Towards a theory of the technology-based firm

Ž .Research Policy 27 1998 465–489

Towards a theory of the technology-based firm 1

Ove Granstrand )

Industrial Management and Economics, Chalmers UniÕersity of Technology, S-412 96 Goteborg, Sweden¨

Abstract

The modern firm is a very viable economic institution, drawing strength from a competitive market economy, withembedded ‘super-markets’ for corporate control and ‘sub-markets’ for internal organization. The technology-based firm in

Ž .addition draws strength from its co-evolution with modern science and technology and vice versa , and thereby becomesincreasingly important. However, received theories of the firm, of which there are many, have not particularly taken accountof technology and technology-based firms, nor their management. This paper takes an empirical point of departure fromrecent findings regarding the positive relationship between technology diversification on the one hand and corporate growthand business diversification on the other. These findings are not readily explainable by received theories of the firm, whichthe paper reviews, and the findings are thus taken as an explanandum for a proposed approach to formulate a theory of thetechnology-based firm. The approach is compatible with various other theoretical approaches such as the resource-based, the

Žtransaction-cost and the evolutionary approach, but specifically takes the idiosyncrasies of technology i.e., technical.competence as well as management into account. Through notably strong economies of scale, scope, speed and space

associated with the combination of different technologies and resources, the technology-based firm is subjected to specificdynamics in its growth and diversification and shifts of businesses and resources. In particular, a technology-based firmtends to engage in technology diversification, thereby becoming multitechnological. As such the technology-based firm hasincentives to economize on increasingly expensive new technologies by pursuing strategies of internationalization on bothinput and output markets, technology-related business diversification, external technology marketing and sourcing, R&Drationalization and technology-related partnering. q 1998 Elsevier Science B.V. All rights reserved.

Keywords: Technology-based firm; Technology; Management; Diversification; R&D

1. Introduction: co-evolution of technology andthe firm

In the history of institutions the modern businessŽfirm is a fairly recent innovation emerging in 19th-

) Tel.: q46-31-772-1209; Fax: q46-31-772-12401 Paper originally presented at the workshop on ‘Technology

and the Theory of the Firm’, organized at the University ofReading, 15th–16th May 1995, with the support of the EuropeanScience Foundation. The comments from Erik Berglof, SvenLindmark, Luigi Orsenigo and participants in the workshop aregratefully acknowledged.

.century Europe , preceded by far by institutions likeŽthe church, the farm, the university emerging in

. Ž12th-century Italy and the bank emerging in me-.dieval Italy as well , and also by the modern nation-

state. 2 As an institutional species the modern busi-ness firm in market economies has developed a

2 Inserting a qualifier like ‘modern’ obviously opens the doorfor vagueness in dating. Here ‘modern business firm’ refers to ajoint-stock limited-liability firm and ‘modern nation-state’ to thetype emerging in connection with the American and Frenchrevolutions in the late 18th century.

0048-7333r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved.Ž .PII: S0048-7333 98 00067-5

( )O. GranstrandrResearch Policy 27 1998 465–489466

remarkable viability and variety. 3 In the 20th cen-Žtury larger and more diversified MNCs multina-

. Žtional corporations and MPCs multi product cor-.porations have emerged, typically internalizing some

teaching, R&D, and banking functions and at thesame time forging links with universities and banks,while the latter institutions at the same time increas-ingly adopt organizational features from businessfirms in advanced countries.

The viability of the firm as an institution partlystems from its recombinant property through themarket for corporate control. The latter market is infact a kind of ‘super-market’ since it works as a

Žselection mechanism ‘selection mechanism’ in the.sense of Nelson and Winter, 1977, 1982 of a higher

order, speeding up both variety generation and selec-tion. Secondly, the viability of a firm also stemsfrom its ability to combine, cumulate, and recombineresources, as well as the ability to rapidly respond tobusiness opportunities and threats, an ability fosteredby competition, and the legal framework around afirm together with the signalling system of account-

Žing. This, however, may in turn produce monopolis-tic as well as myopic behaviour, both potential dis-

.functions of the firm as an economic institution .A third basis for the viability of the firm stems

from the development of governance structures andmanagerial capabilities. From a transaction-cost per-spective the emergence of increasingly large-sized,diversified and internationalized corporations couldbe interpreted as a sign of increasingly loweredgovernance or management cost on an average, com-pared to market transaction costs. However, themodern firm has also increasingly learnt to utilizeinternal ‘sub-markets’ in order to reduce gover-

Žnance cost andror raise innovativeness e.g., through.the M-form , as well as having learnt to link up with

external partners in networks. Thus, the modern firmhas developed into a quasi-integrated hybrid form

3 Despite the terms used here no biological analogies areintended. On the contrary, such analogies are often more mislead-ing than helpful when applied to firms and technologies; see

Ž . Ž . ŽPenrose 1952 and Granstrand 1994a, Ch. 19 . The admittedly.biology-inspired terminology is used here as a language of

general evolution, not necessarily confined to its application inŽ .biology cf. von Bertalanffy, 1968 .

of organization, still with a considerable source ofcentral power, however, thereby distinguishing itfrom a mere network.

A fourth part of the viability of the firm stemsfrom the long-run evolution of S&T, including the

Žscientific ‘revolutions’ during the Renaissance in-troducing, e.g., the method of systematic experimen-

.tation and the Age of Enlightenment. The S&Tevolution has continually generated business oppor-tunities and consequential needs, while the basichuman needs related to the fundamentally differentbiological evolution have remained much the same. 4

An increasing range of business firms has then inŽ .one way or another product- or process-wise be-

come reliant or based upon technology in exploitingbusiness opportunities, thus giving rise to the tech-

( )nology-based f irm TBF as a growing sub-speciesof the modern business firm. 5

Ž .As technology engineering knowledge has be-come an economic engine in technology-based firms,these have collectively also become an increasinglyimportant and now major source of technology. 6

Thereby firm-based technology has increased, abso-lutely as well as relatively. Thus, the increasing

4 Humans have not, for example, developed three arms forcocktail parties, separate talking and eating body organs or en-tirely new senses with new forms of stimulation and entertain-ment.

5 There is no need to operationalize here through some cut-offpoint when a firm is technology-based and when it is not.However, it is important for the rest of the paper to spell out thatthe concept of technology is taken here in the narrow sense ofnatural science and engineering or technical knowledge, thereby

Ž .sticking to an old tradition according to Cantwell 1994 . Morespecifically I will equate technology with a body of knowledge oftechniques that falls in areas containing in principle patentableknowledge, apart from the novelty and non-obviousness require-ment of the patent system. Operationalization of technology in thisway is immensely aided by the international patent system, whichthus provides classification operations which in turn are important

Ž .for codification cf. with chemistry . In the same spirit, newnessŽof technology could be defined as less than 20 years smaximal

.patent lifetime of occurrence and a unit of technological advanceas the minimum invention level required for patentability. Thepatent system thus also offers an operationalization of novelty aswell as of the size of a novelty in form of a simple metric norm.In general the patent system is underutilized as a way to opera-tionalize technology. It is far from error-free but it is the best athand, generally speaking.

6 Ž .See, e.g., Pavitt 1991 .

( )O. GranstrandrResearch Policy 27 1998 465–489 467

importance of technology-based firms and firm-basedtechnologies is a perfect example of virtuous 7 co-evolution of an expansionary economic institutionand S&T knowledge, at the same time independentof biological evolution in general. 8

Given the prevalence and increasing importanceof this co-evolution of firms and technologies, it isnatural to theorize about it as a phenomenon. Thispaper is an attempt to do so, and then with theprimary focus on how technology-based firms can becharacterized and how they form, grow and expandinto many products, markets and technologies. Al-though heterogeneities in the firm may be importantsources of potential economies to be reaped throughmanagement, the specific internal organization oftechnology-based firms is not the primary focus ofthis paper.

2. Received theory

2.1. General surÕey

Keeping in mind the ever-present difficulty ofjudging current times, one can venture to say thatcurrently there is a surge of efforts among scholarsto formulate new or improved theories of the firm,its nature, reasons for existence, behaviourŽ .actualroptimal , differences and effects. This sec-tion cannot comprehensively review these efforts,nor put them in their historical context. 9 A fewcommon features of current theorizing will be pointedout, however, and then the role of technology and

7 In some sense the Christian church as an institution on theone hand and S&T on the other hand would rather be an exampleof adversary co-evolution, at least in the Middle Ages.

8 That is to say, independence at large and so far. Biology andgenetic engineering develop as a S&T field and biotech-basedfirms increase in numbers. New evolutionary patterns in naturemay take root as a result, desirable or not. Incidentally, this mayopen up possibilities for new analogies between biology, firmsand S&T.

9 This is a large research undertaking in itself, for which thereis currently a need, although there are several good works of such

Ž .a review type; see, e.g., Fransman 1994 , and the whole journalissue of ‘Industrial and Corporate Change’, Vol. 3, No. 3, 1994,for a review of current theorizing, related to the concept ofdynamic capabilities.

management in received theory will be briefly sur-veyed.

A first common feature of contemporary theoriz-ing about the firm is that the firm is viewed as‘something A of something B’, using two sets ofterms for A and B. A-terms are then pool, set,bundle, nexus, cluster, portfolio, repository etc. andB-terms are resources, contracts, treaties, relations,capabilities, competences, activities, incentives, peo-ple, products and so on. In this way two fundamentalproperties of the firm are pointed out. First, thecoherence in some sense 10 of the firm is expressedby the A-terms which are conceptually similar. Sec-ond, the heterogeneity of the firm is indirectlyexpressed by the range of conceptually differentB-terms, the choice of which depends upon what issuggested to be considered as being of primary focalconcern. 11 In passing one may note that no AB-combination strictly contradicts another and thus theymay in principle be seen as complementary partialviews.

A second fairly common feature is the emphasisŽon time-related properties of the firm although there

.is considerable theorizing which is essentially static .Besides general time-oriented concepts such as dy-namic, historic, evolutionary etc., commonly usedterms refer to path-dependency, continuity etc. Hereone can discern a dimension along which authorsorient themselves, and that is from emphasis oncontinuity to emphasis on discontinuity in viewingthe evolution of firms and their features. 12 How-ever, although there is a great deal of writing aboutactivities inducing discontinuities such as start-ups,mergers, acquisitions, spin-offs, restructuring,bankruptcies and the like, one is left with an impres-sion that the currently prevailing emphasis is ratheron continuity, stressing the firm’s resilience, adapta-tion, cumulation, long-run build-up of competitiveadvantage, sustainability and so on. 13 This espe-

10 The very concept of coherence of the firm has been elabo-Ž .rated by Dosi et al., Dosi et al., 1992 .

11 Thus, a firm is like a coherent but ‘white elephant’, withpluralistic explanations, not necessarily contradictory.

12 A similar continuity–discontinuity dimension exists amongviews on evolution of technologies.

13 Ž .See, e.g., Pavitt 1991 .


cially pertains to views on large firms. Thus, onemay be led to believe that continuity and large sizeof the firm are associated, but one may as well be ledto believe in a general increase of continuity inindustrial society. 14

The continuity view of the firm implies the exis-tence of inertia and various speed limits in the

Ž .processes of evolutionary rather than revolutionarychange of the firm, speed limits beyond which there

Žare no timercost trade-offs e.g., regarding learning.or mobility of resources . That is, beyond some point

time cannot be bought, and thus time is an indepen-dent analytical concept, irreducible to money andother economic resources.

The third common feature of theories of the firmis the notion of existence of a firm’s interior in

Žsome sense, as opposed to its exterior or its environ-.ment . The notion of an interior and an exterior

follows from a notion of a boundary of a firm butdoes not require such a notion. That is, a concept ofa boundary of a firm is sufficient but not necessaryin order to have a concept of an interior of a firm. 15

Although perhaps overly abstract, this neverthelesshas implications for transaction cost theorizing.

Authors tend to orient themselves along the inte-rior–exterior dimension as well. Authors emphasiz-ing the role of the exterior environment often havedifficulties to explain why firms facing similar envi-ronments differ. 16 Also a reverse statement is true.The resource-based theories of the firm, which nowattract a great deal of effort, 17 may be tempted toemphasize the interior, although resources do notnecessarily have to be seen as entirely belonging tothe interior. 18 In particular, the existence of re-

14 The old vs. the young Schumpeter is the classic example ofthis, of course.

15 Cf. corresponding notions in topology or in fuzzy set theory.Cf. also the notion of a set of boundary-spanning activities.

16 Since separate different firms belong to each other’s exterior,however, their exteriors differ in principle.

17 Ž . Ž . Ž .See, e.g., Wernerfelt 1984 , Barney 1991 , Grant 1991 .18 Examples of this would be resources shared with external

partners or external relations seen as resources. New informationtechnologies offer increasing possibilities for firms to developresource-sharing relations rather than narrow exchange relationswith the environment, giving rise to notions such as ‘the networkfirm’ or ‘the virtual corporation’ or ‘the imaginary organization’.

sources in some kind of interior is taken as given,which makes resource-based theories similar to neo-classical theory of the firm in their difficulty ofexplaining how the firm once came into existenceand started to develop its resources in the first place.

Emphasis along the interior–exterior dimensionalso tends to reflect on the emphasis given in strat-egy formulations for firms. A strategy may empha-size economizing upon resources on the one hand,their ‘push’ and ‘stretch’, economies of scale andscope, costs etc. and on the other hand economizingupon external opportunities, ‘market pull’ etc. Ofcourse, both views are possible and advisable toreconcile in an iterative, interactive process of strat-

Ž .egy formulation see Section 4 .Business transactions constitute a particular form

of interactions between the interior of the firm andits exterior, giving rise to costs and benefits. Trans-action cost theory traditionally uses the firm and themarket as two distinct types of organizing an econ-omy, which enables one to look at transaction costsfor one type relative to the other as an opportunitycost, subsuming any relative benefits. The permissi-ble range of boundaries of the firm in an optimaleconomic organization could then in principle befound by equalizing the transaction costs for the twopolar types. 19

A particularly important resource, attracting muchattention in contemporary theorizing, is competenceor capabilities. The process of competence-building

Ž .or learning could be seen partly but only partly asresulting from interaction between the interior andexterior. 20 The notion of core competence, advo-

Ž .cated by Hamel and Prahalad 1994 and PrahaladŽ .and Hamel 1990 , is closely linked not only to the

notion of an interior of the firm, but to the notion ofsome central part of this interior. 21 In passing one

19 To the extent that the modern firm has developed into aquasi-integrated hybrid of the pure polar types of organizations,i.e., markets vs. hierarchies, as mentioned in Section 1, one canquestion the empirical validity of this dichotomy. This in turndoes not invalidate the transaction cost concept, but it severelycomplicates its use as an opportunity cost concept.

20 Ž .Cf. the notion of learning by interacting in Lundvall 1992 .21 The term ‘core’ is etymologically related to the Latin word

for heart. ‘Core competence’ may be defined somewhat ambigu-Žously but is used mostly in the sense of something central not

.only distinctive in the interior of a firm.


may also note that centrality notions of this sort alsotend to be associated with notions about continuityand stability, and may even invite static thinking. 22

A particularly important competence in turn isŽ . Žmanagerial or organizational competence or capa-

. 23bility . Managerial competence is necessary tomobilize, combine and exploit other resources inresponse to business opportunities. Viewed as a re-source it differs fundamentally from all other re-sources since it is the only resource that is trulynecessary for a firm, functionally as well as legally. 24

2.2. DiÕersification of firms

An important feature in the evolution of firms isŽdiversification, that is, in a general and a bit am-

.biguous sense the firm’s process as well as state ofexpansion into a wider variety of activities. Diversi-fication along various dimensions is characteristic offirms, as both specimens and species. Typically theconcept of diversification among scholars refers toproduct diversification, but it is sometimes also usedto refer to market diversification, of which interna-tionalization would then be a special case. Interna-tionalization and the MNC have received a great dealof attention, 25 product diversification and the MPCless so, at least until fairly recently. 26 This is sur-

22 Cf. the notion that there is a certain industry the firm operatesŽ .in and should define for strategic purposes, with Levitt 1960 as a

classic proponent. Similarly, notions of a fundamental corporateŽidentity or timeless vision invites rigidity in the long run cf. the

.well-known C&C concept in the Japanese NEC .23 ŽAgain several distinctions are possible managerial, organiza-

.tional, economic, entrepreneurial etc. but not of primary impor-Žtance at this stage, except to say that neither technology s

.technical competence nor managerial competence should be sub-sumed one under the other, if catch-all concepts are to be avoidedŽ .cf. Eliasson, 1994a .

24 Self-organizing entities devoid of any central management areŽconceivable just as there are seemingly self-organizing communi-

.ties of social insects like an ant colony or a bee-hive , for examplea small cooperative or a network organization, but some kind ofhierarchy with some kind of management is still likely to developbeyond a certain scale for various reasons.

25 For good reviews of the MNC literature, see, e.g., DunningŽ . Ž . Ž .1988a,b , Cantwell 1989, 1994 , Casson 1990 and BuckleyŽ .1989 .

26 Ž .Penrose 1959 is a classic pioneering work on the subject.For good literature reviews, see Ramanujam and VaradarajanŽ . Ž .1989 and Montgomery 1994 .

prising since the MPC has long been the rule ratherthan the exception, and more so than the MNC infact. 27 Nevertheless, neo-classical theory of the mul-tiproduct, multifactor firm appeared before any spe-cific theory of the MNCs. However, it was not reallyuntil the 1980s that other types of economic MPC-theorizing emerged, after the introduction of a formalconcept of economies of scope by Pantzar and Willigin the mid-1970s, the revived interest in transaction

Ž .cost theory triggered by Williamson 1975 and thegeneral aftermath of the diversification failure waveof the 1970s. 28

Both MNCs and MPCs are reconcilable with mostreceived theories, be they transaction cost-based,resource-based or evolutionary. 29 Such current theo-ries offer some general explanations and predictionsregarding MNCs and MPCs. The challenge ahead isto make them more specific and precise.

2.3. Technology and management in theories of thefirm

The distinction between technology and manage-Žment is fundamental for our purposes here. For an

overview of definitions of these concepts, see.Granstrand, 1982 . Technology and management as

two separate concepts have implicitly or explicitlyfeatured in theories of the firm since long ago; seeTable 1 for an overview. However, differences intechnology and management in and among firms

Žover time including technical and managerial inno-vations and learning, and a skewed distribution oflevels of technology and management among firms

.at any point in time have not featured in theory-making until fairly recent attempts. Thus one may

27 A possible reason behind the relative preoccupation withMNCs may be their closer links to policy issues, at least in market

Žeconomies. For a planned economy of the old Soviet type, theMPCs in form of so-called MNTKs were probably receiving more

.attention than the MNCs .28 Ž .A good example of such theorizing is Teece 1982 . A classic

Ž . Ž .empirical work on MPCs is Rumelt 1974 , just as Hymer 1976was for MNCs.

29 Exceptions would be the OLI-paradigm of Dunning for MNCs,Ž .a paradigm which does not easily lend itself to incorporating

MPCs, and neo-classical theory of MPCs, a theory which does notnaturally extend to MNCs.

()

O.G

ranstrandrR

esearchP

olicy27

1998465

–489

470

Table 1Typical representations of technology and management in theories of the firma

Ž .Theory central authors Focal concern Basic categoriesrfactors Representationrrole of techno- Representationrrole of manage-logy factor ment factor

Ž .Neo-classical NCT Optimal factor input and product Factor inputs; Product outputs; Production function Flawless, costless profit maxi-output decisions in shortrlong- Production function; Profit, Factor input mizerrun profit maximization prices, costs

Mathematical programming Optimal activity level decisions Resource inputs; Product out- Resource consumption coeffi- Flawless, costless profit maxi-Ž .MP for maximizing output or profit puts; Activity levels; Profit, cients mizer

under resource constraints prices, costs Factor inputŽTransaction cost Coase, Existence and boundaries of Transaction cost as a function of Process technology-oriented Embedded in transaction costs

. Ž .Williamson TCT firms in a market economy uncertainty, bounded rationality, Reducible to transaction costsopportunism, information asym-metry, asset specificity

Ž .Principal agent PAT Relation between resource own- Contractual design for the orga- Implicit as just any source of Explicit maximizers of different,Ž .ers principals and resource nization facing asymmetries of asymmetries interdependent utility functions,

Ž .managers agents information and incentives incurring a management cost toamong owners and managers of owners

Ž .resources propertiesŽEvolutionary Nelson, Win- Evolution of the firm Selection mechanisms, variety Innovative and imitative R&D as Embedded partly in routines,

. Ž .ter ET generation, routines, dynamic randomizer and variety generator partly in entrepreneurial activi-processes etc. ties

Ž . Ž .Resource-based RBT Competitive success and evolu- Resources e.g., competences As a particular competence As a particular competencetion of firms

Ž .OLI Dunning MNC existence and behaviour Advantages of MNCs stemming Embedded in ownership Embedded in internalizationfrom ownership, location and in-ternalization

a The table is not complete. Various variants and extensions of the theories are ignored here, and only a sample of illustrative concepts is given.


claim that although there has been a co-evolution oftechnology and firms, there has been no correspond-ing co-evolution of technology and theories of thefirm. In all fairness, one must admit that technologyhas penetrated the firm more in practice than intheory. 30

The technology factor is increasingly attended toin economic theorizing, however, but the manage-ment factor much less so. While its role in practice ispervasive, its role in theory is evasive. Part of thedifficulty stems from the lack of good representa-tions and operationalizations of managerial compe-tence. While no one doubts the presence of techno-logical evolution, many doubt whether there hasbeen any evolution of management, although argu-ments for the latter evolution could be made andcases of managerial innovations apart from merefads could be demonstrated. 31

The consequences of assuming a managerial evo-lution in general, and differential managerial learningamong firms specifically, are vast, however. 32 Theproblem is how to represent management theoreti-

30 A similar observation may be made regarding technology andŽ .other economic theories Granstrand, 1994a, Ch. 1 . Sidney Win-

ter’s words apply here as well: ‘‘ . . . the economy is much better atchanging itself than economists are at changing their minds.’’Ž .Winter, 1988, p. 179 .

31 Ž .See, e.g., Granstrand 1982, Ch. 12 . Managerial evolution ascumulation of managerial knowledge should be distinguishedfrom a stream of managerial fads, a stream that for variousepistemological and utilitarian reasons is characteristic of manage-ment as a field. The concepts of cumulation and evolution in thisas well as in other contexts should then not be interpreted asprocesses with strictly monotonic increases at every point in timetowards some ‘higher forms’, but rather with increases in somesense on average over long periods of time.

32 Ž .One example is given in Granstrand and Sjolander 1990 ,¨where it is argued that managerial learning regarding technology

Ž .transitions shifts stabilizes industrial or market structure undercertain conditions. To give another example, suppose managerial

Ž .learning decreases the governance management costs of the firmfaster than any learning associated with recurrent contracting onthe market. In a dynamic transaction cost theory, this wouldexplain and predict gradual internalization over time better thanwould any theory based on comparative static transaction costs asin the usual TCT. Introducing dynamic transaction costs formallyis not necessarily troublesome. A simple analogy with the tradi-tional learning curve could serve as an illustration. If TC is theN

transaction cost for the Nth contract in recurrent contracting, onecould let TC be a function of N, e.g., TC sTC Nyl , with lN N 1

a parameter depending upon type of contracting context.

cally 33 and how to relax assumptions in receivedtheory without letting the management factor be acatch-all concept.

Finally, in brief terms of the categories in Table1, the theoretical approach in this paper has as itsfocal concern the technology-based firm, its forma-tion, existence, growth and diversification. Basic fac-tors are especially resources and their characteristics,notably technology and management, managementcost and economic characteristics of resource-related

Žprocesses economies of scale, scope, speed and.space , static as well as dynamic. Technology is

represented as a body of knowledge, together withphysical characteristics of its embodiments. As aresource, technology has certain idiosyncrasies, giv-ing rise to specific and strong economies. Similarlymanagement is a body of knowledge, together withbehavioural characteristics of its embodiments. As aresource, management is viewed as a specific meta-

Žresource. As a set of ‘behavioural algorithms’ in-.cluding routines as well as random elements , man-

agement is costly and imperfect but has a property ofself-improvement.

3. Empirical findings and analysis

3.1. The key role of technology diÕersification incorporate economics

This section summarizes the empirical findings ofa research project on the economics and managementof large, technology-based firms in Europe, Japan,and the US, carried out in 1987–1994 by the authorand his colleagues. 34 The project consisted of sev-

33 In formal optimization models, management could be seen asan algorithm or a set of algorithms operating at several levels,including a meta-algorithm for composition and decomposition ofthe firm as a multi-period, multi-level optimization problem. Firmbehavior in game-theoretic modeling could further be seen asinteracting management algorithms for multi-objective optimiza-tion.

34 The contributions of S. Alange, S. Jacobsson, C. Oskarsson¨and S. Sjolander are gratefully acknowledged. A series of papers¨

Ž .and reports have resulted, notably Granstrand et al. 1989 ,Ž . Ž .Granstrand and Sjolander 1990 , Granstrand et al. 1992a ,¨

Ž . Ž .Granstrand and Oskarsson 1994b , Oskarsson 1993 , GranstrandŽ .and Sjolander 1994c . For overviews, see especially the latter¨

two.


eral sub-studies, altogether covering by interviews,questionnaires and published data a sizable sampleof the world’s largest technology-based corpora-tions. 35

Major empirical findings of the project were asfollows.

1. Technology diversification at firm level, i.e.,the firm’s expansion of its technology base into awider range of technologies, was an increasing andprevailing phenomenon in all three major industrial-ized regions, Europe, Japan and US. This finding has

Ž .also been corroborated by Patel and Pavitt 1994 .The prevalence of technology diversification as aphenomenon in technology-based firms prompted the

Ž .notion of a multi-technology corporation MTC as,roughly expressed, a firm having diversified into

Žseveral technologies i.e., technical competence ar-.eas .2. Technology diversification was a fundamental

causal variable behind corporate growth. This wasalso true when controlled for product diversificationand acquisitions. 36

3. Technology diversification was also leading togrowth of R&D expenditures, in turn leading to bothincreased demand for and increased supply of tech-nology for external sourcing.

4. Technology diversification and product diversi-fication were strongly interlinked, often in a pull–push pattern in economically successful firms, lead-ing to hybrid MPCrMTCs.

5. The high-growth corporations followed a se-quential diversification strategy, starting with tech-nology diversification, followed by product andrormarket diversification. This result was independentof region and industry.

35 The main interview and questionnaire study covered 14ŽJapanese large corporations Hitachi, NEC, Toshiba, Canon, Toy-

. Ž .ota etc. , 20 European Ericsson, Volvo, Siemens, Philips etc. andŽ .16 US IBM, GE, AT&T, GM, TI etc. , representing, e.g., 38% of

total US industrial R&D in 1988. Analysis of published data, asŽ .reported in Oskarsson 1993 , covered 57 large OECD corpora-

tions. Additional case studies of companies and product areashave been conducted as well.

36 This finding has later been confirmed also by GambardellaŽ .and Torrisi 1997 for 32 of the largest European and US electron-

ics firms.

6. Japanese companies had on an average themost developed managerial capability for concertedtechnology, product and market diversification.

These findings were not readily explainable interms of received theories of the firm, as reviewed inSection 2. Without going into detail about the prosand cons in using received theories to describe,explain and predict the behaviour of technology-based firms, one can note that technology diversifica-tion, being a central feature in the empirical findings,does not feature at all in received theories. More-over, most theories do not explicate the dynamicsand the heterogeneity of technology, and many re-

Žstrict their focus to process technology see, e.g.,Milgrom and Roberts, 1990, for an otherwise novel

.approach .

3.2. Theoretical analysis of technology diÕersifica-tion at firm leÕel

ŽAt a given point in time an MPCrMTC by. Ždefinition comprises several products or busi-

.nesses and technologies, comprising the productbase and the technology base respectively for thecorporation. These bases are linked to each other,

Ž .e.g., as in Table 2 built on an actual case . The moretechnology-product couplings, the more synergiesthere may be. Here one could note, for example, thefollowing.

Table 2The technologyrproduct matrix of a corporation

Technology Product business

P P P P P P P P P P1 2 3 4 5 6 7 8 9 10

Electronic hardware X X X X X X X X X XManrmachine communication X X X X X XElectronic packaging X X XMicro-mechanical design X X X X X XSoftware X X X X XComputer communication X X XElectro-optics X X X X XInfrared technology X X X XMicrowave technology X X XLaser technology X XHydro-acoustics X XImage processing X X XArtificial intelligence X X X X X XSystems technology X X X X X X X X


.a The technology base of product P is highly2

technologically diversified, while for P it is not.9.b Electronic hardware is a generic technology,

while hydro-acoustics has a narrow applicability inthe product base of the corporation.

How has an MPCrMTC developed such a cou-pled technology-product structure? Generally speak-ing, it may of course result from radical restructuringdone for various reasons, but here we will focus on asecond route with more organic evolution. 37 For agiven product, new customer requirements and tech-nological opportunities create a need for makingtransitions in the technology base of a product, whichleads the company to develop or acquire new tech-nologies through product-related technology diversi-fication. Once acquired, these technologies may thengive the company an incentive to engage in technol-ogy-related product diversification through

Ž .economies of scale and scope see further below .This type of ‘first pull then push’ interplay betweentechnology diversification and product diversifica-tion, simply speaking, results from ‘market pull’,

Žfollowed by internal ‘technology push’ or ‘resource.push’ .

Fig. 1 illustrates this stepwise corporate evolutionŽin principle over time with time points t to t and1 6

an arrow denoting entry into a TP-combination and †.denoting exit from a TP-combination . The com-

bined technologyrproduct base thus shifts over time,just as the customer base may do in accordance.Product invention-based companies in electrical andmechanical engineering then often display a kind of‘rooted’ diversification, sticking to the original prod-uct business area, while diversifying into others. Akind of ‘floating diversification’ on the other hand iscommon for a raw material-based company or achemical company over long periods of time, in the

Ž .sense that the company diversifies away i.e., exitsfrom its original product business area. For this typeof companies, physical by-products in extraction andprocessing also play an important role in the diversi-fication process.

Technology diversification as an empirical phe-nomenon has only recently attracted an interest

37 This route has been typical for diversification in Japanesecorporations.

Fig. 1. The product businessrtechnology matrix of a companyŽ .over time t .

among researchers, and then primarily outside theeconomic tradition, with pioneering works by PavittŽ . Ž . Ž . 381988 , Pavitt et al. 1989 and Kodama 1986a,b .The apparently key role played by this variable inthe economics of technology at corporate level, asdescribed above, is a new finding for which anyexplanation at this stage must be tentative.

Regarding causes and consequences of technologydiversification, tentative modeling is presented in

Ž .Granstrand and Sjolander 1992b and Granstrand et¨Ž .al. 1992a , which emphasizes progress in S&T

together with differentiation of both S&T fields andmarket needs. This view is closely related to theview that an expanding set of opportunities, techno-logical as well as business opportunities, is a majorfactor in a number of phenomena. The notion ofopportunity has attracted a great deal of scholarly

Ž .interest since long ago. Schumpeter 1976 discardsthe notion of a set of investment opportunities van-ishing over time with reference to ever-changing

Ž .capital-consuming technology. Penrose 1959 treatsŽtechnology and industrial R&D as one out of sev-

.eral important sources of new opportunities forŽ .product diversification in general. Scherer 1980

identified technological opportunity as the most im-portant factor behind differences in innovativeness

Ž .between different industries. Jaffe 1989 , in tryingto quantify technological opportunities, argued thatR&D and technology must be looked upon as con-

38 Incidentally no reference is made to it in surveys by Ramanu-Ž . Ž .jam and Varadarajan 1989 and Montgomery 1994 . The connec-

tion between R&D and product diversification has been dealt withŽ .in literature, however, with Nelson 1959 as a pioneering theoret-

Ž .ical work, later verified by Link and Long 1981 .


sisting of a number of distinct technological areas.This view is consistent with the frame of reference in

Ž .Granstrand and Sjolander 1990 which introduced¨the notion of a multitechnology corporation, acquir-ing and exploiting a variety of technologies. In fact,it may be argued that technological opportunities aregenerated in a fundamentally important and inex-haustible way through the combination and recom-bination of various technologies, new as well as old.Of course, far from all combinations are technicallyandror economically feasible, but the technologicalopportunity set might still grow rapidly and evenprogressively, due to the exponential growth of thenumber of combinatorial possibilities in principle.Such a process of combinations and recombinationscould be considered to lie at the heart of the inven-tion and innovation processes, in which technolo-gists, managers, and markets filter out technicallyand economically infeasible combinations.

Theoretically, in the process of taking advantageof technological opportunities, technology diversifi-cation at the corporate level may lead to increasedsales in four different and complementary ways.Firstly, there are static economies of scale to theextent that the same, or close to the same, technolo-gies could be used in several different products withminor adaptation costs. Because exploiting knowl-edge in various applications is characterized by rela-tively smaller variable costs per additional applica-tion in relation to the fixed cost of acquiring theknowledge, the static economies of scale are signifi-cant when a technology has a wide applicability tomany different product areas in a corporation. This isthe case for generic technologies by definition. Sec-ondly, knowledge is not consumed or worn out whenapplied, but it is typically improved through learningprocesses when applied several times, which allowsfor dynamic economies of scale. Thirdly, differenttechnologies have a potential to cross-fertilize othertechnologies, yielding new inventions, new function-alities and increased product andror process perfor-mances when combined, regardless of whether thetechnologies in question have a wide applicability tomany product areas or not. This cross-fertilizationthen yields what could be called true economies ofscope, which are not the kind of economies of scopethat arise from shared inputs, and are special cases ofeconomies of scale. This third type of economy,

potentially associated with technology diversifica-tion, depends on specific technologies which couldbe combined or integrated, and also vary over time,depending upon the intra-technology advancements.Fourthly, combining technologies mostly requires

Žsome technology transfer, and under certain condi-.tions intra-firm technology transfer is faster and

more effective than inter-firm, giving rise to speedand timing advantages in an MTC, advantages thatwill be labelled economies of speed here.

According to the empirical findings above, tech-nology diversification leads not only to sales growthbut also to growth of R&D expenditures. Tenta-tively, this is because when a larger number oftechnologies is involved, a larger amount of coordi-nation and integration work is needed, apart from thecost of acquiring each new technology, as difficultiesarise in connection with conducting multidisciplinaryR&D. These difficulties are widely reported andtypically involve conflicts between professional sub-cultures in science and technology, NIH-effects and

Ž . 39other innovation barriers see Granstrand, 1982 .Thus, in order to reap net benefits from technologydiversification leading to growth of both sales andR&D expenditures, the integrative capabilities ofboth technologists and managers become decisive.

Finally, it should be noted that the causal relation-ship between R&D costs and sales has been widelyobserved, although with no recognition of the impor-tance of technology diversification. In brief, success-ful R&D with a time lag tends to lead to salesgrowth on an average. In some corporations there isalso a tendency to let annual R&D budgeting deci-sions be influenced by a certain fraction of past

39 If N is the number of important technologies, some of whichhave mutual relations giving rise to combination costs, the R&Dcosts at product level could specifically be hypothesized to dependquadratically on N, that is, R&D costss aq bNqcN 2. If someof all the possible interdependence relations among the N tech-nologies are assumed to contribute roughly equally to R&D costs,then these would hypothetically grow exponentially, i.e., R&D

Ž .costss a)exp bN , as a result of technology diversification. Theoften observed progressive rise in R&D costs in a product areaover time, or over successive product generations, could thus beexplained, at least partially, by technology diversification, with theshape of progressive growth being somewhere in between quadraticand exponential growth.


years’ sales, which sets up a causal relationshipbetween past sales and future R&D costs. Thus,there is an observed tendency for R&D costs andsales income to mutually reinforce each other overtime but with different time lags and differentstrengths in the causal couplings. Moreover, compar-isons of R&D intensities among competitors, whichare commonly made in some industries when makingR&D budgets, may yield industry-wide similaritiesin R&D intensity. However, such practices mayhave to be reconsidered when taking the conse-quences of technology diversification and technol-ogy-related product diversification into account.

4. Elements of proposed theory

The purpose of this section is to put the empiricalfindings and their tentative theoretical explanationsin Section 3 into a more general theoretical frame-work for the technology-based firm, and therebydraw as well on received theory as briefly recapitu-lated in Section 2.

This can only be a first step towards a theory ofthe technology-based firm. As such it departs primar-ily from resource-based theory but also from evolu-tionary theory and systems theory in general aspects.

4.1. The enÕironment-resource-management Õiew ofthe firm in general

An outline of the proposed theory of the technol-ogy-based firm is as follows. A business firm ingeneral is viewed as a legally identified, dynamichuman system, consisting of a set of heterogeneous

Žresources in an institutional setting defining, e.g.,.property rights . The firm has an interior and an

Ž .exterior or environment subjected to managementŽand business ideas with a dynamic goal or incen-

.tive structure having commonalities for coordinatingpurposeful action. The interior interacts with theexterior in various ways, in particular through busi-

Ž .ness transactions economic exchange on a market,taken as a network of meetings between buyers andsellers. The business transactions are essentially ex-changes of resources between the firm and its envi-ronment, typically production factors and products

Ž .in a wide sense, including any services for money.A business is defined here as a set of businesstransactions, which is coherent in some sense interms of resources, products and markets.

As the system evolves, resources are transformedas a result of interior operations or activities as wellas of interaction with the exterior. In particular,resources are acquired and exploited through bothinterior operations and business transactions. To theextent that resources are acquired and exploited bythe firm through business transactions, one can speakof them as inputs and outputs of the firm respec-tively. However, resources could also be classified asinputs and outputs of a resource transformation pro-

Ž .cess production process in a wide sense inside thefirm, not necessarily directly linked to businesstransactions with the exterior.

The environment must be explicitly recognizedsince it provides business opportunities without whichthe firm will atrophy. At the same time a firm facesenvironmental challenges or threats and may gobankrupt or disappear through merging or acquisitionor liquidation. This normally results from interactionwith the environment, which thereby provides notonly opportunities but also threats. The opportunitiesand threats in the environment are changing, partlyinfluenced by the firm, partly beyond its influence.The influence depends on the firm’s management

Ž .and other resources, which in turn are influencedby the environment and its past, current and antici-pated interaction with the firm. The major strategictask for management is to position the firm in thestream of opportunities and threats, having a suitableresource structure that can also be transformed overtime to meet new opportunities and threats. 40

In a TBF, then, all of the elements in the generalview of the firm have technology-oriented compo-nents. Thus, technologies constitute a vital part ofthe firm’s resources. A substantial part of the firm’sinteractions with the environment are influenced byinternal and external technical and technological

40 This formulation does not imply that opportunities are fullyexogenous to the firm. For example, a firm that comes up with aninnovation may create entirely new opportunities, some of whichalso spill over to other firms.


changes, opportunities as well as threats, and thecorresponding part of the environment istechnology-oriented and constitutes the firm’s tech-nological neighbourhood. Moreover, technologymanagement is a vital part of management and tech-nology-oriented ideas and goals are a vital part of thefirm’s business ideas and goal structure, althoughtypically more or less aligned to economic goals.

So far there is not much new about this view ofthe firm in general as a system, which is dynamic,self-organizing, resource-transforming, interacting,human, purposeful etc. 41 Nor is it yet clear that theTBF is a particularly interesting special case fortheorizing, although empirical evidence points in thatdirection. It is when the structure of the resource setand its associated processes of acquisition and ex-ploitation are specified that something new and use-ful can hopefully be achieved. Next this specificationwill be made, or at least forwarded a bit.

4.2. The resources of a firm

4.2.1. General resource structureThe resource set is considered as the firm’s capi-Ž . 42tal or capitalized assets , and is decomposable

Žinto physical, financial and immaterial or intangible. Ž .or intellectual capital IC . Immaterial capital plays

a key role, and encompasses both disembodied ICŽ .including business ideas and embodied IC, thelatter in the form of relational capital and compe-

Ž . 43tences or capabilities possessed by humans. Dis-embodied IC is partly protectable by intellectual

Ž .property rights IPRs , including trade secret protec-Ž .tion whether weak or strong of business ideas.

Table 3 gives an overview of the resource categoriesŽ .used here, which in fact is one way out of several

41 This does not mean that there is nothing disputable about sucha view.

42 At this level of analysis, concepts such as resource, capital,and asset could be used interchangeably.

43 Several terms are usable and distinctions could be introducedbut, at this level of analysis, terms such as competences, capabili-

Ž .ties, abilities, skills, knowledge and so on can be used roughlyinterchangeably. Also the terms immaterial, intangible and intel-lectual could be used interchangeably in this connection althoughsome would argue that intellectual capital has a more narrowconnotation, than immaterial capital.

to present a structure of asset categories on a bal-ance-sheet. 44

IPRs constitute an important part of IC. 45 TheIPRs are both registered and unregistered. The com-pany know-how in the form of trade secrets is animportant part of unregistered IPRs. This know-howis partly disembodied human competence. The em-bodied human competence is part of the company’sIC, controlled through employment contracts. Thus,it has to be looked upon as leased human capital orIC rather than human capital owned by thecompany. 46

Ž .The capital structure resource structure differsacross companies and sectors and may then be usedfor constructing taxonomies of, e.g., companies, e.g.,classifying them into raw material-based, IC-inten-sive, knowledge-based, technology-based etc. Thecapital structure of a company also is reflected in its

Žculture and management style and vice versa, of.course . Thus, organizational features and manage-

Žment skills are differentiated across sectors forestry,.banking, pharmaceuticals etc. , and management as

an asset acquires a certain specificity to other assets,as well as to the local environment. Such adaptations

44 Needless to say, there are large problems in valuation of IC.This is a contemporary subject for accounting research. Theproblems are not insurmountable and there are several approachesbeing tried to take a fuller account of IC. The traditional balancesheet in double-entry accounting could be extended, of course, butit could also be complemented by separate IC balance sheets,essentially with non-consolidated multiple value concepts. Thelatter is probably a more viable approach. If nothing else, variousbalance sheets could be used continually in internal accounting, aswell as for valuating the entire firm in connection with anacquisition on the stock market.

45 The importance of IPRs is often downplayed with reference tothe traditionally limited value of patents and the weak enforce-

Ž .ment of IPRs in general at least until the 1980s . Nevertheless,Žthe quoted value of top-level brand names is high indeed ca. 39

.BUSD for Marlboro in 1992 and the value to the firm of its tradesecrets may be high, despite weak legal enforcement of theirassociated rights.

46 The original concept of human capital, pioneered by BeckerŽ .1964 , pertains to individuals. The results produced by the humancapital typically become the property of the company when it isdisembodied, however. For example, inventions made by R&Dpersonnel and any associated patents may be stipulated in theemployment contract to be the company’s property.


Table 3Resource categories of a firm

aŽ . Ž . Ž .Material tangible Immaterial intangible IC

Physical capital Financial capital Intellectual Good-will and HumanŽ .property rights power in embodied

Ž .Disembodied IC internalrexternal competenceŽ .relations capital

Žrelationalb.capital among

Natural resources Liquid capital Patents Employees ManagerialRaw materials DatabasesBuildings Bonds Know-how Customers TechnologicalMachinery Shares Licenses Suppliers CommercialWork in progress Securities Brand names Competitors FinancialInventories etc. Designs Universities Legaletc. Software Investors Manual

Copyrights Interest etc.Concessions organizationsetc. Societies

etc.

a Ž .Exactly what IC should encompass is debatable but it should definitely include IPRs as well as human competence or capital or capabilityand good-will. For simplicity, IC will here be taken to comprise all immaterial or intangible resources or assets, admittedly with someconceptual borderline problems.b Ž .It is possible to distinguish the category ‘relational capital’ including, e.g., trust and internal motivation as well as ‘organizational capital’Ž .or ‘capability’ from IC. Here we rather use these categories as sub-categories of IC.

of resources in general give rise to specific comple-mentarities and concomitant difficulties in separateresource transfer and trade.

4.2.2. Acquisition and exploitation of resourcesThe processes by which resources are acquired

Žand exploited including processes for generation,combination, transformation, regeneration and

.recombination of resources vary widely across com-panies and sectors, of course, and give rise to varia-tions in economic performance, in turn affectingfurther acquisition of resources. Resources by them-selves then have different intrinsic economic proper-ties that make their acquisition and exploitation pro-cesses different from each other, just as their trans-formations could be characterized in economic termsin various ways. We will distinguish between fourgeneral types of process related economies involved:

Žstatic and dynamic economies of scale in the ordi-. Žnary sense of declining average cost , scope in a

. Žbroad sense of synergies , speed advantages of abso-. Žlute and relative pace of a process and space ad-

.vantages of location .

4.2.3. Management as a meta-resourceIt is not sufficient to view the firm simply as a set

of resources without any reference to the way inwhich this set came into existence and then evolved.In order to get a business firm started operationallyŽ .not only legally , there must be some businessopportunity in the environment and a business ideain some person’s mind of how to exploit this oppor-tunity. The business idea, which may be treated as a

Ž .piece of IC e.g., a patent or a trade secret and theŽ .person as an entrepreneur manager constitute the

initial necessary resources.Managerial competence is also part of the firm’s

IC and is a decisive resource for reaping differenttypes of economies for the firm’s formation, sus-tained existence and development. Management 47

47 No essential distinction is made at this level of the analysisbetween management, entrepreneurship, leadership, administrationetc. Neither is a distinction made here between ownership andmanagement. Ownership can rather be viewed as management at aspecific managerial level, when need arises to make this admit-

Ž .tedly important distinction as in PAT .


acquires, combines and exploits resources in re-sponse to business opportunities in the firm’s exte-

Žrior, thereby performing a control function in a wide.sense in the system. In that sense, management

could be viewed as a meta-resource. As such, man-agerial competence is a unique resource that is notsubstitutable in its entirety. The use of this resourcetakes time and is associated with a cost, which wecan call management cost. 48 Such a concept ofmanagement cost can then be used as an alternativeto a transaction cost concept when only two gover-nance structures are used, i.e., markets and hierar-chies, and it can be used as a complementary analyti-cal concept in a general case. What makes manage-ment so special as a resource is that it is a resourcewhich is of unique decisiveness for any firm, andthat it is a resource which has to be ‘self-sourcing’and ‘self-allocating’ in some sense. That is, manage-

Ž .ment has to decide by itself owners apart howmuch effort to allocate to different tasks. This deci-sion in turn requires some effort, and thus a theoreti-cally difficult recursive problem of mental economyarises, usually resolved in practice by attending totime limits and a dominating need for correctiveaction rather than calculation.

However, management cannot be viewed as ameta-resource, incurring a management cost, solelyin a traditional rational economic perspective. Vari-ous behavioural characteristics at individual and or-ganizational level have also to be considered, just as

Ždone in TCT with bounded rationality and oppor-.tunism . To illustrate, a behavioural characteristic

among entrepreneurs of importance for transactioncost considerations in general is what we can calltheir ‘entrepreneurial hubris’ giving them a bias to-wards overvaluing their IC, i.e., their business ideaand managerial competence. 49 This means that theytend to perceive a market failure even if there isnone regarding the firm’s initial capital value. This

48 The occurrence of managerial evolution, innovations andlearning serves to lower this cost, just as new technologies maydo, e.g., new information and communication technologies; seeSection 2 above.

49 There are other dimensions of entrepreneurial characteristicsas well, of importance for the formation and existence of a firm,such as need for achievement, need for autonomy and need forpower.

Ž .behavioural bias ‘economies of hope’ tends to keepthe firm together, as well as tending within limits ofcourse, to develop the firm, giving it more ‘animalspirits’ and ‘sweat equity’ in the pursuit of en-trepreneurial goals and acts of will. That is, this typeof expectational bias among entrepreneurs, which inthat sense is to be regarded as a managerial defi-

Žciency or ‘management failure’, to allude to the.market failure concept , increases corporate coher-Žence and sustainability. Again within limits. Often

there is too much of entrepreneurial hubris, then.risking disaster and dissolution of the firm . Addi-

tional influence from any market failure, e.g., regard-ing market valuation of knowledge or complementar-ities, only serves to reinforce corporate coherence.

Ž .The latter is thus explainable at least partly byŽreference to market failure or management failure in

.the above sense or both, each factor being in itself afavourable condition in principle for coherence, pro-vided it is suitably biased.

Managerial competence is moreover to be re-garded as a bounded, difficult to codify, non-protec-table by patents, dynamically evolving, heteroge-neous resource, especially in a large corporation,with many sub-competences pertaining to different

Žmanagerial areas, tasks and functions marketing.management, financial management etc. . Technol-

ogy management, then, is of particular concern in aTBF.

4.2.4. Knowledge properties in generalŽ .Since knowledge competence, capability plays a

key role as an IC resource, its intrinsic economicproperties need to be recognized. This is often doneand there are various ways to characterize theseproperties, the characterization below being just one.However, it is important to make a fairly completecharacterization of properties of economic relevancein order to limit discussion of epistemological issues.

Ž .Thus, knowledge competence embodied in hu-Ž .mans 1 consumes much time and effort to acquire

Ž .it, especially at individual level; 2 consumes littleŽ .time and effort to use it, once acquired; 3 improves

and cumulates through use without deteriorationŽknowledge is limitlessly reusable or inexhaustiblewithout wear, implying non-rivalry in consumption,

.and possibly with negative depreciation , but maydeteriorate without use, creating a need for mainte-


Ž .nance; 4 is irreversibly transferable to others, if it issuitably codified and adapted to the recipients’knowledge, while still being kept by the knowledge-

Žholder and possibly even improved through ‘learn-.ing by teaching’ . Thus knowledge is cheaply cloned

Ž .or reproducible if codified; 5 is impossible to bedispossessed of, for an individual, and difficult todispose of, i.e., to unlearn or scrap. Machines arescrappable, but knowledge is not, thereby inducing

Ž .irreversibility of its advances; 6 often has strongŽ .complementarities with other knowledge parts; 7 is

easily rendered obsolete by new knowledge parts,Ž .appearing over time; 8 is possible for an individual

Ž . Ž .to keep in almost perfect secrecy control; 9 isimpossible to distribute quite equally among agents;Ž .10 is generally more expensive to generate than to

Žregenerate or imitate e.g., knowing that something.is possible, or exists, cheapens the search for it and

the generation of knowledge is uncertain and filledwith surprises.

Ž .Thus, compared to physical resources capital ,knowledge resources have some fundamentally dif-ferent properties, like being inexhaustible, irre-versibly produced and transferred, due to a dispos-session impossibility, and reusable and reproducibleat no or low cost. Economic and legal concepts and‘laws’ for physical resources thereby could be ex-pected to apply in some respects fundamentally dif-ferent from knowledge resources.

Acquisition and exploitation of knowledge arethus characterized by a large fixed initial investmentcost with a small variable cost in application. Thevariable cost may even be negative sometimes, be-cause of learning by using the knowledge. The ‘tech-nical’ lifetime of knowledge is infinite and, throughlearning by using, its depreciation is negative as longas it is not rendered obsolete by other knowledge.

Thus knowledge has strong economies of scale,both static and dynamic, and has increasing returnsin use as long as it is not obsolete, thereby creating a

Ž .risk of lock-in i.e., the competence trap . Because ofthe scope of complementarities among its parts,knowledge also often has strong economies of scope.Moreover it is not time-consuming to use, once

Žacquired, and it is highly mobile if codified al-.though there may be multiple, incompatible codes .

Ž .The codification or disembodiment process may betime- and effort-consuming, however, and the less

Ž .knowledge is codified into one language , the moretime and effort are needed to transfer it amonghumans. 50 Science and R&D are important in orderto improve the codification process, but are by nomeans the only ways to do so.

Codification is important not only for facilitatingtransfer of knowledge per se, but also to enable itscumulation. Cumulation takes place within an indi-

Ž .vidual and a group through use of memories , amongŽ .humans through transfer and between human gen-

erations, and results in a common pool or stock ofknowledge. 51 This pool may partly be publiclyavailable, and some knowledge may adopt a ‘publicgood’ characteristic. 52

All in all, knowledge has properties that give it agreat economic potential as a resource, even uniquelygreat in some respects. However, there are alsoproperties that strongly limit exploiting this eco-nomic potential. As is well known, knowledge isdifficult to value, price and sell. This is partly be-cause knowledge is highly differentiated, whichmakes it difficult to match demand with a competi-tive supply, and partly because knowledge may eas-ily be stolen with little chance of IPR enforcementŽ .cf. Arrow’s information paradox, Arrow, 1974 .New ideas and knowledge, which may be difficult tocodify and specify partially without risking full dis-closure, are especially vulnerable to theft and inad-vertent diffusion. New ideas and knowledge can bekept perfectly secret by an individual for cost-free

Ž .eternity, torture aside but then benefits from learn-ing by using the new knowledge and complementari-ties from combining it with the knowledge of othersare lost.

Thus, the intrinsic properties of knowledge createan economic potential, at the same time as theseknowledge properties create limits to knowledge ex-ploitation, not least to exploitation through market

50 Throughout the text here, both time and effort are empha-sized. This is because there are very limited possibilities to make

Ž .trade-offs between time and efforts or costs . That is, in learningand teaching you can buy time through increased efforts only upto a point, due to speed limits.

51 The terms ‘pool or stock of knowledge’ are deceptive sincethey are often used together with the term ‘flow’ which indicates aphysical transfer rather than diffusion or cloning of knowledge.

52 This does not mean it is a free good in the sense that it isŽ .costless to acquire and use see Nelson, 1992 .


exchange mechanisms. This gives a natural rationalefor having some kind of system in an economywhich stimulates knowledge exploitation without toomany negative side-effects and costs for ‘running’ it.In fact, if a system could be designed that stimulatesnot only knowledge exploitation but also its acquisi-tion, the system would be doubly effective. This isindeed what the IPR system is intended to accom-plish regarding technical knowledge.

4.2.5. Special properties of technologyTechnology is a special kind of knowledge and as

such has, in addition to the general knowledge prop-erties above, special properties, not all of them sharedby other types of knowledge. These properties in-clude the following.

.1 An artifact link, i.e., technology is linked toŽ .artifacts materials, products or systems of artifacts

and to the processes by which they are produced.These artifacts and their production processes arepossible to characterize by physical design and per-formance parameters, which typically evolve overtime as their underlying technologies develop. Acertain technology may, moreover, be linked to manyartifacts, i.e., having a wide applicability or beingmulti-purpose, just as a certain artifact may be linkedto many specific technologies, i.e., being multitech-nological.

.2 A science link, i.e., technology is linked tonatural sciences and their methodology.

.3 A relatively high degree of codifiability ingeneral, partly stemming from the links to artifactsand the links to natural sciences, through the use of

Žformulae in a formal language mathematical, chemi-.cal, computational etc. , drawings, models, patent

documents, textbooks, and a scientifically orientedlanguage. Also the artifacts serve as codifiers ofparts of technical knowledge, the embodied technol-

Žogy. However, in certain areas e.g., new andror.less science-based there is an important tacit knowl-

edge component as well. The tacit component is notstatic, since codification is a dynamic process, linkedto R&D and scientification of technologies. On theother hand, R&D uncovers new technologies with,sometimes, a low degree of codification initially.High codifiability facilitates the specification andtransfer of technical knowledge as well as its accu-mulation.

.4 A ‘practical purpose’ link, i.e., technicalknowledge is generated largely with an intention tohave something working in practice or have sometechnical performance level achieved.

.5 Links to globally oriented common systems forŽ .its operationalization specification and assessment,

especially the patent system and the educationalsystem but also systems for standardization, testing,regulation, classification etc.

An implication of these idiosyncrasies of techni-cal knowledge is that it is easier to have a system forregistering technical knowledge, due to its highercodifiability and its artifact link. The patent system issuch a system and it also stimulates the codificationof technology. At the same time, as its main purpose,

Žit stimulates technology generation, diffusion clon-. 53ing and technology cumulation. The patent sys-

tem thus reinforces some of the knowledge proper-ties of technology, giving rise to economies of scale,

Ž .scope and mobility speed and space of technology,although at the possible cost of some R&D andoutput market distortion plus the cost of administer-ing the patent system. Thus we have still anotherspecial property of technology, namely that it isŽ .6 Possible to protect by patent rights.

The heterogeneity of technology must be empha-sized, although this property is not unique to technol-ogy as a body of knowledge. Technology beingheterogeneous means that different technologies and

Žtypes of technologies may be identified not without.difficulties and combined, and that their economic

properties typically differ. These properties of a spe-cific technology, or combination of technologies,pertain to its genericness or degree of applicability,customer utility impact and production cost impact,mediated through the corresponding technical perfor-mance parameters, cost of acquisition, complemen-tarities, substitution effects, potential for advance-

53 Sometimes, however, complementarities are not gained sincetechnology exploitation may also be hindered rather than fosteredby a fragmentation of IPRs among agents that raise transactioncosts even to the point of hindering exchange through market

Ž .transactions. One example of this is the digital audio tape DATtechnology; another may occur in what is now termed multimedia,in which case numerous interests in media, telecom and comput-ing industries converge and conflict.


Žment, excludability or protectability e.g., through.patents and state and rate of diffusion on the market.

Needless to say, technologies are dynamicallyevolving. Technological changes that give rise toinnovations at a closer look typically then involvesome new combination of partly old technologies.An empirically important phenomenon is the increas-ing technology diversification at business level, giv-ing rise to multitechnology products and processes,in turn giving rise to multitechnology corporations astypical TBFs. At the same time, fundamental break-throughs in S&T give rise to generic or multipur-pose technologies. Altogether, these two phenomenacreate more and more technologyrbusiness cou-plings with associated economies of scale and scope,possible to reap by active and innovative manage-ment.

Finally, it is important to distinguish technologyfrom management. Both technology and manage-ment are knowledge resources and as such bounded,heterogeneous and dynamically evolving and, viatheir embodiments, both affect transformations ofother resources. However, management primarily haslinks to human resources, behavioural characteristicsand social sciences. Compared to technology, man-agerial knowledge is less codifiable, more localizedand non-protectable by patents.

4.3. DiÕersification

4.3.1. Types of diÕersificationA firm can be viewed as composed of one set of

Ž .businesses or productrmarket combinations , con-stituting its business base, and one set of resources,constituting its resource base, and a many-to-manycorrespondence between resources and businessesŽwith a standard production function as a very spe-

.cial case , subjected to environmental changes andmanagement and organizational behaviour. A firmmay engage in two fundamental types of diversifica-

Žtion—business diversification with product, ser-.vice and market diversification as special cases and

Žresource diversification with, e.g., technology di-. 54versification as a special case . Resource and busi-

54 Such a firm could be labeled multi-business, multi-resourceand corresponds to the multiproduct, multifactor firm in neo-classical theory.

Fig. 2. Shifts in the resource and business base over time.

ness diversification corresponds to input and outputdiversification, from the point of view of the firm asa whole. Note that ‘business diversification’ heretypically refers to business on the firm’s outputmarkets, while diversification on the firm’s inputmarkets is subsumed under resource diversification.This is an essential distinction in the argument be-low, saying that the interaction between these twodiversification processes is one important source ofdynamics in the evolution of the firm.

Each business of the firm has in turn a resourcebase, and each type of resource may be exploited inseveral businesses. These links can be summarized in

Ž .a resourcerbusiness matrix see below , analogousto the productrmarket matrix commonly used indescribing diversification. 55

As the firm evolves over time, its resource baseand business base may shift, with some resources R ,i

businesses B , and mutual BR-couplings scrappediŽ . Ž .or substituted , some kept conserved and someadded. See Figs. 2 and 3 for two graphic representa-tions of this phenomenon.

Strictly speaking, the concept of diversificationrefers to the ‘adding part’ of these shifts. As a basisfor a typology we can then distinguish between theprocesses of substitution or exit, conservation, anddiversification or entry, pertaining to both businessesand resources. Any change in the business base as

55 Ž .In fact, a key contribution by Wernerfelt 1984 was toemphasize the resource perspective in strategy-making, whichpreviously had commonly focused only on productrmarket diver-

Ž .sification with internationalization as a special case .


Fig. 3. Shifts over time in the resourcerbusiness matrix.

well as in the resource base then involves some or allof these processes, which gives a number of combi-

Žnations or types some of which are logically infeasi-.ble, though . Sequences of changes of different types

may be further considered, which gives additionalŽ .possibilities for typification see further below . For

example, over long periods of time the resourceandror business base may shift entirely and becomedisjoint from what they originally were. As men-tioned in Section 3, this may happen to rawmaterial-based firms and chemical firms with ‘float-

Ž .ing’ diversification as in Figs. 2 and 3 , whileinvention-based firms in mechanical and electricalengineering rather display ‘rooted’ diversification,with the original business and product area remain-ing, at least for a long period of time until it isscrapped, thereby turning the diversification into‘floating’. Sometimes a stage of mainly resourcediversification, driven by ‘demand pull’ andeconomies of scope, is followed by a stage of mainlybusiness diversification, driven by ‘resource push’and economies of scale, which is then followed by a

Žstage of business and resource scrapping see Section.3.1 . However, different types of diversification in a

firm do not necessarily have to evolve in a stage-wiseor sequential manner but can be conducted concur-rently. Concurrent diversification puts larger de-mands on managerial resources, though, and sequen-tial diversification has historically been more com-mon. 56

56 Ž .Cantwell and Piscitello 1996 argue that, for the largestEuropean and US firms, technological diversification and interna-tionalization of technological activity have historically occurredsequentially, while a new contemporary complementarity betweenthem has been emerging since the 1980s.

The rate at which the resource and business baseshifts over time is in principle possible to measureby using a distance measure between two sets, possi-bly overlapping, be they sets of resources or busi-nesses. 57 An innovation typically leads to changes inthe resource andror business base, and to the extentthat resources and businesses are kept, compared tothose that are scrapped and added due to the innova-tion, the change is gradual rather than radical. Tush-

Ž .man and Andersson 1986 make the distinctionŽ .between competence-destroying i.e., scrapping and

Ž .competence-enhancing i.e., adding technologicalinnovations. Many if not most technological innova-

Žtions shift the competence base including the tech-.nology base without inducing disjointness, and with

Ž .an enhancement diversification bias, i.e., with moreŽadding than scrapping of competences see

.Granstrand, 1994a, Ch. 7 . In addition, several tech-nologies in the technology base are usually retainedor conserved.

Thus, to the extent that technology conservationand diversification dominate over technology substi-tution, the resulting change becomes more gradualthan radical, which is in line with the cumulation and

Ž .continuity view of the firm see Section 2 .Generally speaking, then, the size and rate of

change in the resource and business base implied byany diversification move provide a basis for distin-guishing between, on the one hand, continuousrin-crementalrrelated etc. diversification and, on theother hand, discontinuousrradicalrunrelated re-source andror business diversification. 58

4.3.2. DiÕersification dynamicsAs the firm is commonly conceived it derives its

dynamic features primarily through interacting with

57 Ž .See Granstrand 1994a, Ch. 7 for such a measure defined ona technology base.

58 This latter type is also referred to as conglomerate diversifica-tion. As is well known, most studies show the poor economicperformance of such diversification, although it could be argued

Ž .that perfect capital markets e.g. should lead to neither economiesnor diseconomies. However, conglomerate or unrelated diversifi-cation is in practice related to some extent through the use ofcommon managerial resources. Management failure in the form ofmanagerial or entrepreneurial hubris mentioned earlier then easilycomes into play, as does the principal agent problem arising frommanagement seeking job security in corporate diversification.


its environment. There are other sources of dy-namism as well, fairly well recognized. One is man-

Žagement in a broad sense, including entrepreneurial.acts in the organization as a whole to the extent that

it is not entirely adaptive to environmental changes.Another is endogenous innovations, rather than adap-tations. However, two additional sources of dy-namism need to be recognized as well.

The first is the increasing returns that accrue inthe resource acquisition and exploitation processes

Žper se, in particular learning in a broad sense,including learning in recurrent contracting, leading to

.reduced transaction costs . The second source is theinteraction between resource and business base shifts,diversification in particular. As resources are ac-quired to support a specific business, some of theseresources may have multiple uses with propertiesthat improve the economic prospects of going into anew business, e.g., through the provision ofeconomies of scale and scope. 59 This in turn mayrequire that new resources are acquired as well.

Moreover, resources already acquired for a busi-ness that is later scrapped may be difficult to scrapor dispose of in a market in the short term, leading topossibly sunk fixed costs, in turn creating an incen-tive to use surplus resources for business diversifica-tion. 60

A resource may also be lost or scrapped for otherŽ .reasons e.g., loss of key people or concessions ,

leading to scrapping a business, thereby possiblyreleasing other resources with alternative uses and soon. These dynamics in scrappingrkeepingraddingbusinesses and resources with alternative uses are inreality influenced by the firm’s interaction with itsenvironment, of course, but it is important to note

Žthat external dynamics is a sufficient at least in the.long run but not a necessary condition for internal

dynamics. That is, a firm could in principle go on

59 This case has been well recognized in literature. PenroseŽ . Ž .1959 is definitely a pioneering work, followed by Teece 1982

Ž .and Chandler 1990 and others. Normally, reference is made toresource slack, indivisibility and under-utilization.

60 Ž .One example being diversification induced by life or longtime employment, e.g., in Japan.

Ž . Ž .scrapping substituting and adding diversifyingŽ . 61even in a static let alone stable environment.

The internal diversification dynamics is driven bythe economic properties of resources and their trans-formation, notably of four types considered here:economies of scale, scope, speed and space. Theseeconomies derive from the physical or intrinsic prop-erties of resources, in conjunction with the many-to-

Ž .many transformation or production in a wide sensecorrespondence between resources and businesses.Different diversification patterns then put differentrequirements on management in reaping the relevanteconomies involved. Too much of simultaneous re-source and business diversification, as in conglomer-ate diversification, may then overtax management aswell as other resources. Commonly observed se-quences of corporate evolution like internationaliza-

Ž .tion market diversification followed by resourcediversification, spurring subsequent product diversi-fication, can then be explained by dynamicallychanging mixes of economies of scale, scope, speedand space. Concurrent diversification, e.g., more orless simultaneous internationalization and productdiversification, may then be achieved through com-plementarities, giving rise to economies of scope;resource sharing, giving rise to economies of scale;managerial learning, removing certain diseconomiesof scale; a premium on time to certain markets; plus

Ž .locational economies economies of space . Knowl-edge resources in general and technology in particu-lar then have properties, many of which are unique,that provide a strong form of dynamics to the firmand its diversification process.

4.3.3. Technology-related diÕersificationSection 3 above described how technology diver-

sification had been empirically found to be the pri-mary variable leading to growth of the technology-based firm, while at the same time leading to growthin the firm’s R&D expenditures. The sales growthinduced by technology diversification can be hypoth-

61 Of course, there are environments that would not allow this tohappen, but the point is that if there are admissible environmentsthey do not need to be dynamic.


esized to result from four sources: static as well asdynamic economies of scale in applying technologiesacross products and markets, i.e., across businesses;

Ž .economies of scope in a broad sense in combiningdifferent technologies; and economies of speed inacquisition and transfer of technologies in the firm.High sales growth was particularly associated with asequential diversification strategy, with technologydiversification followed by technology-related busi-ness diversification. The growth of R&D expendi-tures resulting from technology diversification washypothesized to derive from the cost of new tech-nologies plus the cost of overcoming difficulties incombining various technologies, i.e., a kind of disec-

Ž .onomies of scope but not congestion . Increasinglyexpensive R&D needed to support existing busi-nesses then gives an incentive for technology-relatedbusiness diversification to economize upon theŽ .quasi-fixed R&D investments, be it through prod-uct diversification or market diversification or both.The relative failure of MPCs vs. MNCs observed inliterature could then be hypothesized to result from a

Žhigher degree of technology-relatedness implying.scale, scope as well as speed economies in MNCs,

everything else equal. In addition, internationaliza-62 Žtion may provide locational economies or

.economies of ‘space’ in relation to R&D. Sincediverse S&T activities tend to agglomerate in vari-

Ž .ous regions of the world Cantwell, 1989, 1994 anMTC gains further locational advantages by locatingR&D and technology-sourcing activities in such‘multitechnology regions’.

Skeptics of business diversification, of whom thereare many, might object that technology diversifica-tion does not necessarily have to lead to technology-related business diversification. The need for tech-nology diversification is largely generated by forcesexogenous to the firm. As empirical studies haveshown, many firms also grow through technologydiversification without engaging in business diversi-fication. However, as R&D expenditures growthrough technology diversification, a need arises torecover them through expanding their business base.Many firms have in the past responded to this by

62 This corresponds to the L factor in Dunning’s OLI-paradigm.

market diversification, especially internationaliza-tion, thereby becoming MNCs. Internationalizationof TBFs has usually concerned the firm’s outputmarkets, but increasingly internationalization ofR&D also takes place.

An alternative, sometimes complementary andsometimes not, depending on limited managerial ca-pabilities, is to engage in technology-related productdiversification and spread R&D costs over severalproduct areas. Compared to internationalization,however, this puts other and seemingly larger de-mands on management.

Another general response to rising R&D costs inTBFs is the development of markets for technologywith increasing division of R&D labour amongTBFs. Rising R&D costs, in conjunction with gener-icness of new technologies, increase both supply anddemand for them on a technology market. Empiricalstudies have also shown an increase in externalsourcing of new technologies in TBFs. More generalrationalization of R&D work is also possible, e.g.,through cost-reducing innovations in the R&D pro-

Ž .cess itself e.g., in software development .Still another possibility for mitigating the rise in

R&D costs so common to TBFs is to engage intechnology-related partnering, either pure R&D part-nering or technology-related business partnering. Infact the growth of partnering among TBFs of allsizes is to a considerable extent motivated by risingR&D costs, influenced by technology diversifica-tion. Again technology-related partnering puts spe-cial demands on management, general managementas well as technology management, but in this case itis management linked to several firms, which re-quires a quite different type of quasi-integrated gov-ernance structure.

Thus, some predictions for TBFs, based on risingR&D costs through partly exogenously driven tech-nology diversification, are that they will respond inone or more of five ways: internationalization, tech-nology-related business diversification, external tech-nology sourcing as well as technology marketing,rationalization of R&D, and technology-related part-nering.

In summary, the economic properties of technol-ogy as a resource create, through technology diversi-fication and technology-related business diversifica-tion, an economic potential for the firm. Other physi-


cal and intellectual resources may also provide aneconomic potential in similar ways, of course, but itis argued here that technology has some unique andparticularly strong properties in these respects, in-

Žcluding the associated tendencies toward costly and.risky R&D. That physical capital and intellectual

capital have some fundamentally differing economicproperties is clear, but how then does technologydiffer from other IC resources? Don’t other compe-tences or knowledge resources provide economies ofscale and scope much as technology does? Yes, theydo, but as described in Section 4.2 above, technologyhas some unique properties which lead to a particu-larly strong economic potential. For example, it is

Ž .possible to embody ‘productify’ technology in arti-facts to decrease transactional hazards, and more sothan in the case of pure knowledge-based services.The possibilities to codify can be argued to be largeron an average for technology than for other compe-tences, which improves the prospects of cumulationand transfer, which in turn improves the prospects ofreaping economies of scale, scope, speed andspace. 63

ŽS&T discoveries and inventions e.g., radio.waves, photoconductivity, the laser have a feature

of indivisibility, in the sense that the magnitude withŽ .which they are made acquired is far from fully

determined by the need for them in a particularsituation of resource utilization. This property theyshare with discoveries in other knowledge areas in

Žgeneral cf. the discovery of the American continent.by explorers and exploiters . Mother Nature simply

has a way of revealing herself in chunks, not alwaysproportional to the effort of her explorer. However,

Žthe patent system is unique to technology not sci-.ence and technical inventions, and gives by design a

unique economic potential to technical advances be-

63 The whole issue of codifiability could be elaborated at length,Žbut must be omitted here see above and also Arrow, 1974 and

.Nelson, 1992 . Suffice it to say that codification is a dynamicŽ . Ževolutionary process, driven partly by R&D cf. the importancein this respect of contributions by Lavoisier and Mendelejev in

.chemistry . As a technology matures, it becomes more codified—in fact the level of codification could even be taken as a definingcharacteristic of maturing, until new discoveries challenge the

Žaccepted code, thereby temporarily disrupting cumulation cf..Kuhn, Lakatos, Popper .

Ž .yond a certain technical as opposed to economicsize or level of invention. The indivisibilities inducedby the patent system yield possible slack andeconomies of scale, but this may be a minor consid-eration compared to the role of the patent system inproviding potential economies, e.g., of speed, for thepatent holder relative to competitors, through theordinary temporary, restricted monopoly protectionassociated with patent rights. 64,65

ŽThe role of patents in firms and economic devel-.opment in general is an issue which could be elabo-

rated at great length. Most authors do not regard therole of patents as very important, however. E.g.,

Ž .Chandler 1990 attributes a secondary role to themin firm formation historically. On the other hand,there are not many studies with this focus. At thesame time a ‘pro-patent’ era has emerged in the1980s as IC has become generally more important.And insofar as IPRs as a whole are concerned withtrademarks, trade secrets etc., their historical role hasbeen larger than commonly recognized.

4.3.4. The management factorThus, omitting further elaborations here, we may

infer that technology has properties which createstrong economic potentials. Some of these propertiesare shared with other resources, in particular withother competences, while some properties are unique.Some properties are intrinsic in the nature of tech-nology, some are man-made. The IPR system inparticular is part of the institutional setting that bydesign endows technology with unique economicproperties, although it is debatable how strong theyare.

However, the general economic potential of tech-nology is not realized automatically, nor are itsreturns automatically appropriated, not even partlyby a firm which has invested in acquiring it. In order

Žto accomplish that, management in a broad sense,

64 If the indivisibilities are small they could be taken to ‘con-vexify’. The number of patents related to a product is often largeand also increases empirically. At the same time, it may beclaimed that the genericness of many patents increases.

65 The timing properties induced by the patent system are no-table. First, there is a ‘winner takes all’ race to get a patent

Ž .granted; second, single patent protection expires fully after afixed time interval.


.including organizational capabilities is a necessarybut not sufficient competence. This is not the properplace to elaborate upon how it can be achieved, morethan to indicate a few challenges to current manage-rial thinking, based on the empirical findings and thetheoretical analysis above. First, a specializationstrategy of ‘sticking to the knitting’ could be severely

Žcriticized on the above grounds, for one thing see.Granstrand et al., 1992a . Second, in order to reap

technological economies of scale, scope, speed andspace, technology transfer and integration are ofdecisive importance in acquiring and exploiting tech-nology, often with in-house R&D as a dominantmode of acquisition. This opposes the current ap-

Ž .proach at least in the West that favours one-sideddecentralization of R&D and technology manage-ment. Moreover, down-sizing middle technologymanagement in that connection may jeopardize criti-cal integrative functions for reaping the economicbenefits of technology diversification.

Hence, the management factor must explicitlyenter into a theory of the technology-based firm andits diversification processes as a variable intermedi-ate between technological and economic changes atfirm level. Management is then not only to be repre-sented by its embodiment in humans, expressing

Žitself in managerial performance of which the yard-sticks are much poorer than for technical perfor-

.mance , but also to be represented by a knowledgearea, subjected to evolution through learning, cumu-lation and innovations just as for technological evo-lution. At the same time, managerial knowledge and

Žpractices co-evolve with new technologies e.g., new.computer and communication technologies with an

interplay between technological and managerial in-novations. Management serves as an explanatoryfactor together with the technology factor not onlybehind the formation, sustained existence and diver-sification of the technology-based firm, but alsobehind differences in economic performance of firmsŽ .exogenous factors apart , and in an even widerperspective as a partial explanation of the strength inthe co-evolution of S&T and the firm as an institu-tion. Unfortunately for many economic theories,management has somewhat paradoxically been a big,dark box, perhaps even bigger than the black box oftechnology. Fortunately for economic research, muchexciting work thereby remains to be done.

5. Summary and conclusions

As an economic institution the firm has developeda strong form of co-evolution with S&T, with in-creasing importance of TBFs as well as firm-basedtechnologies. Empirical studies have shown that theMTC is the rule rather than the exception amongtechnology-based firms, and that technology diversi-fication is a primary variable leading to growth ofthe firm as well as to growth of its R&D expendi-tures, often leading subsequently also to technology-related product diversification.

ŽReceived theory of the firm has with some ex-.ceptions not particularly addressed technology-based

firms and their management. This paper has pro-posed an approach for theorizing about such firms byfocusing on technology as part of the firm’s intellec-tual capital and general resources, interacting with anenvironment, under active and innovative manage-ment. The firm’s resources, including managementas a special part, can be acquired and exploited invarious ways. Technology then has intrinsic andpartly unique properties leading to particularly strongeconomic potentials through any or all of four basic

Žtypes of economies involved—scale static and dy-. Ž . Ž .namic , scope in a broad sense , speed and timing

Ž .and space location and mobility . The IPR system isone but only one source of these unique properties,and serves to strengthen the economies associatedwith technology.

The approach for theorizing is first to characterizea firm in general and a TBF in particular, next togive descriptions and explanations of why TBFsform, grow and diversify in various ways, and finallyto offer some predictions.

A firm in general is taken to be a human systemcharacterized by its dynamic and heterogeneous re-sources, its environment, with an institutional set-ting, its internal and external interactions, includingtransactions, its business ideas, its goal structure andits management. Resources are being purposely ac-quired, transformed, and exploited under manage-ment as an active and even innovative control func-tion, giving rise to a mix of the four basic types ofeconomies, those of scale, scope, speed and space.Management could partly be seen also as a resourcefor the firm, but then as a uniquely decisive meta-re-source with very special features. For the formation


of a firm, management and a viable business idea arenecessary. For the sustained existence of the firm,the net management cost must also be perceived tobe lower on average than net market transactioncosts.

A technology-based firm is then a firm for whicheach of these firm-characterizing elements containsor is influenced by technology and technical artifactsin a vital way in some sense. Technology is charac-terized by both its ‘soft’ and ‘hard’ sides, being adynamic body of knowledge, strongly linked to

Ž .changing technical artifacts typically products , withphysically measurable technical attributes, impactingin uncertain and innovative ways user and producerutilities and goals, as well as the firm’s businessinteractions and competitive situations. Technologycould partly be seen as a resource, but only partlysince it enters into both the input and output charac-terizations of a firm, and influences competition onboth input and output markets. As a knowledgeresource, being part of the firm’s intellectual capital,technology shares the general characteristics ofknowledge, but in addition displays unique features,such as being protectable by patents. These uniquefeatures give technology particularly strongeconomies of scale, scope, speed and space, whichhowever need active and innovative management tobe reaped within a TBF, putting emphasis on thequalities of its technology management.

Empirical studies have also shown the importanceof technology diversification for the growth of aTBF, which then typically evolves into a MTC.However, technology diversification also leads togrowth of R&D expenditures, which induces theTBF under competition to engage in one or more offive strategies: internationalization, technology-re-lated business diversification, external technologysourcing and marketing, R&D rationalization, andfinally technology-related partnering with others, in-cluding other TBFs. The latter leads to the creationof quasi-integrated systems or families of firms, inorder to reap technology-related economies of scale,scope, speed and space.

In relation to received theories of the firm, thetheoretical approach offered here has similarities anddifferences. Various perspectives and basic cate-gories are similar to the ones in received theory, notonly resource-based theory and evolutionary theory,

but also transaction cost theory and neo-classicaltheory. The main difference is the emphasis andexplication of technology and its idiosyncrasies aswell as the emphasis on management, including tech-nology management.

The central role of technology diversification inthe evolution of a technology-based firm, then be-coming multitechnological, is moreover a new per-spective with strategic implications for management.The strong potential economies of scale, scope, speedand space associated with combining heterogeneoustechnologies and their characteristics, the realizationof which depends upon the qualities of management,are thus central explanatory categories. In particular,the relative superiority of probably less than perfecttechnology management over conducting transac-tions on imperfect technology-related markets is de-cisive.

Finally, in line with the emphasis on technology,it must be remembered that technology as well asfirms are such pervasive and complex phenomenathat there are long, probably never-ending, paths totake in further theorizing about their co-evolution.To mention one direction for further research: asfirm-based technologies increasingly develop, theo-rizing about their evolution is also called for. Afterall, Mother Nature has ‘managed’ to reveal herself instrange ways, be it to firms or other institutions.

List of abbreviations

C&C Computer and communicationsET Evolutionary theoryIC Immaterial or Intellectual capitalIPR Intellectual property rightMNC Multinational corporationsMP Mathematical programmingMPC Multiproduct corporationMTC Multitechnology corporationNCT Neo-classical theoryNIH ‘Not invented here’OLI Ownership–Location–InternalizationPAT Principal agent theoryR&D Research and developmentRBT Resource-based theoryS&T Science and technologyTBF Technology-based firmTCT Transaction cost theory


References

Arrow, K.J., 1974. The Limits of Organization, W.W. Norton andCo., New York.

Barney, J., 1991. Firm resources and sustained competitive advan-Ž .tage. Journal of Management 17 1 , 99–120.

Becker, G.S., 1964. Human Capital, The University of ChicagoPress, Chicago.

von Bertalanffy, L., 1968. General System Theory. Foundations,Development, Applications, George Braziller, New York.

Buckley, P.J., 1989. The Multinational Enterprise. Theory andApplications, MacMillan, London.

Cantwell, J., 1989. Technological Innovation and MultinationalCorporations, Basil Blackwell, Oxford.

Ž .Cantwell, J. Ed. , 1994. Transnational Corporations and Innova-tory Activities. UN Library on Transnational Corporations,Vol. 17, London, Routledge, Introductory chapter.

Cantwell, J., Piscitello, L., 1996. The Diversification and Interna-tionalization of Corporate Technology: From a Strategic ChoiceHistorically to a Complementary Combination Today, Mimeo,University of Reading, UK.

Casson, M.C., 1990. Enterprise and Competitiveness: A SystemsView of International Business, Clarendon Press, Oxford.

Chandler, A.D., Jr., 1990. Scale and Scope. The Dynamics ofIndustrial Capitalism, Harvard University Press, Cambridge,MA.

Dosi, G., Teece, D., Winter, S., Toward a Theory of CorporateCoherence, Preliminary Remarks. In: Dosi et al.

Ž .Dosi, G., Giannetti, R., Toninelli, P.A. Eds. , 1992. Technologyand Enterprise in a Historical Perspective, Oxford UniversityPress, Oxford.

Dunning, J.H., 1988. Explaining International Production, UnwinHyman, London.

Dunning, J.H., 1988. Multinationals, Technology and Competi-tiveness, Unwin Hyman, London.

Ž .Eliasson, G., 1994. In: Granstrand, O. Ed. , Technology, Eco-nomic Competence and the Theory of the Firm.

Fransman, M., 1994. Information, knowledge, vision and theoriesŽ .of the firm. Industrial and Corporate Change 3 3 , 713–757.

Gambardella, A., Torrisi, S., 1997. Does technological conver-gence imply convergence in markets? Evidence from the

Ž .electronics industry, Research Policy forthcoming .Granstrand, O., 1982. Technology, Management and Markets,

Frances Pinter Publ., London.Granstrand, O., Sjolander, S., Alange, S., 1989. Strategic technol-¨ ¨

ogy management issues in Japanese manufacturing industry.Ž .Technology Analysis and Strategic Management 1 3 , 259–

272.Granstrand, O., Sjolander, S., 1990. Managing innovation in¨

Ž .multi-technology corporations. Research Policy 19 1 , 35–60.Granstrand, O., Bohlin, E., Oskarsson, C., Sjoberg, N., 1992a.¨

External technology acquisition in large multi-technology cor-Ž .porations. R&D Management 22 2 , 111–133.

Granstrand, O., Sjolander, S., 1992. Internationalization and diver-¨sification of multi-technology corporations. In: Granstrand, O.,

Ž .Hakanson, L., Sjolander, S. Eds. , Technology Management˚ ¨and International Business. Wiley, London.

Ž .Granstrand, O. Ed. , 1994a. Economics of Technology, Elsevier,Amsterdam.

Granstrand, O., Oskarsson, C., 1994b. Technology diversificationin ‘MUL-TECH’ corporations. IEEE Transactions on Engi-

Ž .neering Management 41 4 , 355–364.Granstrand, O., Sjolander, S., 1994c. Managing innovation in¨

multi-technology corporations. Chapter contribution to Dodg-Ž .son, M., Rothwell, R. Eds. , Handbook of Industrial Innova-

tion, Edward Elgar.Grant, R.M., 1991. The Resource-Based Theory of Competitive

Advantage—Implications for Strategy Formulation, CaliforniaManagement Review, pp. 114–135.

Hamel, G., Prahalad, C.K., 1994. Competing for the Future, HBSPress, Boston, MA.

Hymer, S., 1976. The International Operations of National Firms,MIT Press, Cambridge, MA.

Jaffe, A., 1989. Characterizing the technological position of firms,with application to quantifying technological opportunity andresearch spillovers. Research Policy 18, 87–97.

Kodama, F., 1986a. Japanese innovation in mechatronics technol-ogy—Fumio Kodama studies technological fusion. Science

Ž .and Public Policy 13 1 , 44–51.Kodama, F., 1986b. Technological diversification of Japanese

industry. Science 233, 291–296.Levitt, T., 1960. Marketing myopia. Harvard Business Review 38

Ž .1 .Link, A.N., Long, J.E., 1981. The simple economics of basic

scientific research: a test of Nelson’s diversification hypothe-Ž .sis. Journal of Industrial Economics 30 1 , 105–109.

˚ Ž .Lundvall, B.-A. Ed. , 1992. National Systems of Innovation.Towards a Theory of Innovation and Interactive Learning,Pinter Publishers, London.

Milgrom, P., Roberts, J., 1990. The economics of modern manu-facturing: technology, strategy, and organization. American

Ž .Economic Review 80 3 , 511–528.Montgomery, C.A., 1994. Corporate diversification. Journal of

Ž .Economic Perspectives 8 3 , 163–178.Nelson, R., 1959. The Simple Economics of Basic Scientific

Research. Journal of Political Economy, pp. 297–305.Nelson, R., 1992. In: Dosi et al., The Roles of Firms in Technical

Advance: A Perspective from Evolutionary Theory.Nelson, R., Winter, S., 1977. In search of a useful theory of

innovation. Research Policy 6, 36–76.Nelson, R., Winter, S., 1982. An Evolutionary Theory of Eco-

nomic Change, The Belknap Press, Cambridge, MA.Ž .Oskarsson, C., Doct. diss. 1993 . Technology Diversification—

The Phenomenon, Its Causes and Effects, Department ofIndustrial Management and Economics, Chalmers Universityof Technology, Goteborg, Sweden.¨

Patel, P., Pavitt, K., 1994. Technological Competencies in theWorld’s Largest Firms: Characteristics, Constraints and Scopefor Managerial Choice. STEEP Discussion Paper No. 13,Science Policy Research Unit, University of Sussex.

Pavitt, K., 1988. Uses and abuses of patent statistics. In: vanŽ .Raan, A. Ed. , Handbook of Quantitative Studies of Science

Ž .and Technology. Amsterdam, Elsevier North-Holland .Pavitt, K., Robson, M., Townsend, J., 1989. Technological accu-


mulation, diversification and organisation in UK companiesŽ .1945–1983. Management Science 35 1 .

Pavitt, K., 1991. Key characteristics of the large innovating firm.British Journal of Management 2, 41–50.

Penrose, E., 1952. Biological analogies in the theory of the firm.American Economic Review 42, 804–819.

Penrose, E., The Theory of the Growth of the Firm, Oxford, BasilBlackwell and Mott, 1959, 5th impression, 1972.

Prahalad, C.K., Hamel, G., 1990. The core competence of theŽ .corporation. Harvard Business Review 68 3 , 79–91.

Ramanujam, V., Varadarajan, P., 1989. Research on corporatediversification: a synthesis. Strategic Management Journal 10,523–551.

Rumelt, R.P., 1974. Strategy, Structure and Economic Perfor-mance, Harvard University Press, Cambridge, MA.

Scherer, F.M., 1980. Industrial Market Structure and EconomicPerformance, 2nd edn. Cambridge, MA.

Schumpeter, J.A., 1976. Capitalism, Socialism and Democracy,Ž .George Allen and Unwin, London, Originally published 1943 .

Teece, D.J., 1982. Towards an economic theory of the multiprod-uct firm. Journal of Economic Behaviour and Organization 3Ž .1 , 39–63.

Tushman, M., Andersson, P., 1986. Technological discontinuitiesand organisational environments. Administrative ScienceQuarterly 31, 439–465.

Wernerfelt, B., 1984. A resource-based view of the firm. StrategicManagement Journal 5, 171–180.

Williamson, O., 1975. Markets and Hierarchies: Analysis andAnti-trust Implications. A study in the Economics of InternalOrganization, The Free Press, London.

Winter, S., 1988. On coase, competence and the corporation.Ž .Journal of Law, Economics, and Organization 4 1 , 163–180.

Towards a theory of the technology-based firm

Documents

Transcript of Towards a theory of the technology-based firm