The Competitiveness of Nations
in a Global Knowledge-Based Economy
April 2003
Nathan Rosenberg; W. Edward Steinmueller
*
Why are Americans Such Poor Imitators?
The American Economic Review
Volume 78, Issue 2,
May 1988, 229-234.
Despite American success in
previous historical eras in imitating the technology and organizational
structure of our industrial rivals in other nations, there is mounting evidence
that our capacity to absorb and adapt our rivals’ advantages to our own purposes
has diminished in recent years. While
these concerns are voiced with regard to a number of nations, the recent
success of Japanese firms has been noteworthy and deserves special attention.
One reason Americans have
been such poor imitators is that, until very recently, we were not even aware
that there was much in Japanese industry that was worth imitating. Japanese economic competitiveness was, for a
long time, dismissed as simply reflecting lower labor costs, which were
regarded as decisive in certain industries. Later, Japanese success was dismissed as
ephemeral, reflecting the ease of rapid growth on the part of a “mere imitator”
following the innovative leads of other nations, particularly our own.
More recently, as
competition has become more heated and as certain American industries have
suffered heavily from Japanese imports, the successes of Japanese firms have
been attributed to policies of “industrial targeting” orchestrated by the
Ministry of International Trade and Industry (MITI), usually said to involve
extensive government subsidies and coordination of import policies that
unfairly tilted what should have been a level playing field.
We do not wish to deny that
there may have been some truth to each of these beliefs at one point in time. However, an unfortunate consequence of such
beliefs has been that they have delayed efforts to monitor
*Fairleigh S. Dickinson, Jr.
Professor of Public Policy, Department of Economics, Stanford University,
Stanford, CA 94305, and Deputy Director, Center for Economic Policy Research,
100 Enema Commons, Stanford University, Stanford, CA, 94305, respectively.
229
and study the performance of the Japanese
manufacturing sector with any care.
Certainly an earlier
complacency has now unravelled. It is abundantly clear that there is much to
admire, and perhaps to emulate, in some parts of the Japanese manufacturing
system, particularly in the production of goods with a high degree of systemic
complexity - for example, plain paper copiers, automobiles, some machine tools,
and consumer electronics. In retrospect
it is obvious that there has been much that has been worth imitating, but
Americans, even when they have become aware of this, have been poor imitators. Why should this have been so? Why have the Japanese been so much better at
imitation than the Americans? What has
made the Japanese, if we may be permitted to use the phrase, such “creative imitators?”
During the past twenty
years, the composition of Japanese exports to the United States has shifted
dramatically away from industries where labor intensity provided comparative
advantage, to industries where sophisticated manufacturing skills and technology
are of central importance. This shift
has been so marked that, in 1985, over 80 percent of Japanese exports to the
United States were in electrical, electronic, transportation equipment, and
machinery sectors. In short, Japanese
“imitation” has been concentrated in a few specific sectors where American
industry had previously been dominant for many years following World War II. The imitative processes that were used by
Japanese industry in catching up to U.S. firms in these sectors should now be
recognized as having major implications for how Japanese firms are prepared to
succeed at the more difficult task of forging ahead (Moses Abramovitz,
1986).
The first part of our
answer to the question of why Americans have been such poor imitators is that
there has been a distinct asymmetry in the strengths developed by each of these
industrial economies. This asymmetry
partly accounts for why it has taken so long to appreciate fully the sources of
Japanese industrial capabilities. The
Japanese have been very successful in borrowing and developing technologies
initially created by American firms. These
technologies have been largely of a hardware nature, in particular, a stream of
highly visible product innovations.
By contrast, what may be
most worth imitating on the Japanese side is much more subtle and much less
visible. It includes ways in which
certain activities are carried out, rather than readily identifiable
pieces of hardware. These differences
lie at the levels of organization and incentives for improvement. The first is the efficient coordination of
product design and manufacturing functions. The second is effective solutions to the
myriad small problems that are key to efficient mass production techniques. An important part of the reason that it has
been so difficult to appreciate the nature of these Japanese achievements is
that they are heavily concentrated in a collection of activities - development -
that economists have, so far, failed to unpack and subject to detailed and
critical analysis.
This is surprising for a
number of reasons, not the least of which is that R&D is, in fact, overwhelmingly D. Yet, we know more
about the 12 percent of R&D that
constitutes basic research than of the 68 percent that constitutes development.
While this may be understandable on the
part of natural scientists, it is less so on the part of economists. Nevertheless, American thinking about the
innovation process has focused excessively upon the earliest stages - the kinds
of new products or technologies that occasionally emerge out of basic research,
the creative leaps that sometimes establish entirely new product lines, the
activities of the “upstream” inventor or scientist rather than the “downstream”
engineer. American discussions of
technical change are more likely to be presented in terms of major innovations
and pioneering firms, rather than in terms of the success of particular sectors
or firms at catching up and overtaking other organizations through sustained
effort and small improvements. In this
respect, the dominant view of the innovative process is still overly
Schumpeterian, in its preoccupation with discontinuities and creative
destruction, and its neglect of the cumulative power of numerous small,
incremental changes. We suggest that the
Japanese have had a much deeper appreciation of the economic significance of
these vital development activities than their American counterparts.
Development, of course,
covers a range of activities whose content differs widely from one industry to
another. It generally includes the
designing of new products, testing and evaluating their performance (which in
some industries may involve the building and testing of prototypes, or
experimentation with pilot plants), and inventing and designing new and
appropriate manufacturing processes. In
each of these activities, the role of minor modifications and small improvements
that better integrate design and production, establish closer feedbacks from
users to suppliers, and more effectively “tune” existing production methods,
are critically important. Individually,
each of these modifications and improvements will bring about some slight
reduction in cost or improvement in performance. Their cumulative effects may, however, be
immense, as when the semiconductor industry moves, through a multitude of small
steps, from a handful of transistors on a chip to a million such transistors,
or when the channel capacity of a 3/8” coaxial cable expands, through a
succession of small improvements, by more than an order of magnitude, or when
the speed of computers increases by several orders of magnitude.
It is the essence of these
development activities that they have no well-defined terminus. They do not end when a new or improved product
is brought to market. Quite the
contrary. A continual stream of small
improvements is often the essence of success in the competitive process. In industries such as those that currently
account for the bulk of Japanese exports to the U.S., development is a never-ending
activity. They are not, from some points of view, very exciting activities. They are activities that do not win Nobel
Prizes; nor, for the most part,
230
do they even win recognition at the Patent Office. This low visibility accounts for the very
limited awareness of their economic importance. Nevertheless, poor performance in the development
process can be commercially fatal to firms that are highly successful at
research. Such poor performance can
readily translate into final products of inferior design, lower quality, and
poor reliability. It can also translate
into higher cost and, therefore, inability to sustain a market position originally
achieved through the innovation process. These shortfalls can convert a technological
head start, resulting from successful innovation, into a scramble to retain
what turns out to be a shrinking market share against the cost and performance
advantages of competitors, including those who may have had no role in the
initial innovation or in the antecedent research that made it possible.
These possibilities are, of
course, not offered as mere idle rumination. There is an accumulation of evidence that many
Japanese successes in recent years are a consequence of greater effectiveness
in organizing and providing strong incentives for these “downstream”
development activities. In the internal
organization of their firms, the Japanese commonly provide for much closer
interaction between product designers and production engineers, they devote far
more attention to the refinement of the appropriate process technologies, and
they also assign a more prominent role to the engineering department. [1] In considerable measure, then, their skill in
imitation has been an accompaniment of their skill in, and concern with,
development activities. The significance
of these activities is heightened by a recognition that the ability to imitate
and improve upon one’s own prior performance, rather than starting from
scratch, is often central to success at development activities. If American industry were to improve its
development skills it would also, simultaneously, improve its capacity to
imitate. The two capabilities overlap heavily.
This statement applies with
particular force to the cultivation of a strong interface between product
design and engineering. Japanese strength
at this interface has facilitated its technology-imitation activities by the
ease with which it enables foreign products to be quickly adapted and modified
to suit domestic requirements, and low production costs to be speedily achieved
(see Mansfield). Furthermore, it has
made it possible to move to positions of leadership where new technologies call
for simultaneous optimization on both the process and product sides. The more rapid exploitation of robotics in
Japan appears to be due, in important measure, to the alacrity with which
Japanese firms modified and simplified product design in order to accommodate
the new robotics technology. It has
probably been more sensible to simplify the design of products so that robots
could readily assemble them - reducing the number of component parts and
simplifying the method by which parts are attached to one another - than to
design robots of more general, and therefore more sophisticated, assembling
capabilities.
A central theme in the
study of the development process has been its integrated, interactive, and
iterative nature. In sharp contrast,
American firms have often compartmentalized research and manufacturing
functions. Often, this has led to
breakdowns in the development process characterized by “finger pointing,” in
which functionally specialized groups within the firm assign blame to each other
or to external suppliers. In spite of
this, U.S. firms are often very good at innovation since individual ingenuity
and sharply focused specialization can overcome many obstacles. But these same firms often find it difficult
to make the small steps that are crucial to the ongoing development process. This leaves competitors with a host of
opportunities for imitation and modification for improving performance or
reducing costs,
1. In a recent comparison of innovation in Japan and the United States,
Edwin Mansfield (1988) has observed a striking difference with respect to the
allocation of R&D budgets
between product and process technology. According
to Mansfield, the American firms in his sample devoted 2/3 of their R&D budgets to improved product
technologies and only 1/3 to improved process technologies, whereas among the
Japanese firms only 1/3 of the R&D
budgets were devoted to improved product technologies and 2/3 to
improved process technologies.
231
thereby truncating the appropriation of returns from
innovation.
The Japanese have, on
numerous occasions, been the leaders in the commercialization of new products,
in spite of the fact that the new product, or some essential component, was
invented elsewhere. Although the United
States pioneered both the scientific and technological frontiers in the
invention of the transistor, Japanese firms were the first to succeed in
large-scale application of this technology for radios, and later obliterated
America’s earlier dominance of the market for color television receivers. Japanese success at quality and design improvements
for mass-produced goods such as compact automobiles and consumer electronics
are highly visible. Products requiring
smooth coordination of different technologies (for example, electrical,
electronic, and mechanical) for such things as plain paper copiers, facsimile
machines, floppy disk drives, and personal computer printers, are strongholds
of Japanese commercial and export success. None of these technologies rests on a single
critical innovation. Instead, Japanese
success in each of these areas can be traced to the cumulative impact of its
great development capabilities.
Japanese success in
development has often been able to overcome America’s much-heralded innovative
capabilities. The more specialized an
activity becomes, the greater the importance of efficient information exchanges
if inappropriate tradeoffs or inappropriate optimization criteria are to be
avoided. For specialists to work well in
a large organization, there must be an intimate familiarity with one another’s
goals and priorities. There must be a
set of shared understandings and concerns. The development efforts of Japanese firms
strongly emphasize rotation of personnel among departments in ways that lead to
the exchange of useful information and the formation of common goals. In many cases, close communication among
functionally separate specialists is strengthened by the awareness of a commonality
of interest flowing from stable, long-term employment (and supplier) relationships.
Japanese firms appear to make more
systematic use of engineering skills and production worker experience
throughout the entire sequence of development activities associated with the
introduction of new products, including the most minute aspects of the eventual
manufacturing process.
These activities are not
well appreciated when, as is commonly the case, development is thought of as
the application of scientific knowledge. [2] Development in fact incorporates knowledge
from many sources. Even in those instances in which new scientific knowledge does
provide the initial stimulus for a new product, the subsequent development
process will draw upon a wide variety of sources, the most common of which is
likely to be the existing “in-house” engineering knowledge. Organizational structures and incentive
systems that can exploit these sources effectively will create economic advantages
over competitors who cannot do so, even if these competitors have superior research
capability. If these development capabilities
are sufficiently strong, the stage of commercialization may be reached sooner,
and will certainly be reached by firms in a better position to subsequently
reduce cost and improve performance (Masahiko Aoki and Rosenberg, 1987).
In short, the economic
value of “first-mover” advantages in capturing the economic returns from
innovation is overrated, because innovations are commonly very poorly designed
in their earliest stages and in numerous ways ill-adapted to their ultimate
applications (Rosenberg, 1976, ch. 11). The incremental improvements underlying
development play a critical role in the eventual capture of returns from
innovation.
Thus, there are two reasons
for the primacy of development in capturing the returns from innovations in
markets such as those in which the Japanese have demonstrated success. The first is that efficiency gains in mass production
are often easier to achieve through large numbers of small improvements than
2. The National Science Foundation defines Development as “...the systematic
use of knowledge or understanding gained from research directed toward
the production of materials, devices systems, or methods, including design and
development of prototypes or processes” (1985, p. 221, emphasis added).
232
through major revisions. The second is that cost reduction and
performance improvements in a well-established technology are often capable of
overtaking efforts to advance technology through discontinuous “leaps” or major
innovative steps. The creative elements
of imitation involve not only the adaptation of new or externally created technology,
but a continuing refinement of existing technologies and manufacturing methods.
The expansion of Japanese
industries in the sectors accounting for most of their exports to the United
States is largely the consequence of success at development activities (see our
1988 paper). In the consumer electronics
industry, successful Japanese development of video cassette recording (VCR)
technology employed multiple prototype development efforts, close coordination
of design and manufacturing, and numerous small improvements in a product of
considerable systemic complexity (Richard Rosenbloom
and Michael Cusanamo, 1987). In addition to providing Japanese firms with a
major export market, continuing improvements in VCR products well after their
initial introduction were a major factor in RCA’s failure to gain a market for
video player technology (Margaret Graham, 1986). The same sorts of development activities,
involving numerous small improvements and precise coordination of design and
manufacturing, have been important in the commercialization of laser
technologies for compact disc players and laser printers, markets that are
currently experiencing major expansion.
The role of sustained
incremental improvement and focus on manufacturing processes have also been
important in electronics industries located upstream from the production of
systems for end users. The technological
innovativeness of the integrated circuit (IC) industry at producing intermediate
goods for other electronics companies is by now well known (Ernest Braun and
Stuart MacDonald, 1982). What has been
less well appreciated by economists is that innovations in the IC industry have
been strongly influenced by the incremental improvement of process technology (Steinmueller, 1987). Recent successes of Japanese IC firms in
international competition have been heavily dependent upon success at
manufacturing improvement. In IC production,
the proportion of workable devices emerging from the production process, production
yield, is the most important manufacturing cost factor (Steinmueller,
1987), and yield is very sensitive to both the extent of production experience
and its successful integration with the design process. Japanese accomplishments in international IC
competition involves several factors (Steinmueller,
forthcoming), including successful product and process development and high
yields in the large-scale production of IC memory devices. The implications of development success have
not been confined within the Japanese IC Industry. The demands of Japanese consumer electronics
provided important impetus to the development of CMOS (complementary metal
oxide semiconductor) technology. In part
due to continuing development efforts, CMOS recently has emerged as the leading
technology for future very large scale IC devices. As a consequence, several international joint
ventures and other agreements are creating a flow of technological knowledge
from Japanese to American IC firms (Steinmueller,
forthcoming 1988).
Close communication links
between suppliers and users play a role at the interfirm
level that is analogous to our emphasis on effective communication links among
functional specialists at the intrafirm level. Recent detailed studies of the organization of
parts purchases in the Japanese automobile industry by Banri
Asanuma (1985) demonstrate the existence of long-term relationships with
important institutional mechanisms for coordinating design and assuring timely
supply. Aoki (1987, p. 335) cites an
(unnamed) major auto manufacturer as having 122 stable “first tier” suppliers. More importantly, Aoki characterizes these
relationships as “quasi-permanent,” noting that between 1973 and 1984 only 3
firms exited from this relationship while 21 firms entered. The consequences of such stable supplier
relationships are that development efforts can be jointly initiated and pressed
forward, further extending the coordination of product
233
design and manufacturing beyond the level of the
individual firms as well as improving the flow of information for making
modifications and improvements in the manufacturing process.
We draw an ironic
conclusion from our examination of American and Japanese technological skills. The Japanese have indeed been excellent
imitators. But instead of flourishing a
trump card stating that Americans are excellent innovators, we need to
fix our attention on the disconcerting prospect that innovative skills may
count for a great deal less than we once thought - unless we can learn to
become better imitators ourselves.
References
Abramovitz,
Moses, “Catching Up, Forging Ahead, and Falling Behind,” Journal of Economic
History, June, 1986, 46, 385-406.
Aoki, Masahiko, “A Microtheory of the Japanese
Economy: Information, Incentives and Bargaining,” Kyoto Institute of Economic
Research, Discussion Paper No. 241, Kyoto University, October, 1987.
____________ , and Rosenberg, Nathan, “The Japanese Firm as an
Innovating Institution,” Center for Economic Policy Research, Publ. No. 106, Stanford University, September, 1987.
Asanuma, Banri, “The Organization of Parts
Purchases in the Japanese Automotive Industry” and “The Contractual Framework
for Parts Supply in the Japanese Automotive Industry,” Japanese Economic
Studies, Summer 1985, 13, 32-78.
Braun, Ernest
and MacDonalds Stuart, Revolution in Miniature:
The History and Impact of Semiconductor Electronics, 2d ed., Cambridge:
Cambridge University Press, 1982.
Graham, Margaret B. W., RCA and the Video Disc: The Business of Research, Cambridge: Cambridge University Press, 1986.
Mansfield, Edwin,
“Industrial R&D in Japan and the United States: A Comparative Study”
American Economic Review Proceedings, May 1988, 78, 223-28.
Rosenberg, Nathan, Perspectives
on Technology, Cambridge: Cambridge University Press, 1976.
______________ , and Steinmueller, W. Edward, “Can Americans Become
Better Imitators?,” Center for Economic Policy Research, Publ.
No. 117, Stanford University, 1988.
Rosenbloom,
Richard S. and Cusumano,
Michael A., “Technological Pioneering and Competitive Advantage: The
Birth of the VCR Industry,” California Management Review,” Summer, 1987,
29 55-76.
Steinmueller,
W. Edward, “Microeconomics
and Microelectronics: Economic Studies of Integrated Circuit Technology,” unpublished
doctoral dissertation, Stanford University, 1987.
____________________ , “Industry Structure and Government
Policies in the U.S. and Japanese Integrated Circuit Industries,” in John B. Shoven, ed., Government Policies Toward Industry
in the United States and Japan, Cambridge University Press, forthcoming.
____________________ , “International Joint Ventures in
the Integrated Circuit Industry,” in David C. Mowery, ed., International
Collaborative Ventures in U. S. Manufacturing, American Enterprise
Institute, Ballinger forthcoming, 1988.
National Science Foundation, National Science Board, Science
Indicators 1985, Washington, 1985.
234
The Competitiveness of Nations
in a Global Knowledge-Based Economy
April 2003