The Japanese Firm May be Becoming too Rigid for Information-Sharing in the Digital Age

Masahiko AOKI (Professor of Economics, Stanford University)

It has became a sort of fashion in Japan to refer to the 1990s as a "lost decade" because of its poor economic performance. In my view, however, the problem of the 1990s was not created simply by a mistake in macro-economic management, rigid bureaucratic regulations, lack of political leadership, moral hazard behavior of financial institutions and large borrowers, and so on, that happened only in that decade. For example, the financial crisis in the 1990s was the inevitable culmination of the failure of the financial sector to have adapted to changing international and technological environments already visible in the 1980s. In that sense, the lost decade rather started in the 1980s. Rather, in the 1990s the Japanese political-economy may be thought to have finally started its overdue efforts for changing the institutional arrangements, albeit only slowly and often in an invisible way. This institutional framework had worked well in several decades of economic growth in the twentieth century but have been rapidly becoming misfits with the emergent international, demographic, and technological environments. The symptom of a change may be seen as latent in the outcome of the recent election, but also in various spheres of micro-economy, including corporate organization domains. In this small commentary, I would like to point to some important organizational agenda of Japanese firms from historical and international perspectives. Specifically I would like to focus on the important impacts of IT revolution on firm's organization, by referring to the so-called Silicon Valley phenomena as a frame of comparison.

In the 1980s the Japanese corporate firms in the export industry were considered to enjoy a considerable competitive edge. Some of those firms, as represented by Toyota and Sony, still maintains competitiveness in global markets, but the general perception prevailing around the turn of the century is that the Japanese firms are somehow failing to sustain the past luster. Why? What reform is needed for Japanese firms to regain vigor in the age of IT revolution and the globalization of markets?

Let us recall that in the 1950s and 60s it was the large integrated American firms, such as GM, IBM, US Steel, that dominated industries. Those firms were organizationally characterized as hierarchically integrated. For example, in development of new models, whether they were Cadillac, IBM 360, their conceptual design was initiated centrally, and then it was gradually decomposed as analytical design, process design, testing, manufacturing, and so on. Some scholars describe this kind of development process as the "water-fall model."

However, as product systems got more complex, it became increasingly harder to foresee all possible engineering problems from the beginning. However, in the water fall model, information feedback from downstream design and manufacturing processes to upstream processes was inherently limited and as a result the problem of quality-control became increasingly difficult. It is at this point (some time around the late 1970s or the beginning of the 1980s) that the Japanese firms made an inroad into global markets, aided by their own organizational innovation. Globally competitive Japanese firms structured their organizations in such a way that information fed-back from lower levels to upper levels, as well as horizontal information sharing across different task units, was utilized effectively for fine-coordination of their internal activities. Best examples of these practices were the so-called kanban system, development teams comprising not only development engineers, but also manufacturing and marketing representatives as well as engineers from suppliers. This type of organization may be referred to as the "horizontal hierarchy".

Then, in the 1980s and 90s the so-called Silicon Valley phenomena became emergent and started to exhibit a still new model of industrial organization. In my opinion, the essence of the Silicon Valley phenomena can not be fully appreciated only by seeking its innovativeness in terms of risk-capital supply by venture capitalists and the incentive effect of IPO (initial public offerings). In my opinion, its innovative nature lies more in its information systemic characteristics viewed as a system comprised of a cluster of entrepreneurial firms and various intermediary-services including venture capitalists, lawyers, accountants, university researchers, etc. This system works in the following way distinct from the water fall model or horizontal hierarchy. Instead of centrally conceiving the design of a new technological product system within a single organization, new start-up firms are decentrally engaged, and compete with each other, in the development of products in niche markets that may potentially form a module part of new technological product systems (such as internet). Innovative product systems can be formed evolutionarily by combining those module products developed by successful firms. Here the contrast between "centralized design" and "evolutionary formation" is important. Venture capitalists and leading firms in niche markets (e.g., Intel, Cisco Systems, Microsoft) may play a very important role in drawing a road map for a potential system as a guide for entrepreneurial firms and mediating information necessary and useful for realizing it. But they do not dictate the design of new product systems centrally by themselves. The leading firms grow by acquiring and combining element technology developed by entrepreneurial firms [Thus instead of in-house R&D, people talked about A&D (acquisition and development)]. By this way, the Silicon Valley model has achieved a remarkable innovative capacity. This innovation has had tremendous impacts on the organizational landscape of American economy in general. Even large firms, particularly better- run ones such as GE, are not strictly run centrally any more in operational domains, but their headquarters concentrate their functions more in flexibly de-coupling and re-coupling business units, while spinning off operational activities as much as possible outside. [See more for this my article entitled "Information and Governance in the Silicon Valley Model," to be published in X.Vives (ed.), Corporate Governance: Theoretical and Empirical Perspectives, Cambridge University Press, forthcoming in 2000. This is available from http://www-econ.stanford.edu/faculty/workp/swp99028.html

Thus, the organizational innovation that has been occurring in the US has made an once advantageous characteristic of the Japanese firms into somewhat disadvantageous one. Intense information sharing within an organization and emphasis on firm-specific expertises, and so on are becoming relatively more costly in the age of digital revolution. Under the digital revolution even small units (such as entrepreneurial start-up firms in Silicon Valley) can enlarge its scope of information processing and can be engaged in integrative module-product development and design. Also, if interfaces and protocols are standardized, module products from different firms can be flexibly combined to form a new innovation system. By this way, the speed of innovation can become fast. In contrast, the Japanese firms may be becoming too large for information-sharing based on face-to face communications in the age of IT revolution and unnecessarily too closed in their attempt to develop a new innovative system all in-house. This observation may apply more to emergent industries such as internet but not so much to industries where constituent activities are closely correlated and complementary (e.g., auto industry). But it may still suggest some generic problem that the Japanese firms are facing. That is, the Japanese firms may need to restructure themselves in such a way that information processing and decision process can be made quicker, and initiatives for new development may be made more decentralized.

Obviously, I am not proposing that Japanese firms should adopt the Silicon Valley model as it is. I have referred to the Silicon Valley model only in order to understand a generic nature of informationally-efficient and innovative industrial organization in the age of IT revolution. The Japanese firms may be able to, and ought to, realize the generic organizational innovation in their own path-dependent way. How can it be possible?

Considering that competence for drawing a road map for potentially innovative systems are still largely residing in established firms in Japan, there may be a hint for one possible way of reforming them, although it cannot be almighty. That is to utilize the recently-deregulated holding-company framework. There are many reasons why the holding company may be worth consideration for more wide adoption, among which the possibility of separating strategic decision making (map-drawing so to speak) and operational business activities can be an important one (also employment contracts and personnel administration may be diversified depending on business lines). By this way, economically variable face-to-face information-sharing may be encapsulated within various subunits and effectively utilized within each unit for strategic decision, development design, manufacturing, etc., without interfering each other's decision and continual innovation. Then Japanese firms may become more agile in operation and flexible in restructuring.

It is not advantageous, however, for firms to adopt the holding company scheme without the introduction of a consolidated tax scheme. It is highly desirable for the government to introduce a consolidation tax scheme as soon as possible to make Japanese firms and industrial structure more flexible.

Reference for the holding company scheme and its implications: M.Aoki; Information, Corporate Governance, and Institutional Diversity: Competitiveness in Japan, US and Transitional Economies, Chapter 7, Oxford University Press, forthcoming in Fall 2000.

Also, a book dealing with the subject matters of this commentary in a more systemic and academic way will be published in 2001 from MIT Press (in English), NTT Publishers (in Japanese) and Shanghai Publishers (in Chinese) under title, Toward a Comparative Institutional Analysis.