Haru ichiban (literally "spring number one"), is a strong, southern wind that first blows sometime between Risshun (February 4, the first day of spring on the lunar calendar) and Shunbun (March 20, the vernal equinox). Haru ichiban signals a loosening of winter's icy grip and is a harbinger of spring. This year, haru ichiban blew into Tokyo on February 23. And with warmer days ahead, our spirits naturally rise.
As Japan emerges from its long economic chill, innovation holds the key to sustainable growth. The technological knowledge required to develop new products is becoming increasingly specialized and complex, and product life cycles are becoming shorter. Accordingly, it is necessary for companies to innovate and develop new product systems by integrating their specialized knowledge. They must also develop a system of coordination that allows them to apply such knowledge promptly. On February 14, 2005, RIETI held a Policy Symposium, "Japan's Innovation System: Its Strengths and Weaknesses." The symposium analyzed the strengths and weaknesses of Japanese industrial innovation, focusing on the linkage between science, technology and the creation of new products; industrial clusters and industry-university tie-ups; software platform theory in the information-based home appliance industry; and product architecture theory in manufacturing. Prior to the symposium, we spoke with Faculty Fellow Hiroyuki Chuma about the organizational limits of Japan's semiconductor industry and what sort of specialized and integrated knowledge structure it should strive to build.
RIETI Faculty Fellow Hiroyuki Chuma has been a professor at Hitotsubashi University's Institute of Innovation Research since 1999. Prior to assuming his current post, Dr. Chuma was assistant professor at the University of Southern Illinois, associate professor of economics at Tokyo Metropolitan University and visiting professor at Yale University. An expert on labor economics and the role of skilled workers in the innovation process, his recent research focuses on sources of international competitiveness in manufacturing industries.
The author of numerous scholarly papers and several books, Dr. Chuma's publications in English include "Permanent Labour and Land Tenancy Contracts in Agrarian Economies: An Integrated Analysis," (co-author), Economica, 1993; and "Why Do People Demand Life Insurance Savings?" Economic Studies Quarterly (now Japanese Economic Review), 1990. He holds a Ph.D. in Economics from the State University of New York at Buffalo and a B.A. in Economics from Hitotsubashi University.
Interview
RIETI Report: How has the international competitiveness of the Japanese semiconductor industry changed through the years?
Chuma: Generally speaking, the golden age of Japan's semiconductor industry was the roughly 10-year period from the signing of the Japan-U.S. Semiconductor Agreement in 1986. In the 1980s, when dynamic random access memory (DRAM) chips were very important among the various semiconductor devices, Japan was one of the world's top semiconductor producers. Japan held a roughly 90% share of the global DRAM market in the mid-1980s, but by the early '90s its share had begun to decline. In 1990, Japan accounted for nearly 50% of the world's semiconductor shipments, but this figure has continued to decline and since 2000 it has fallen below 30%.
Industry pundits often use the phrase "technology driver," meaning a semiconductor device that drives the global semiconductor market. Since the late 1990s, Japanese semiconductor makers have been unable to come up with such technology drivers. As I mentioned earlier, DRAMs were a very important semiconductor device during the 1980s. In fact, other semiconductor devices such as central processing units (CPUs) and microprocessor units (MPUs) were largely dependent on the technological innovations in DRAMs. However, in the 1990s we entered the age of the personal computer and DRAM prices crashed from 1995 to 1996. If one takes 1995 prices as a base year and sets an index of 100, by the following year the index had fallen to 30. This price crash led to a further decline in the competitiveness of the Japanese semiconductor industry.
RIETI Report: Against this backdrop, how do you assess the current situation?
Chuma: I think the most serious issue is that the Japanese semiconductor industry has lost its ability to come up with world-leading technology drivers. Although Japanese semiconductor makers have reaped huge profits in such areas as flash memory and charge coupled devices (CCDs), for a long time they have been unable to take the lead in high value-added products such as CPUs and field programmable gate arrays (FPGAs).
One theme of my research is to study why this has happened. I have been looking into the problems faced by the Japanese semiconductor industry in three areas - research and development, production technology and manufacturing - mainly through field research, with the cooperation of various people in the industry. What I have found, in short, is that the industry's competitiveness appears to be weakening in all three areas.
RIETI Report: What is behind the decline in competitiveness?
Chuma: One major factor is that there are serious limits to organizational management when an industry is faced with a rapid rise in the complexity of both the market and technology. In other words, companies are facing the "limits of organization." Of course, it is extremely important for semiconductor device makers to understand the potential requirements of customers and incorporate them into device functions. The ability to read the market is thus critical. However, the market has become extremely diversified and specialized due to globalization and economic growth. In other words, the market has become highly complex, and its degree of complexity is increasing rapidly. In order to accurately understand the changes in such a complex market, wide-ranging analysis of market trends and developments is indispensable but, relatively speaking, Japanese semiconductor makers have a problem interpreting the market. This is one factor behind the decline in their competitiveness.
In addition, once customers' potential requirements are incorporated into the functions of a semiconductor device, one must consider how to design the device itself. It is necessary to determine sort of technology to use, and it is therefore important to be able to accurately interpret key trends and developments in technology. And in order to select the best technology, one has to choose the proper process technology to actually put the technology to use. However, in the semiconductor industry, where the pace of innovation is extremely fast, the complexity of such technologies is increasing as rapidly as that of the market, if not more so. I think another factor behind the decline in competitiveness is that organizations are not being managed in a way that properly deals with this technological complexity.
Semiconductor makers are facing increased complexity in both the market and in technology. Unless companies deal with these issues properly, they will not be able to provide high value-added devices for the market at a sufficiently profitable price in a timely fashion. Japanese semiconductor makers appear to be weak in both these areas because of limits in their organizational management.
RIETI Report: Can you be more specific about the increasing complexity of the market?
Chuma: Even in the 1980s, Japan was not providing platform-type technology, at least in terms of semiconductor devices. In fact, the breadwinner for Japanese semiconductor manufacturers in the 1980s was DRAMs manufactured for IBM or IBM-compatible mainframes. In that sense, they were making semiconductor devices that met IBM's specifications. However, in the 1990s, DRAMs came to be increasingly used in personal computers and home appliances rather than in mainframes and mini-computers (workstations). In other words, whereas in the past companies only had to monitor the businesses of mainframe and mini-computer manufacturers to understand market trends, the market became increasingly complex from the 1990s, and even in the case of DRAMs it no longer sufficed to just keep an eye on Intel. For example, it has become more important for companies to think about such questions as how the functions of DRAMs and CPUs will be separated in the future, what the main use of personal computers themselves will be, and what the main use of various home appliances and telecommunications devices such as routers will be - and then to accurately determine where they should position themselves and what sort of market and technology strategy is needed. Japanese semiconductor makers have been losing their competitiveness as they have failed to meet these new challenges.
RIETI Report: What about a company such as ASML (see note), which you cite in your paper?
Chuma: ASML is a good example of a company that has adjusted to this new environment. In the case of highly complex and advanced devices that bring together the best that science can offer, such as the microlithography industry, the limits of what is called the linear model in industrial organization are increasingly apparent. With such devices, it is important to identify the architecture and the manufacturing process simultaneously. So companies must change how they bring knowledge together. Integrated knowledge spanning 10 to 20 subjects is necessary for such a complex process. Companies must continuously go beyond corporate boundaries in seeking expertise. ASML has been successful in managing this process by using effective communication tools.
In my paper I point to the importance of a modular architecture as one tool for such communication. Such architecture need not be perfect from the outset. The knowledge of various experts can be amassed to improve it and make it into a more suitable modular architecture later. However, in order to bring together the knowledge of various experts in a process of trial-and-error, an effective modular architecture must be created to enjoy the benefit of "understanding at glance."
RIETI Report: Are there any efforts among Japanese semiconductor manufacturers to reform their organizations based on these considerations?
Chuma: Unfortunately, such efforts are not common. However, there are some examples. One company that has made changes based on this model is Hiroshima Elpida Memory Inc. (the manufacturing firm of Elpida Memory Inc.). Interference among manufacturing processes is becoming very serious at the state-of-the-art semiconductor plants: A minor glitch in process "A" can lead to serious problems in process "B" or "C." Preventing such adverse effects and solving problems that actually arise depends largely on how promptly the people closest to the actual manufacturing process - operators and engineers - become aware of "changes and abnormalities." It is quite ironic that in semiconductors fabs the more automated production systems advance, the more critical role work incentives tend to play because of the intense interference between systems and incentives. This is because the tolerances in the devices and processes that make up the production system are rapidly shrinking in line with the more miniaturized and multifunctional requirements for semiconductor devices. The more repetitive and patterned the tasks to be conducted are, the greater the tendency for them to be automated; and the difficulty of tasks dependent on human intelligence grows in inverse proportion. As a result, fabrication engineers, whose main task is to establish a mass-production system, will increasingly be called upon to consider the incentives of workers as well as the technology involved. In other words, they need to have knowledge not only of natural sciences but also social sciences. Hiroshima Elpida Memory has developed a unique semiconductor production system that allows it to adapt to the changing business environment; it now closely resembles Toyota's manufacturing system in the sense that it skillfully manages the organization so that local optimization leads to global optimization. Companies such as Hiroshima Elpida Memory, which manage knowledge effectively, enjoy higher productivity than their competitors.
RIETI Report: What are the implications of your research for policymakers?
Chuma: I wonder. The limits of organization faced by the present administrative system are even more serious than those faced among the private sector companies we have been discussing! As things stand, I believe it will be very difficult for administrators to present policy options that take the integration of global and local optimization into consideration. To make such an outcome more likely there will have to be a more comprehensive overhaul of the administrative system, including reform of the civil service system. I would like more policymakers to understand the basic structure I have been discussing here so that they can see the need for such reforms more clearly. After that, they may be able to establish a legal system that makes it easier to bring knowledge together and to present policy options that will help Japanese industries.
Note) ASML Holdings N.V. is a Dutch microlithography company.
Interview conducted by Toko Tanimoto, chief editor of the RIETI website (February 8, 2005).
Handouts for the RIETI Symposium "Japan's Innovation System: Its Strengths and Weaknesses" are available at http://www.rieti.go.jp/en/events/05021401/report.html.
To view Professor Chuma's paper, "Increasing Complexity and Limits of Organization in the Microlithography Industry: Implications for Japanese Science-based Industries," go to http://www.rieti.go.jp/en/events/05021401/pdf/4-2_chuma.pdf [PDF:724KB].
For your reference
Hot Issues: "Modularity and Management"
http://www.rieti.go.jp/en/special/hot-issues/categories/modularity.html
*******
Brown Bag Lunch Seminars
Coming Soon
==========
All BBLs run 12:15 - 13:45, unless otherwise stated.
3/01 KAWAKAMI Naotaka (Director for Policy and Communications Division, Planning and Coordination Bureau, Financial Services Agency)
"Program for Further Financial Reform" (in Japanese)
3/03 YAMAGUCHI Mitsutsune (Professor, Faculty of Economics, Keio University)
"The Post-Kyoto Framework" (in Japanese)
3/15 FUKAO Kyoji (Faculty Fellow, RIETI/Professor, Institute of Economic Research, Hitotsubashi University)
Title: TBA (in Japanese)
3/25 Jacques BOURGEOIS (Partner, Akin Gump Strauss Hauer & Feld LPP)
Marco C.E.J. BRONCKERS (Partner, Wilmer Cutler Pickering Hale and Dorr LPP)
Title:TBA
For a complete list of past and upcoming BBL Seminars, see http://www.rieti.go.jp/en/events/bbl/index.html
Fellow titles and links in the text are as of the date of publication.
For questions or comments regarding RIETI Report, please contact the editors.
RIETI Report is published monthly.
