex," poems of cam menacing' ~~ or b~-c=~=d Is Coupes, ad He willingness me War R&D risks. Fin~ci~ Ague ma Bus loosed He gap of none ovations OD J~='s economy, Being maw forces play a Aver role. 0~ cages under w~ include: (1) lands toga He "spinning On' of mom R&D Divides mom lie paw Is ~ smog sun ad
THE JAPANESE CHALLE!JGE IN HIGH TECHNOLOGY 565 subsidiaries; (2) the creation of small subsidiaries designed to serve as the functional equivalent of small venture start-ups; (3) the appearance of the rudiments of a venture capital market; (4) signs of at least some shift toward greater labor mobility; (5) some erosion in the government's power to in- tervene in the market; (6) changes in the substance and role of Japanese industrial policy for high technology; and (7) greater attention to basic and precommercial, prototype research. As win financial deregulation, these changes could have the effect of freeing up market forces. They might also create a research environment Rat is more conducive to the kind of bold, new-product designs and state-of-the-art breakthroughs for which Japan has not heretofore been known. In short, some of the rigidities of Japan's old R&D systemone geared to latecomer catch-up in the heavy manufacturing sectorsmay be giving way to an adaptable R&D system that is more suitable to meeting We func- tional needs of high technology. Whether this heralds the onset of a new era of technological originality is still too early to tell. The evidence is too mixed to make simple projections possible. To this point, this discussion has assumed that in order to compete in high technology, Japan will have to find ways of being more innovative. But is this assumption valid? Must Japan innovate? Why would it not be possible for Japan to continue doing exactly what it has done so successfully in the past: namely, follow a conservative, second-to-market strategy, letting Amer- ica pay the high costs and take the risks of developing new industries and markets? Why not simply continue concentrating on the less glamorous but commercially more decisive areas of process and production technology and mass marketing? Is not the history of technology replete win examples of inventors being soundly thrashed in the commercial marketplace by tech- nological second-comers? There is no doubt that a distinction needs to be drawn between technological innovation and commercial success. The two are not necessarily linked. The first is no guarantee of the second. Nevertheless, because high technology sectors have steep learning curves and comparatively short product life cycles, the advantages of being first-to-market can be worm far more than the costs and risks of early investment. First-comers can secure dominant market share, win brand name recognition, move down the learning curve, raise the bamers to entry and, in some cases, push second-comers right out of existence. Relying on foreign technology, as an alternative to domestic innovation, can leave companies at the mercy of foreign firms, which may or may not be willing to grant licenses in return for royalty payments. If patent holders believe Hey can gain more Man they lose by withholding basic patents, Japanese second-to-market firms could find themselves closed out of bur- geoning markets. Moreover, Japanese companies must also accept the reality of attempts by the U.S. government to impose restrictions on Me international
566 DANIEL I. OKIMOTO transfer of militarily sensitive, dual-purpose technologies. What would Jam anese firms do if there was a groundswell of technological nationalism that restricted Weir access to foreign know-how? Of course, for We latter, worst-case scenario to materialize, the interna- tional situation would probably have to deteriorate. Even then, it would be hard to shut off Me flow of knowledge completely. In this era of high technology, We United States and Japan have common interests in keeping Me transfer of technology open. The volume of technology transfer across Me Pacific, including licensing, second-sourcing, original equipment man- ufacturer (OEM) agreements, and cross-licensing, is greater today than ever before. Nevertheless, even under an open international regime, Japanese companies would have to have Weir own technology in order to cross-~cense foreign know-how. In ~nterna~aonal technology transfers, more bartering seems to be taking place, a manifestation perhaps of Me mounting costs and value of innovation. To obtain foreign technology, therefore, the Japanese believe they must develop their own In order to obtain something of comparable value from abroad. If their perception is correct, it means that they must be able to innovate. Perhaps the most compelling reason why Japan needs to innovate is because the rapidly developing countries in Asia are moving quickly up the ladder of manufactunug value-added into, for example, the low end of consumer electronics. As South Korea, Taiwan, and Singapore develop the ~nfra- structure to mass-produce consumer electronics products, Japanese producers will have no choice (short of protectionist recourse) but to scramble up the ladder of value-added. They will have to move, for example, from consumer to industnal electronics, from hardware to software, from components to integrated systems. As "min,-Iapans" spring up all around it, Me only way of staying ahead will be to accelerate the pace of R&D. For all these reasons, therefore, Japan may have to embark on a crash program to expand and upgrade its infrastn~cture in science and technology so that it can innovate. Japan can no longer exploit the advantages of late- comer status. It may not be able to follow a low-cost, low-risk strategy of second-to-market by capitalizing on low production costs. As comparative advantage shifts to the newly industrializing states, Japan will have to com- pete head to head with Me United States in what has been the traditional U.S. stronghold, high technology. The challenge facing Japan will almost certainly be harder than the past challenge of industrial catch-up in the smokestack sectors. Known for their adaptability, however, the Japanese are Lying hard to overcome some of Me old institutional constraints that have impeded innovation in the past. Whether they succeed remains to be seen. Notwithstanding the image of infallibility, success is by no means assured. ButJapan's formidable storehouse of s~aeng~s, combined with Me changes Rat are now taking place in its old R&D system, suggest that it certainly would be foolhardy to count Japan out.
THE JAPANESE CHALLENGE IN HIGH TECHNOLOGY REFERENCES 567 Davis, Neil W. 1985. Japanese space activities: Taking off in a big way. The ACCJ Journal 22(3): 21-28. Feigenbau~n, Edward A., and Pamela McCorduck. 1983. TkeFzfth Generation: Artificial Intelligence and Japan's Computer Challenge to the World. Menlo Park, Calif.: Addison-Wesley. Imai, Ken' ichi. 1984. Japanese Industrial Policy for High Technology. Unpublished paper, prepared for Northeast Asia-United States Forum on International Policy, Stanford University. Ministry of International Trade and Industry. 1982. Japan Science and Technology Outlook. Tokyo: Fuji Corporation. Montani, h£asanori. 1982. Japanese Technology. Tokyo: The Simul Press. Mowery, David C., and Nathan Rosenberg. 1985. The Japanese Commercial Aircrew Industry Since 1945: Government Policy, Techzical Development, and Industrial Structure. Stanford: Northeast Asia-United States Forum on International Policy, Stanford University. National Science Foundation. 1976. indicators of lntern~onaZ Trends in Technological Innovation. Washington, D.C. Okirnoto, Daniel I. 1983. Pioneer and Pursuer: The Role of the State in the Evolution of the Japanese and American Semiconductor Industries. Occasional Paper of He Northeast Asia-United States Forum on International Policy. Calif.: Stanford University. Okimoto, Daniel I. Forthcoming. Between Mm and the Market: Japanese Industru~l Policy for High Technology. Calif.: Stanford University Press. Okimoto, Daniel I., Takuo Sugano, and F~anldin B. Weinstein. 1984. Competitive Edge: The Semiconductor Industry in the U.S. and Japan. Calif.: Stanford University Press. Organisation for Economic Cooperation and Development. 1982. Innovanon in Small and Medium Firms. Pans. Ozawa, Terutomo. 1974. Japan's Technological Challenge to the West, 1950-1974. Casubndge, Mass.: MIT Press. President's Compression on Industrial Competitiveness. 1985. Global Competition: The New Reality, vol. 1. Washington, D.C.: U.S. Government Printing Office. Saxonhouse, Gary R. 1982. Japanese High Technology, Govemment Policy, and Evolving Com- paranve Advantage in Goods and Services. Unpublished paper, prepared for Japanese Political Economy Research Conference, Honolulu, Hawaii. Semiconductor Industry Association. 1983. The Effect of Government Targeting on World Serru- conductor Compennon. Cupertino, Calif.
