Technologies, Products and Organization

in the Innovating Firm: What Adam

Smith Tells Us andJoseph Schumpeter
Doesn't
KEITH PAVITT

(Science Policy Research Unit, Mantell Building, University of Sussex,

Brighton BN1 9RF, UK)

Adam Smith's insights into the increasingly specialized nature of knowledge production
art crucially important in understanding the contemporary problems of managing
innovating firms. Products and firms are based on an increasing range of fields of
specialized technological understanding. Competition is not based on technological
diversity, but on diversity and experimentation in products, etc. Firms rarely fail
because ofan inability to master a new field oftechnology, but because they do not succeed
in matching the firm's systems of coordination and control to the nature of the available
technological opportunities.

$
m

jj
|
P.
|
$
j>
5

All the improvements in machinery, however, have by no means been the

inventions of those who had occasion to use the machines. Many . . .
have been made by the makers of the machines, when to make them
became the business of a peculiar trade: and some by . . . those who are
called philosophers, or men of speculation, whose trade is not to do
anything but to observe everything: and who, upon that account are
often capable of combining together the powers of the most distant and
dissimilar objects. . . . Like every other employment. . . it is subdivided into a
number of different branches, each of which affords occupation to a peculiar tribe
or dass of philosophers; and this subdivision of employment in philosophy, as well
as in every other business, improves dexterity and saves time. (Smith, 1776, p.
8, my italics)

433

Technologies, Products and Organization in the Innovating Firm

1. Setting the Scene

Evolutionary Theory and the Innovating Firm

Attempts over the past twenty years to build an evolutionary theory of the
firm have grown in part out of a dissatisfaction with the inability of
mainstream theory to deal satisfactorily with two important, interrelated and
empirically observable characteristics of contemporary society: continuous
technical change, and the central role of the business firm in generating
learning, improvement and innovation, through deliberate and purposive
action. At the beginning, concepts like 'technological trajectories' and
'routines' were introduced by Nelson and Winter (1977, 1982) to reflect the
cumulative and path-dependent nature of technical change, the often tacit
nature of the knowledge underlying it, and the trial-and-error behaviour of
business practitioners trying to cope with a complex and ever-changing
world.1
Since then, two influential streams of analysis have helped deepen our
knowledge of the innovating firm. First, numerous attempts have been made
to apply the tools and techniques of evolutionary theory (and particularly
biological evolution) directly to modelling and understanding technical
change (e.g. Dosi and Marengo, 1993; Metcalfe and de Liso, 1995). Second,
numerous empirical studies have attempted to formulate generally applicable
laws that explain when and why established firms succeed in innovation, and
when and why they fail (e.g. Iansiti and Clark, 1994; Teece, 1996). This paper
will argue that, whilst both schools have made notable contributions to
understanding of both the nature of the innovating firm and the conditions
for successful innovation, much remains to be done.

Multi-technology Products and Firms

In particular, greater care and attention needs to be devoted to the
distinctions between the artefacts (products, etc.) that the firm develops and
produces, thefirm-specifictechnological knowledge that underlies its ability to
do so, and the organizational forms and procedures that it uses to transform
one into the other. We shall argue that, in the late twentieth century, lack of
technological knowledge is rarely the cause of innovation failure in large firms
' Thii paper concentrates on these aspects of firm behaviour, and not on the wider implications of
evolutionary theory for the theory of the firm. For a concise evaluation of the latter, together with that of
other recent theoretical developments, see Coriat and Wanstein (1995).

434

Technologies, Products and Organization in the Innovating Firm

based in OECD countries. The main problems arise in organization and, more
specifically, in coordination and control.
This can best be understood if more attention is paid to what Adam Smith
said about the division of labour, and less to what Schumpeter said about
creative destruction. Smith's identification of the benefits of specialization in
the production of knowledge has been amply confirmed by experience.
Professional education, the establishment of laboratories, and improvements
in techniques of measurement and experimentation have increased the
efficiency of discovery, invention and innovation. Increasingly difficult problems can be tackled and solved.2 Two complementary forms of specialization
have happened in parallel.
First, new disciplines have emerged, with all the benefits of the division of
labour highlighted by Smith himself at the beginning of this paper. These
specialized bodies of knowledge have become useful over a growing range of
applications, so that products incorporate a growing number of technologies:
compare the eighteenth-century loom with today's equivalent, with its fluid
flow, electrical, electronic and software elements improving the efficiency of
its mechanical functions. In other words, products are becoming increasingly
'multi-technology', and so are the firms that produce them. Each specific body
of technical knowledge cannot be associated uniquely with a single, specific
class of product. 3 Products and related technologies co-evolve within firms,
but their dynamics are different. For example, Gambardella and Torrisi (1998)
found that the most successful electronics firms over the past ten years have
been those that have simultaneously broadened their technological focus and
narrowed their product focus. In other cases, firms have used their range of
technological skills to create or enter new product markets (see Granstrand,
1982; Granstrand and Sjolander, 1990; Oskarsson, 1993; Granstrand and
Oskarsson, 1994).
Second, in addition to the benefits of the cognitive division of labour into
more specialized fields, the rate of technical change has been augmented by
the functional division of labour within business firms, with the establishment
of corporate R&D laboratories and similar groups devoted full-time to
inventive and innovative activities. In addition to the Smithean benefits of
specialization, professionalization and improved equipment, these laboratories
enabled firms to monitor and benefit more systematically and effectively from
2
The dassic texts on thii are Rosenberg (1974), de Soil* Price (1984) and Mowery ind Rosenberg
(1989) See, for example, the reasons why problems in mechanics were solved more easily than those in
medicine.

' Amongst other things, this a a source of frustration for economists who would like to match statistics
on inventions from technology-based patent classes with product based trade and production statistics.
See, for example, Scherer (1982).

435

Technologies, Products and Organization in the Innovating Firm

the outside advances in specialized academic disciplines. And with growing

experience in the development and testing of prototypes, they have allowed
systematic experimentation with a wider range of products and processes than
had previously been possible through incremental improvements constrained
by established products and production lines. In fields of rich technological
opportunity, firms have in consequence become multi-product as well as
multi-technology.

Two Bodies of Knowledge

Hence the importance, in analysing the innovating firm, of distinguishing
clearly artefacts (products) from the knowledge sources on which they are
based.4 Nelson (1998) identifies two, complementary elements in firmspecific knowledge. First, there is a 'body of understanding', based on
competencies in specific technological fields, and reflected in the qualifications
of corporate technical personnel, and in the fields in which they patent and
publish.5 The second element is what Nelson (1998) calls a 'body of practice',
related to the design, development, production, sale and use of a specific
product model or a specific production line. This firm-specific practical
technical knowledge is often obtained through the combination of experimentation, experience, and information and other exchanges amongst different
parts of the organisation.6 As such, it is an organizational task, so that 'a body
of practice' consists largely of organizational knowledge that links 'a body of
understanding' with commercially successful (or, more broadly, useful)
artefacts.
4

For an earlier discuuion of this distinction, see Archibugi (1988).

' For measurement of corporate technological competencies through the fields of qualifications of
technical personnel, see Jacobsson and Oskarsson (1995); through patenting, see Patel and Pavitt (1997);
ind through scientific papers, see Narin and Olivastro (1992), Godin (1996) and Hicks and Kan (1997).
' The difference between the two forms of knowledge isnicelyillustratedin the following passage from
Iansiti and dark (1994), in relation to the firm-specific capabilities for the design and development of dies
used in the production of body panels for automobiles: The knowledge that underlies that capability
includes an understanding of metallurgy, the flow of metal under pressure and che relationship between
the characteristics of the material, the forces and pressures applied to the material and the surface
properties that result. These kinds of knowledge pertain to the fundamental properties of the die and its
production system. But the firm must also have knowledge about how the fundamental concepts can be
operationalised into effective actions. These include knowledge of techniques of die design, die modelling,
die testing and finishing, for example. Additionally, knowledge can take the form of the skill of die
designers in anticipating processing problems, customised software that allows for rapid and effective
testing, patterns of communication and informal interaction between die designers and manufacturing
engineers that allow for early identification of potential problems, and an attitude of co-operation that
facilitates coordinated action between the die designers and the tool makers that will build the dies. These
elements (and many others) define an organisational capability for die design and development' (p. 560).

436

Technologies, Products and Organization in the Innovating Firm

Method
The starting point for our analysis is the large, multi-divisional manufacturing
firm, with established R&D activities and a product range that has grown out
of a common, but evolving, technological competence. In part, this reflects
the focus of this author's recent research (Patel and Pavitt, 1997; Tidd et al.,
1997). More important for the purpose of this paper, large multi-divisional
firms are the largest single source of the new technological knowledge on
which innovation depends. They perform most of the R&D activities, employ
most of the qualified research scientists and engineers, perform and publish
most of the corporate basic research, and maintain the closest links with
academic research (Hicks, 1995). They also contribute to the development of
knowledge and products for their suppliers of production equipment,
components and software (Rosenberg, 1963; Patel and Pavitt, 1994). Finally,
even when they fail in innovation themselves, they remain the major source of
the technological and other competencies which enable new firms with
different organizational approaches to succeed. Understanding the reasons for
their success and failure therefore has the widest implications, not only for
their managers, but also for the distribution of innovative activities amongst
companies of different sizes and ages. It is not the main concern of this paper
to argue in general for or against the large firm's ability to sustain radical
innovation, but to understand better the reasons for its success and failure in
trying to do so.7
We divide our analysis into four parts, reflecting four mechanisms identified
by earlier analysts of the innovating firms: competition, cognition, coordination
and control.8 Table 1 sets out schematically how the division of labour, both in
knowledge production and in corporate innovative activities, has influenced
these four mechanisms. We argue in Section 2 that failure to distinguish
between technologies and products has led to confusion in evolutionary
7
Many analysts in the evolutionary tradition are pessimistic about the ability of large firms to sustain
radical innovations, pointing to recent spectacular failures and to the emergence of new organization
forms: Teece (1996) recently identified four types of firm: conglomerate, multi-product integrated
hierarchy, virtual corporation and 'high flex' Silicon Valley type. Others argue that the obituary of the large
innovative firm may well be premarure, since there remain many examples of their success in developing
and exploiting major innovations: for example, according to Methe a *l. (1996): 'established firms,
including industry incumbents and diversifying entrants, play vital and underemphasized raid as sources
of major innovations in many industries' (p. 1181). And now even the oldest and best established of
capitalistsgrocers (supermarkets) and moneylenders (financial services)have become major players in
the development and exploitation of information technology.
1
These dimensions of the innovating firm emerge from the original work of Nelson and Winter (1982),
and from later work by Cohendet tt tt. (1994). In the language of Teece and Pisano(1994), in their analysis
of the 'dynamic capabilities' of the firm, our competitive mechanisms relate their notions of corporate
pesilin, cognitive mechanisms to corporate paths, and coordination and control mechanisms to corporate
proaua.

437

TABLE 1. Some Consequences of the Division of Labour in the Production of Technological Knowledge
Analytical
implications

oo

Technology ^
products

Division of
labour in
knowledge
production

Laboratories
Disciplines
Trained scientists and
engineers

Increasing output.
range and usefulness
of knowledge

Division of
labour in
business
functions

Specialized technical
functions, inc. R&D
labs

Increasing
competence to
understand and
improve artefacts

Multi-technology
products

Technological
discontinuities *
product
discontinuities

5r
Technological diversity within
firms, and within countries,
but not within industries

5"

ft.

Multi-technology
> Multi-product firms
firms

5Management
implications

Co-ordination
Organizational
competence to
experiment and learn
across organizational
boundaries

Competencies
Technological
competence
enhancement >
competence
destruction

Control
Organizational competence to
reconfigure divisions and
evaluate options in the light
of technology characteristics

Technologies, Products and Organization in the Innovating Firm

writings about what kind of 'diversity' is desirable in competitive processes. In

Section 3, we further argue that, although there are clear cognitive limits on the
range of technologies that a specific firm is capable of mastering, failure in
innovation in established firms is not the result of the destruction of their
technological competencies, but of their inability to match the technological
opportunities with organizational forms and procedures appropriate for their
development and exploitation. In Sections 4 and 5, we analyse in greater depth
how the appropriate forms of two organizational elements that are central to
corporate innovative activitiesmechanisms of coordination and control
depend in part on the nature of the technology itself.

2. Competitive Mechanisms and Technological Diversity

It is around the notion of diversity9 that the distinction between technologies
(bodies of understanding) and products (bodies of practice) is most confused
and potentially most misleading, given the central importance accorded to
diversity in the evolutionary theory of technical change.10 In recent research
undertaken at SPRU on the technological competencies of the world's largest
firms, we have used the level and distribution of corporate patenting by
technical field as a measure of the corporate body of technological understanding (Patel and Pavitt, 1997; O. Marsili, in preparation). This showed
that technological diversity is prominent in some dimensions, but virtually
absent in others.
Large firms are active in a range of technologies broader than the
products that they make. This reflects the multi-technology nature of
their products, and the knowledge required to coordinate in-house
product innovation with innovation in related production systems and
supply chains. What is more, the range of technological competencies
mastered by large firms is increasing over time, as new technological
opportunities emerge.
There is high diversity amongst large firms in the level and mix of their
technological competencies, depending on the products that they
produce. These largely sector-specific mixes of technological competence
change only slowly over time, again in response to changing technological opportunities.
' A reading of the Oxford Oman English Dictionary suggest! that the term 'diversity' is interchangeable
with 'variety' and 'heterogeneity'.
10
'It is a basic proposition of evolutionary theory that a system's diversity affects its development'
(Cohendet and LJerena, 1997, p. 227).

439

Technologies, Products and Organization in the Innovating Firm

There is low diversity in the level and mix of technological competencies

amongst large firms producing similar products. What is more, the
degree of technological diversity is lowest in the product fields with the
highest rate of technical change: computers and pharmaceuticals.
In other words, for the individual firm, technological diversity gives it the
basis to make and improve its products. For the economy as a whole, more
diversity amongst firms in their mixes of specialized technological knowledge
enables them to explore and exploit a fuller range of product markets. But at
the level of the product market or industry, there is similarity rather than
diversity in the level and mix of technological activities in competing firms,
and especially in those with high rates of technical change. Technological
diversity is certainly not a characteristic of competition amongst innovating
firms.11 The diversity exists downstream in the body of practice, namely the
product and process configurations that can be generated from the same or
very similar base of technological knowledge. We know that some of these
configurations do not work out technically, and many more do not work out
commercially (Freeman and Soete, 1997). What emerges is a world where
firms with broad and stable bundles of technological competencies have the
capacity to generate and experiment with a range of product (and process and
service) configurations, some of which succeed, but many of which fail.
At any given time, advances in some fields of technology open major
opportunities for major performance improvements in materials, components
and subsystems (e.g. economies of scale in continuous processes, economies
of miniaturization in information processing). The directions of these
improvements are easily recognized, even if they require the commitment of
substantial resources for their achievement, e.g. Moore's Law in semiconductors.12 Thus, experimentation and diversity do not take place between
different technologies. On the contrary, rich and well-known directions of
" A (frivolous) translation of these results into biological evolutionary terms might be (i) species need a
range of genetic attributes for survival; (ii) since they live in different parts of the forest, the elephant and
the mouse have different genetic mixes, which change only slowly; (iii) there is little room (or need) for
genetic deviance when things are changing fast (and in predictable directionssee below).
13
Recent comments by an IT expert make the point nicely: 'Precious little has happened in digital
technology over the past five years. . . . Steady increases in processor speed and storage size have become
as predictable as a child's growth. . . . Just as the computer industry is predicated on Moore's Lawthat
chips will double in speed every 18 months, which companies can literally plan onthe telecom industry
can be predicated on the transparent network. . . . Change a routine and uneventful. . . . The fiber-optic
backbone has joined the microprocessor on a steady predictable dimb. Processing speeds will double every
18 months. Bandwidth will quadruple every two yean. Corporate planners can rest easy' (Steinberg, 1998,
pp. 80-84). In our framework, the conclusion of the hut sentence does not follow from the preceding
analysis. Such rapid if predictable change in underlying technology is bound to create a plethora of
difficult-to-prtdict products and services.

440

Technologies, Products and Organization in the Innovating Firm

improvement in underlying technologies13 create opportunities for diversity

and experimentation in product configurations. Technological opportunities
create product diversity. There is no convincing evidence that technological
diversity creates product opportunities.14

3. Cognitive Mechanisms and Creative Destruction

Large firms may have competencies in a number of fields of technology but,
in the contemporary world of highly specialized knowledge, the costs of
mastering all of them clearly appear to outweigh the benefits. Firms develop
their technological competencies incrementally, and constrain their search
activities close to what that they already know. Thus, over the past 20 years,
electronics firms have moved heavily into semiconductor technology (but not
biotechnology), and drug firms into biotechnology (but not semiconductor
technology). The firm's knowledge base both determines what it makes, and
the directions in which it searches (Patel and Pavitt, 1997). In this sense, there
are clear cognitive limits on what firms can and cannot do.
The central importance of firm-specific technological competencies has led
some analysts to place technological discontinuities (i.e. major technological
improvements) at the heart of the theory of the innovating firm. In particular, they argue that such discontinuities may either enhance established
competencies and strengthen incumbent firms, or destroy established competencies and undermine them. Again, it must be stressed that technological
discontinuities are not the same as product discontinuities, even if they are
often treated as such. For example, perhaps the most influential paper on
discontinuities by Tushman and Anderson (1986) talks of technological discontinuities in its title, whilst the basis for its empirical analysis are new products
and processes (e.g. jet engines, oxygen steel-making).
Although they may have revolutionary effects, technological discontinuities
rarely encompass allor even most ofthe fields of knowledge that feed into
a product. Typically they may affect the performance of a key component (e.g.
transistors vs. valves) or provide a major new technique (e.g. gene splicing).
But they do not destroy the whole range of related and complementary
13
Nelson and Winter (1977) originally called these 'natural trajectories' and were roundly criticoed by
those arguing that technologies are socially constructed. But perhaps Nelson and Winter were right. The
range of opportunities in different technological fields depends heavily on what nature allows us to do.
Compare rates of increase of information storage capacity over the last 20 years with rates of increase in
energy storage capacity. In the former, newsprint, punch cards and analogue recording have been
overwhelmed by digital methods. In the latter, petroleum remains supreme, in spite of considerable
technological efforts to develop better alternatives.

" In this context, a recent paper, Stankirwia (1998) proposes the notion of interrelated 'design space'
for artefacts, and 'operands' for the underlying knowledge base, techniques, etc.

441

Technologies, Products and Organization in the Innovating Firm

technologies (e.g. sound reproduction in radios, memories in computers,

molecular design in pharmaceuticals) that are necessary for a complete
product. 15 Indeed, they create opportunities for product 'discontinuities' that
often can be achieved only through improvements in complementary but
long-established technologies (e.g. metal tolerances and reliability for robots).
Furthermore, as Gambardella and Torrisi (1998) have shown, corporate
technological dynamics can have different dynamics from corporate product
dynamics, with technological diversification going hand in hand with
increasing product focus: for example (i) when a technological discontinuity
is incorporated in a product family at the mature stage of its product cycle;
or (ii) when a technological discontinuity provokes the emergence of radically
new but technology-related product markets with differentbut as yet
ill-definedcharacteristics; this is probably the case in the electronics
industry studied by Gambardella and Torrisi (1998).
Finally, it should be noted that the predominance given to revolutionary
technologies in the destruction of corporate competence has often been
associated with the notion of paradigm shifts in technology (Dosi, 1982),
similar to those in science (Kuhn, 1962). But this is a misinterpretation of the
notion of paradigm. A new paradigm does not discredit and displace all the
knowledge generated in the earlier paradigms, but instead adds to them.
Newtonian physics still has major theoretical and practical uses, and at least
a quarter of all the new technology created today is still in mechanical
engineering. The development and commercial exploitation of technological
discontinuities turns out to be a more cumulative process than is often
supposed.
Certainly, there are many historical examples of firms that have failed
because they did not master major emerging fields of technology (Cooper and
Schendel, 1976). But competence-destroying technologies are the exception
rather than the rule today, especially amongst large firms, who have
demonstrated a strong capacity through their R&D departments to acquire
and develop competencies in 'discontinuity-creating' technologies like computing and biotechnology (Patel and Pavitt, 1997). The key factors behind
the success and failure of innovating firms must be sought elsewhere, in the
organizational processes linking technologies, products, their production and
their markets.
Cognitive mechanisms also underlie the taxonomy of innovation proposed
by Abernathy and Clark (1985), which distinguishes four types: incremental,
component, architectural and revolutionary. Based on an analysis of innova" On biotechnology, lee McKdvey (1996).

442

Technologies, Products and Organization in the Innovating Firm

tion in photolithographic aligners, Henderson and Clark (1990) argued

that innovations in product architecture 16 destroy the usefulness of the
architectural knowledge in established firms, and this is difficult to recognize
and remedy. More recently, and based on analysis of innovations in computer
disk drives, Christensen and Rosenbloom (1995) concluded that architectural
innovations do not necessarily destroy established competencies. What does
is a change in the 'value network' (i.e. user market) of the innovation.17
Whilst these studies throw interesting and important light on innovation
processes within firms, they must be interpreted with care. An alternative
reading of the Henderson and Clark story is that failure has less to do with
cognitive failure by design engineers to recognize the value of alternative
product architectures, than with organizational factors such as the inability of
design engineers to recognize signals from users or the marketing department, or the unwillingness or inability of corporate management to establish
a new design team or product division. Furthermore, it may be a mistake to
generalize from the experience of US firms specialized in the IT sector to firms
in other sectors and countries. In contrast, for example, to Christensen and
Rosenbloom's emphasis on the difficulties of US firms making computer disk
drives in switching end-user markets, most of the world's leading chemical
firms have been very successful in the twentieth century in deploying their
techniques and products deriving from organic synthesis in markets as diverse
as textiles, building, health and agriculture (Hounshell and Smith, 1988;
Plumpe, 1995).

4. Co-ordination Mechanisms and Learning across

Organisational Boundaries
One of the most robust conclusions emerging from empirical research on the
factors affecting success in innovation is the importance of coordinating
learning and other change-related activities across functional boundaries
(Burns and Stalker, 1961; Rothwell, 1977; Cooper, 1988; Wang, 1997).18
Here we see the second major feature of the division of labour that is central
to contemporary corporate innovative activities, namely coping with
functional specialization, with the emergence of specialized departments for
16
'. . . reconfiguration of an established system to link together existing components in i new way'
(Henderson and dark, 1990, p. 12)
17
The authors liken t change in the 'value network' (i.e. disk configuration and user market) to a
paradigm shift, which implies a much broader definition of the notion of paradigm than that probably
envisaged by Dosi (1982).
11
Problems of such coordination have also figured largely in the works of Coase (1937),
Aolci (1986) and Loasby (1998).

443

ftnrose(1959),

Technologies, Products and Organization in the Innovating Firm

R&D, production, marketing, logistics, strategy, finance, etc. Such coordination cannot realistically be reduced to designing flows of codified information
across functional boundaries. It also involves coordinated experimentation
(e.g. new product launches), and the interpretation of ambiguous or
incomplete data, where tacit knowledge is essential. As the observations of
Iansiti and Clark in footnote 6 (p. 436) show, personal contacts, mobility and
interfunctional teams are therefore of more central importance than pure
information flows.
In our present state of knowledge, effective coordination belongs in the
field of practice rather than the field of understanding. Unlike purely
technological processes, organizational processes are difficult to measure and
evaluate, and do not lend themselves readily to rigorous modelling and
controlled experiments. In addition, the coordination processes in which we
are interested are complex. Experimentation and learning across critical
organizational interfaces are particularly difficult when combining knowledge
from different functions, professions and disciplines, each with their distinct
and different analytical frameworks and decision ruleswhich is another
reason why firms may try to compensate for greater technological complexity
by greater market focus.19
In addition, the identification of the location of critical interfaces is not easy,
for three reasons. First, there are potentially several such interfaces, involving
a multitude of possible linkages between R&D, production, marketing and
logistics within the firm, and a variety of sources of outside knowledge in
universities, other firms (suppliers, customers, competitors, etc.) and other
countries. Second, the interfaces that merit analysis and managerial attention
vary considerably amongst technologies and products. Compare firms in
Pharmaceuticals and automobiles. In the former, strong interfaces between
in-house R&D and the direct output of academic research (in medicine,
biology and chemistry) are essential. In the latter, they are not, but strong
interfaces between in-house R&D and production are of central importance.
These differing characteristics have important implications for both the
appropriate organizational forms, and geographical location of corporate
innovative activities.
Finally, the key interfaces for organizational learning change over time, very
often as a result of changes in technology-related factors themselves. Witness
the growing importance for the pharmaceutical industry of the interface with
19
Models of inert-corporate coordination have not got very far in grappling with these essential features
of innovative activities. For example, in Aoki's models (1986), problems of coordination are in production
and dealt with through information flows, rather than in learning and innovation mediated through tacit
knowledge. Furthermore, sources of instability and change are in an exogenous environment, rather than
created by the firms themselves.

444

Technologies, Products and Organization in the Innovating Firm

research in academic biology, the consumer market interface for producers of

telephones and computers, and of interfaces with software and materials
technologies for firms in virtually all sectors. One major source of failure in
innovation is likely to be inadequate recognition of the importance of these
new interfaces, or the inability of management to take effective action to
establish them. We should look for what Leonard-Barton (1995) calls 'core
rigidities', when individuals and groups with the established competencies for
today's products are either ignorant of, or feel threatened by, the growing
importance of new competencies.

5. Control Mechanisms: Matching Strategic Styles

with Technologies
The lack of a one-to-one link between each product and each technology has
at least two major implications for organizational practice in business firms.20
The first is that firms which master fields of rich technological opportunity are
often able to develop and produce several products based on the same body of
knowledge. In other words, they compete and grow through technologyrelated diversification.21 Second, the very existence of this broadly useful
knowledge means that the classic M-form organization is unable to match
tidily each field of its technology to one product or to one division.
As a consequence, systems of corporate control in the multi-product firm
have a major influence on the rate and direction of its innovative activities.
Chandler (1991) distinguishes two essential functions of corporate control:
the entrepreneurial function of planning for the future health and growth of the
enterprise, and the administrative function of controlling the operation of its
many divisions.22 The administrative function is normally exercised through
systems of financial reporting and controls. The entrepreneurial function for
technology is the capacity to recognize and exploit technology-based
opportunities. This requires an ability to evaluate projects and programmes
where the normal financial accounting techniques are often inoperable and
inappropriate, since exploratory research programmes should be treated as
options, rather than full-scale investments (Myers, 1984; Hamilton, 1986;
Mitchell, 1986; Mitchell and Hamilton, 1988). It may also require the
establishment of a central corporate research programme or laboratory,
funded in part independently from the established product divisions
20
For more extended discussions, see Kay (1979), Prahalad nd Hamel (1990), von Tunzrlmann (1995)
and Marengo (1995)
21
See Rumelr (1974). Numerous examples can be found in the electrical and chemical industries.
22

See also the earlier pioneering work of Goold and Campbell (1987).

445

Technologies, Products and Organization in the Innovating Firm

(Graham, 1986). And it will certainly require the capacity to reconfigure the
composition and objectives of established divisions in the light of changing
opportunities (Prahaled and Hamel, 1990).
Different balances between the administrative and the entrepreneurial
functions are likely to be appropriate to different levels of technological
opportunity. In addition, the appropriate degree of decentralization of the
entrepreneurial function within the corporation will depend in part on the
nature of the firm's core technology.23 The higher the costs of product
development, the greater the need for central control of the entrepreneurial
function. In other words, the appropriate system of corporate control will
depend in part on the nature of the technology.
Thus, Table 2 suggests that firms with low technological opportunity are
likely to be compatible with an emphasis on the administrative rather than
the entrepreneurial function, and with more centralization with increasing
capital intensity. Firms with high technological opportunities and high
costs of product and process developmentsuch as those in drugs and automobilesare likely to be best suited to a strong entrepreneurial function at
the corporate level. Those with high technological opportunities, but low
costs of product developmentlike those in consumer electronics and the 3M
Corporationwill be best served by more decentralized entrepreneurial
initiative.
Table 2 also shows that there can be mismatches between strategic style
and the nature of technological opportunities. For example, tight financial
control and emphasis on short-term profitability do not allow investments in
exploring longer-term options emerging from new technological opportunities: this is one reason why GEC in the UK and ITT in the USA have
progressively excluded themselves from many high-technology markets
{Economist, 1995, 1996). Similarly, the characteristics of technology, and the
corresponding organizational requirements, change over time. Thus, one
reason for the recent deliberate demerger of ICI was the reduced technological
opportunities in the previously fast-moving field of bulk chemicals (Owen and
Harrison, 1995). Similarly, the high costs of mainframe computers in the
1960s and 1970s, and their specificity to the corporate office market, imposed
centralized entrepreneurship. With the advent of the microprocessor and
packaged software, the costs of experimentation tumbled and new markets
emerged. Mainframe firms had great difficulty in adjusting in time to the
requirements of greater decentralization.
2i
Marengo (1995) models learning, and comes to tome intuitively appealing conclusions about the
balance between organizational centralization and decentralization. But his learning is also about changes
in the environment, rather than about internally generated changes.

44 6

Technologies, Products and Organization in the Innovating Firm

TABLE 2. Technology and Corporate Control

Strategic style

Levels of
decision-making

_. . .

Division

Entrepreneurial

Administrative

Higb-tecb opportunity + bigb costs

of product development

Low-tecb opportunity +
bigb cost of investments

-+
-)

drugs
automobiles
bulk chemicals in 1960s >
mainframes in 1970s

GEC(UK)
ITT (USA)
aluminium
steel

Higb-tecb opportunity + low costs

of
J rproduct development

Low-tecb opportunity +
low cost of investments

,
consumer electronics
3M

conglomerates

6. Conclusions
The main argument of this paper is thatas foreseen by Adam Smith
specialization in knowledge production is a central feature of the innovating
firm. It is therefore of great importance to distinguish products (and other
artefacts) from the underlying bodies of technological understanding on
which they are based. Although the two evolve together, they do not have the
same dynamics. Inadequate care in distinguishing the two can result in
mistaken policy prescriptions (e.g. Granstrand etal., 1997). And it can lead to
too much emphasis in evolutionary theorizing on the economic benefits of
technological diversity, on the frequency and causes of creative destruction,
and on the nature and implications of changes in technological paradigms.
The main challenge is to improve understanding of the organizational
processes of coordination and control that make for a successful matching
between the development and deployment of bodies of technological
knowledge, on the one hand, and commercially successful (or useful) working
artefacts, on the other. We have stressed that our practical and theoretical
knowledge of these organizational processes are less well grounded than
our knowledge of the processes of technological advance per se. This why
companies with outstanding technological competenciesXerox and IBM
in the early days of personal computers, for examplefailed to develop
organizational forms to exploit them. Nonetheless, large firms are capable of
restructuring their activities to benefit from the new technological oppor447

Technologies, Products and Organization in the Innovating Firm

tunities that they have mastered. 'Routines' can and do change. 'Creative
destruction' is not inevitable.
The appropriate organizational processes will depend on the characteristics
of the technologies, such as their sources, the rate and direction of their
change, and the costs of developing and building artefacts based on them.
And since technologies vary greatly in these characteristics, and they change
over time, any improved knowledge that we acquire will be highly
contingent. Nonetheless, the research of Woodward (1965) and Chandler
(1977) on the organizational dimensions of changes in process technologies
shows that such research can make a major difference to our understanding
of innovation in firms. The following avenues of research appear to be
particularly fruitful:
1.

2.

3.

Bibliometric studies and surveys that map linkages between knowledge,

products and organization in business firms over a range of sectors. This
is essential given intersectoral variety. The great challenge is to develop
measures of organization that are conceptually clear and empirically
robust.
Detailed studies by historians and sociologists of the interactions between
the development of the technological knowledge base, and the associated
artefacts that emerge from them (e.g. Constant, 1998; Stankiewicz,
1998).
Case studies of the effects on firms, their organization and their products
of the introduction of technological discontinuities, whether in the
form of new sources of useful knowledge, or of order-of-magnitude
improvements in the performance in one field. If then our analysis is
correct, large firms in advanced countries will have few difficulties in
mastering the new technology, resultant product discontinuities will
happen only after a extended period of learning,24 and failure is likely to
result from 'core rigidities', namely resistance from established groups
within the organization.

Finally, our analysis suggests that truths about the real innovating firm will
never be elegant, simple or easy to replicate. It is nonetheless to be hoped that
formal theorizing will try to incorporate more real-world features of the
innovating firm. In particular, evolutionary economics grew out of dissatisM
See Miyazaki (1993) for an account of the extended period that Japanese firms spent learning about
opto-dectronics. It might be argued thai the personal computer began with a component innovation (the
microprocessor) which, after a number of complementary component innovations (e.g. memories), and
architectural innovation] (internaliiing the dislc drive) and incremental improvements, created the
conditions for the emergence of the revolutionary innovation that was the P C

448

Technologies, Products and Organization in the Innovating Firm

faction with mainstream formalizations of technical change. It would be a pity

if it ended up going down the same path.

Acknowledgements
The author has benefited greatly in the preparation of this paper from the
comments of Stefano Brusoni, Mike Hobday, Patrick Llerena, Richard Nelson,
Ed Steinmueller and two anonymous referees. The usual disclaimers apply.

References
Abernathy, W and K. Clark (1985), 'Innovation: Mapping the Winds of Creative Destruction,' Rtseanb
Policy, 14, 3-22.
Aoki, M. (1986), 'Horizontal vj. \fcrtical Information Structure of the Firm," Amtrican Economic Rtview, 76,
971-983.
Archibugi, D. (1988), 'In Search oft Ujeful Measure of Technological Innovation,1 Tkbnological Fonaating and
Socud Cbangt, 34, 253-277.
Burns, T. and G. Stalker (1961), Tbt Managtmat of Innovation. Tavistock: London.
Chandler, A. D. (1977), Tbt VisiUt Hand: Tbt MMxagtrial Rtvolution m Amman Bvinas. Belknap Press:
Cambridge, MA.
Chandler, A. (1991), The Functions of the HQ Unit in the Multibusiness Firm,' Strattgic Mjuagtmtntjotmul,
12,31-50.
Christensen, C and R. Rosenbloom (1995), 'Explaining the Attacker's Advantage: Technological Paradigms,
Organisational Dynamics, and the Value Network,' Rtseanb Policy, 24, 233257.
Coase, R. (1937), The Nature of the Firm,' Economic*, 4, 386-405.
Cohendet, P and R Llerena (1997), 'Learning, Technical Change and Public Policy: How to Create and
Exploit Diversity,' in C. Edquist (ed.), Systems of Innovation: Tabnologies, Institutions and Organisations. Pinter:
London.
Cohendet, P, R lierena and I_ Marengo (1994),: 'Learning and Organizational Structure in Evolutionary
Models of the Firm', EUNET1C Conference 'Evolutionary Economics of Technological Change: Assessment
of Results and New Frontiers', Strasbourg.
Constant, E. (1998), 'Recursive Practice and the Evolution of Technological Knowledge,' in J. Ziman (ed.),
Tkbnological Innovation as an Evolutionary Pnass. Cambridge University Press: Cambridge (in press).
Cooper, A. and D. Schendel (1976), 'Strategic Responses to Technological Threati,' Business Horizons,
February, 61-69.
Cooper, R. (1988),The New Product Process: A Decision Guide for Management,' Journal of MMrktting
MMmapmat, 3, 238-255.
Coriat, B. and O. Wtinstein (1995), Us Namdla Tbtorics <U rExtnprist. Livre de Poche, Librarie Generale
Franchise: Paris.
de Solla Price, D. (1984), The Science/Technology Relationship, the Craft of Experimental Science, and
Policy for the Improvement of High Technology Innovation,' Retard} Polity, 13, 320.
Dosi, G. (1982), Technological Paradigms and Technological Trajectories: A Suggested Interpretation of the
Determinants and Direction! of Technical Gunge,' fttstanb Policy, 11, 147162.
Dosi, G. and L. Marengo (1993), 'Some Elements of and Evolutionary Theory of Organisational

449

Tkcbnobgies, Products and Organization in the Innovating Firm

Competencies', in R. W England (ed.), Evolutionary Catapts in Contemporary Ecmmia. University of Michigan
Preij: Ann Arbor, MI.
Economist (1995), T h e Death of the Geneen Machine1, June 17,86-92.
Economist (1996), 'Changing of the Guard', September 7, 72.
Freeman, C and L. Soete (1997), Tbt Economics of Industrial Innovation. Cassel-Pinter: London.
Gambardella, A. and S. Tbrrisi (1998), 'Does Technological Convergence Imply Convergence in Markets?
Evidence from the Electronics Industry,' Research Policy (in press).
Godin, B. (1996), 'Research and the Practice of Publication in Industry,' Research Policy, 25, 587-606.
Goold, M. and A. Campbell (1987), Strattgits and Styles: Tbt Rolt of tbt Centn m Managing Diversified
Corporations. Bladcwell: Oxford.
Graham, M. (1986), 'Corporate Research and Development: The Latest Transformation,' in M. Horwitch
(ed.), Tkbnohgy in tbt Modern Corporation: a Strategic Perspective. Pergamon Press: New York.
Granstrand, O. (1982), Technology, Management and Markets. Pinter: London.
Granstrand, O. and C Oikarsson (1994), Technology Diversification in "Multi-tech" Corporations,' IEEE
Transactions on Engineering Management, 4 1 , 355364.
Granstrand, O. and S. Sjolander (1990), 'Managing Innovation in Multi-technology Corporations,' Rcstanb
Policy, 19, 35-60.
Granstrand, O., E Patel and K. Pavitt (1997), 'Multi-technology Corporations: Why diey Have 'Distributed'
rather than 'Distinctive Core' Competencies', California Management Review, 39, 825
Hamilton, W (1986), 'Corporate Strategies for Managing Emerging Technologies', in M. Horwitch (ed.),
Technology in tbt Modern Corporation: A Strategic Persptam. Pergamon Press: New York.
Henderson, R. and K. dark (1990), 'Architectural Innovation: The Reconfiguration of Existing Product
Technologies and the Failure of Established turns,' Administrative Sciences Quarterly, 3 5 , 9 - 3 0 .
Hicks, D. (1995), 'Published Papers (1995) Tacit Competencies and Corporate Management of the
Public/Private Character of Knowledge,' Industrial and Corporate Cbangt, 4,401-424.
Hicks, D. and S. Kara (1997), Tbt Changing Shape ofBritisb Industrial Research, STEEP Special Report no. 6,
Science Policy Research Unit, University of Sussex.
Horwitch, M. (ed.) (1986) Technology in tbt Modern Corporation: A Strattgic Perspective. Pergamon Press: New
York.
Hounshell, D. and J. Smith (1988), Science and Corporate Strategy: DM Pmt R&D, 1902-1980.
University Press: New York.

Cambridge

Iansiti, M. and K. Clark (1994), "Integration and Dynamic Capability: Evidence from Product Development
in Automobiles and Mainframe Computers,' Industrial and Corporate Change, 4, 557-605.
Jacobsson, S. and C Oskarsson (1995), 'Educational Statistics as an Indicator of Technological Activity,'
Reuartb Policy, 24, 127-136.
Kay, N. (1979X The Innovating Firm. Macmillan: London,
Kuhn, T. (1962), Tbt Smanrt ofScientific Revolutions. Chicago: University of Chicago Press.
Leonard-Barton, D. (1995), WHhprinp ofKnowkdgt. Harvard Business School Press: Boston, MA.
Loasby, B. (1998), The Organisation of Capabilities,'_/iwrW of Economic Behaviour and Organisation (in press).
Marengo, L. (1995), 'Structure, Competence and Learning in Organisations,' Wirtscbafispolixht Blatter, 42,
454-464.
McKetvey, M. (1996), Evolutionary Imvations: Tht Business of Biotechnology. Oxford: Oxford University Press.
Metcalfe, S. and N. de Liso (1995), 'Innovation, Capabilities and Knowledge: The Epijtcmk Connection', in

450

Technologies, Products and Organization in tbt Innovating Firm

J. de U Motcre and G. Piquet (eds), Evolutionary Economics and tbt New International Public*! Economy. Pinter.
London.
Mcthe, D., A. Swnminarhan tnd W Mitchell (1996), 'The Underemphasiied Role of Established Finns as
Sources of Major Innovations,' Industrial and Corporate Change, 5, 1181-1203.
Mitchell, G. (1986), 'New Approaches for the Strategic Management of Technology' in M. Horwitch (ed.),
Tkfmtlogy in tbt Mtdern Corporation: a Strategic Perspective. Pergamon Press: New York.
Mitchell, G. and W Hamilton (1988), 'Managing RAD as a Strategic Option,' Research- Technology
Management, 31, 15-22.
Miyazaki, K. (1995), Building Competencies in tbt Firm: Leans from Japanese and European Optoelectronics.
Macmillan/St Martin's Press: New York.
Mowery, D. and N. Rosenberg (1989), Technology and tbt Pursuit of Economic Grtuitb. Cambridge: Cambridge
University Press.
Myers, S. (1984), 'Finance Theory and Financial Strategy,' Interfaces, 14, 126-137.
Narin, F. and D. Olivastro (1992), 'Status Report: Linkage between Technology and Science,' Research Policy,
21,237-249.
Nelson, R. (1998), 'Different Perspectives on Technological Evolution,' in J. Ziman (ed.), Technological
Imtevatim as an Evolutionary Process. Cambridge: Cambridge University Press (in press).
Nelson, R. R. and S. G. Winter (1977), 'In Search of a Useful Theory of Innovation,' Research Policy, 6, 36-76
Nelson, R. R and S. G. Winter (1982), An Evolutionary Theory of Economic Change. Belknap Press: Cambridge,
MA.
Oskarsson, C (1993), Technology DiversificationTbt Pbtntmtnon, its Causes and Effects. Department of
Industrial Management and Economics, Chalmers University of Technology, Gothenburg.
Owen, G. and T. Harrison (1995), 'Why ICI Chose to Demerge,' Harvard Business Review, March-April,
133-142.
PareL P ind K. Pavitt (1994), The Continuing, Widespread (and Neglected) Importance of Improvements
in Mechanical Technologies,' Research Policy, 23, 533-546.
Patd, P. and K. Pavitt, (1997), The Technological Competencies of the World's Largest Firms: Complex and
Path-dependent, but not Much \briety,' Research Policy, 26, 141-156
Penrose, E. (1959), The Theory of tbt Growth of tbt Firm. BlackweU: Oxford.
Plumpe, G. (1995), 'Innovation and the Structure of the IG Farben,' in F. Caron, P Erker and W. Fischer
(eds), Innovations in tbt European Economy between tbt Vim. De Gruyter, Berlin.
Prahalad, C. K. and G. Hamel (1990), The Core Competencies of the Corporation,' Harvard Business Review,
May-June, 7 9 - 9 1 .
Rosenberg, N. (1963), Technological Change in the Machine Too! Industry, 1840-1910,' Journal of Ecmemic
History, 23, 414-446.
Rosenberg, N. (1974), 'Science, Invention and Economic Growth,' EanmiiJournal, 84, 333.
Rothwdl, R. (1977), T h e Characteristics of Successful Innovators and Technically Progressive Firms,' R&D

Manapment, 1, 191-206.
Rumdt, R. (1974), Strategy, Structure and Ecmmic Performance. Graduate School of Business Administration,
Harvard University.
Scherer. F. (1982), 'Inter-industry Technology Flows in the US,' Research Policy, 11, 227-45
Smith, A. (1776), The Vtalth of Nations. Dent (1910): London.
Stankiewici, R. (1998), Technological Change as an Evolution of Design Spaces', in J. Ziman (ed.).
Technological Innovation as an Evolutionary Process. Cambridge: Cambridge University Press (in press).

451

Technologies, Products and Organization in the Innovating Firm

Steinberg, S. (1998), 'Schumpeter's Lesson: What Really Happened in Digital Technology in the Past Five
Years,' Wind, January.
Trece, D. (1996), 'Firm Organisation, Industrial Structure and Technological Innovation,' Joxmal of Economic
Btbaviour tad Organisation, 31, 193-224.
Teece, D. and G. Pisano (1994), "The Dynamic Capabilities of Firms: An Introduction,' Industrial and
Orporatt Cbangt, 3, 537-556.
Tidd, J., J. Bessant and K. Pavitt (1997), Managing Innovation: Integrating Technological, Market and
Orgimuatinsl Cbtagt. Wiley: Chichester.
von Tunxelmann, N. (1995), Tkbtnlog) **d Industrial Pngrtss. Elgar: Aldershot.
Tushman, M. and P Anderson (1986), Technological Discontinuities and Organisational Environments,'

Admnistntivt Sriaia Quarterly, 31, 439-465.

Wang, Q. (1997), 'R&D/Marketing Interface in a Firm's Capability-building Process: Evidence from
Pharmaceutical Firms,' InttnutfiomdJournal >f Innovation Managtmtnt, 1, 2352.
Woodward, J. (1965), Industrial Organisation: Tbtory and Practia. Oxford: Oxford University Press.
Ziman, J. (1998), Tkbmiegical Innovation as an Evolutionary Pnass. Cambridge: Cambridge University Press
(in press).

452