1. Introduction

nowledge creation processes of technology-basedcompanies have become increasingly global. The

pioneers of R&D internationalization are high-techcompanies operating in small markets and with littleR&D resources in their home country, as it is the casefor ABB, Novartis and Hoffmann-La Roche (Switzer-land), Philips (Netherlands) or Ericsson (Sweden). Asearly as 1986, Dutch and Swiss companies had morelaboratories outside their home countries than inside(Pearce and Singh, 1991). Between 1985 and 1993,overseas investment in R&D by US firms increasedthree times faster as did domestic R&D. In the US,overseas R&D expenses reached 10% of overall R&Dinvestment, up from 6% in 1985 (National ScienceBoard, 1996, pp. 4–44). In the same period, the shareof majority-owned foreign affiliates’ R&D in the USrose from 9% to over 15% (National Science Board,1996, pp. 4–46). In 1991, Japanese multinationalcompanies performed less than 5% of their R&D

abroad (Buderi et al., 1991, p. 85), but the recentestablishment of Japanese laboratories in Europe andthe US increases the significance of Japanese-basedglobal R&D (see e.g. Dalton and Serapio, 1995).According to Buderi et al. (1991), European com-panies performed about one third of their R&D outsidetheir home countries.

Literature on R&D internationalization has longfocused on the economic and political perspective.Most research on this topic has taken place on a macro-economic or sectoral level (see Patel, 1996; Papanas-tassiou and Pearce, 1994; Beise and Belitz 1996, NIWet al., 1995; Amstad et al., 1996; HØakanson, 1981).Management research has long neglected internationalR&D (see Cheng and Bolon, 1993, p. 3; Granstrandet al., 1993, p. 414; Cantwell, 1992, p. 94). A principalissue in R&D internationalization is the locationdecision, which has been studied by e.g. Brockhoff(1997), Pearson et al. (1993); HØakanson (1992);Gerpott (1991); Wortmann (1990); De Pay (1989) andGassmann (1997a) on project level. Although this

R&D Management 28, 3, 1998. © Blackwell Publishers Ltd, 1998. Published by Blackwell Publishers Ltd,108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street, Malden, MA 02148, USA.

147

Organization of industrial R&D on aglobal scale

Oliver Gassmann1 and Maximilian von Zedtwitz2

1Corporate R&D at Schindler Lifts Ltd., 6031 Ebikon/Lucerne, Switzerland, oliver_gassmann@ch.schindler.com2Institute for Technology Management, University of St. Gallen, 9000 St. Gallen, Switzerland,maxi.zedtwitz@unisg.ch

Although the globalization of industrial R&D processes has increasingly attracted managementscientists and considerable work on R&D internationalization has been done since the 70s,important problems in organizing geographically distributed R&D sites have not yet beenresolved. This analysis is based on 165 interviews in 31 technology-based multinationals fromSwitzerland, Germany, USA, and Japan conducted between 1994 and 1997. We describe theextent of R&D internationalization, its key drivers and principal impediments. Two trends areidentified: (i) a geographic concentration on leading technology centres, and (ii) a tightenedcoordination of R&D units. Overlaying structures, superimposed to regional and functionalstructures, play a central role in overcoming current deficits in global R&D management.Organizational constituents of this structure are technology offices (Schering), technologymanagement as a competence center (Schindler), central project offices (IBM), central projectmanagement (Bayer), project oriented career development (Bosch), virtual projectmanagement departments (Roche), and cross-cultural job rotation (ATR).

K

�

research has produced invaluable knowledge thatimmensely enhanced our understanding of the phe-nomenon per se, interviews with R&D managers haveshown that a practical guiding framework for organiz-ing international R&D is still missing.

In our paper we first summarize the motives andtrends in R&D internationalization. We have foundthat many companies struggle with an inadequate R&Dorganization when attempting to draw on the potentialsof global R&D. In a globalizing world traditional R&Dorganization proves to be ineffective for many newchallenges: R&D directors must learn to manage anadditional organizational level in order to ensure globalR&D efficiency. In particular, the geographical distri-bution of R&D exerts a significant influence on thecompany’s overlaying organization structures. In astudy of technology-intensive companies we selectedstate-of-the-art methods for industrial R&D to besuccessfully organized on a global scale. Not surpris-ingly, the management and development of people is acentral aspect in effectuating optimal use of overlayingstructures. In this paper we focus on the structure thatfacilitates and enables the execution of projects and theeffective exchange of knowledge and information. Inthe future, overlaying structures, which are superim-posed and often independent from geographical orhierarchical structures, will become even moreimportant.

2. Research sample

Our research is mainly based on 165 semi-structuredinterviews in 31 technology-based companies between1994 and 1997. All companies are characterizedby a high degree of internationality and operatein  the  following industries: electrical/electronics,automotive/turbines/heavy machinery, and chemicals/pharmaceuticals. We focused on companies that arepioneering the internationalization of their R&Dactivities. Their home bases are in Europe (16), USA(5) and Japan (10).

Table 1 lists the main R&D locations of companiesin our research sample. The data for this research wasgathered in personal interviews with principal repre-sentatives (R&D directors and managers) of eachcompany, following a semi-structured interviewguideline. This data was complemented by deskresearch and feed-back from our interview partners.

3. Trends in R&D internationalization

3.1 International R&D locationsMost of the analysed companies perform a significantamount of R&D abroad (Fig. 1). There is a strongindication for a correlation between technology inten-

148�R&D Management 28, 3, 1998 © Blackwell Publishers Ltd 1998

Oliver Gassmann and Maximilian von Zedtwitz

Table 1. Technology-based companies in our research sample and principal R&D sites.

Company Data of foreign R&D sites: Number of employees (p.), year of establishment

Electrical/electronics industryABB � Corporate research in Switzerland (Baden: 244 p.), Sweden (V×asteras: 254 p.), Germany (Heidelberg: 187 p.),

Italy (Milan), Norway (Oslo) and Finland (Vaasa)� At least 10 additional development centres world-wide, but no R&D in Japan

AT&T � 15 to 20 R&D sites world-wide (Bell Labs in the USA), research laboratory in the Netherlands� Design centres in Japan, Taiwan, Singapore and in several European countries

Bosch � Research and advanced development in Germany only� Development in Switzerland, USA, Brazil, Mexico, Rep. of South Africa, Australia, Spain

Canon � Interactive systems in Surrey (UK), computer networks and software in California, imaging technology inSydney (Australia), telecommunications in Rennes (France), 22 R&D locations in Japan: Shimomaruko (2,200p.), Canon Research Center (250 p.), Fuji Susono (1,000 p). Total of 7,300 R&D researchers

Fujitsu � 4 R&D sites in JapanIBM � Corporate research centres: USA: Hawthorne and Yorktown Heights (1200 p.), Almaden (450 p.), Austin (30

p.), Switzerland: Z×urich (250 p,), China: Beijing (40 p.), Japan: Tokyo (250 p.), Israel: Haifa (330 p.)� About 40 R&D centres world-wide, most of them in the USA, Western Europe, and Japan

Hitachi � Fundamental research in Great Britain (since 1989, Cambridge, micro-electronics) and Ireland (since 1989,Dublin: information science: 50 p.)

� Technical centres for industrial design in Germany (Munich) and Italy (Milan)� Development in the USA (San Francisco; Detroit; Princeton: 50 p.) since 1989� Technical centre in Singapore since 1995

Matsushita � R&D sites in the USA (since 1980)Electric � Software development in Taiwan (1981, 40 p.); development in Singapore (since 1990) and GreatIndustrial Britain (since 1984, 22 p.)

� R&D site in Germany (since 1991)Mitsubishi � Research centre in Great Britain (Cambridge) and in the USA (Sunnyvale)Electric � Decentral development at some manufacturing sitesNEC � R&D laboratory in Bonn, GE (since 1994, 20 p.) and two other labs in the USA: Princeton (90 p.), San Jose

(5 p.)Nokia � Corporate Research in Helsinki (400 p.), Tampere (130 p.), Dallas and Boston (18 p.), Bochum (22 p.).

� total of 8,000 p. in R&DPhilips � Corporate research in the Netherlands (Eindhoven, 1,750 p.), in Germany (Aachen and Hamburg, 450 p.), France

(Limeil-BrÐevannes, 200 p.), Great Britain (Redhill, 200 p.) and in the USA (Briancliff Manor, 300 p.)Sharp � Development centre in the USA, several other R&D sites since 1995

� R&D centre in Great Britain (since 1990, Oxford: 40 p.)� Technical centre in Taiwan

Continued

© Blackwell Publishers Ltd 1998 R&D Management 28, 3, 1998�149

Organization of industrial R&D

Table 1. (Continued)

Company Data of foreign R&D sites: Number of employees (p.), year of establishment

Siemens � Research laboratory in the USA (artificial intelligence and image processing, Princeton: 137 p.). More than 2,600people in the USA (of which 1,000 in software development)

� Several development sites through high-tech company acquisition; significant increase of foreign R&D personnel.Centres of European R&D in Austria (2,600 p.), Switzerland (880 p.), Italy (700 p.), Belgium (650 p.), France,Great Britain, SwedenIn Western Europe, more than 5,500 people have been employed for software development alone

� R&D in Japan only in joint venture companiesSony � Development in Belgium, Germany, and Great Britain (since 1987)

� A total of 7 R&D labs in the USA (e. g. since 1987 in San Jose)� Technical centre in Singapore (1994)

Unisys � USA: total of 3,500 p. in San Jose, Salt Lake City, South California, Roseville, Pennsylvania. Tokyo (3,000 p.),Australia (500 p.), France (200 p.)

Automotive, Turbines, and Heavy MachineryDaimler-Benz � Palo Alto, USA (40 p.). Portland (10 p.), Shanghai (15 p.), Bangalore (14 p.), Japan (4 p.), Moscow (4 p.)

� Central Research in Germany (6 locations, 1,429 p.), several international cooperative projects� Approx. US-$ 5.9 billion of R&D in business units

Schindler Lifts � Research and development of software, cars, drive, safety and system concepts in Ebikon (CH)� Module development and market-driven development at decentralized sites: landing doors in Melune (F), new

door family in Mulhouse (F), electronics in Locarno (CH), hydraulic systems in Stockholm (Sweden)Gettisburgh and Morristown (USA), gears in Zaragoza (Spain).

� Listening post in Japan until 1996. A development centre for cars and field engineering is planned in Ipoh(Malaysia) for 1998.

Sulzer � Research only in Switzerland (Winterthur); central research incorporated as an independent company (Innotech)� Divisional development in 45 sites in 12 countries; Centres in Switzerland, Germany, UK, France, USA

United � Central R&D in the USA (East Hartford)Technologies � Major development centres in Spain (Madrid, 1992), Germany (Aachen, 1994) and Japan (Tokyo, 1992)

Chemicals/Pharmaceutical IndustryBASF � Biotechnology research in Boston (USA)

� Development in New Jersey (USA)� Materials development in the Netherlands and in Japan

Bayer � R&D centre in Germany (Wuppertal)� European R&D in Belgium; merger of Bayer subsidiary Agfa Inc. with Gevaert (1964)� R&D sites in the USA: since 1954 (Mobay in Pittsburgh); acquisition of Cutter Laboratories (1974), Miles

Laboratories (1978) and Cooper Technicon (1989). Several R&D related involvements in the USA in the 80s.Biotechnology research in West Haven (since 1988, 300 p. in 1992). 20% of total R&D budget and c. 2,000employees in the USA

� Acquisition of Canadian Polysar’s Rubber Division in 1990. Continuation of R&D after government interventionfor foreign investment

� R&D sites in Japan: Yuki Research Centre; pharmaceutical research in Osaka (1990); Kyoto (since 1995, 120 p.).About 300 to 400 employees in Japan

Ciba-Geigy � Central research in Great Britain and Japan� Pharmaceutical research in the USA� Divisional development in 13 other countries

Du Pont � 1350 p. at central facility, 2,780 p. in the US, 425 p. abroad. Total of 35 international sites, up from 25 in 1993.Eisai � Research in the USA (organic chemistry, Boston) and in the UK (central nervous system, London)

� Clinical development mainly outside JapanHoechst � R&D in 16 countries: in Western Europe (19% of R&D expenditures; particularly in France, but also in Austria,

Italy, Great Britain and Spain), in the USA (20%) and in Japan (2%). Other foreign R&D through M&A in theUSA and Western Europe

� Capital holding in pharmaceutical company Roussel Uclaf (since 1974) and acquisition of US chemical companyCelanese (1987) intensified R&D internationalization

Kao � European central R&D in Germany (Darmstadt: 30), Barcelona (60–70 p.)� R&D labs in the USA (Pittsburgh: 40 p.), Cincinnati (30–40 p.) and South-East Asia� Smaller R&D sites in Sweden, France, and Great Britain� Total of 300 researchers abroad in 14 locations

Hoffmann-La � Basic research in Switzerland (immunology, Basel) and in the USA (molecular biology, New Jersey)Roche � Industrial R&D in the USA: Alameda (PCR), Nutley (chemical, biological and medical research), Research

Triangle Park (Diagnostics, PCR), San Francisco (biology and medical research);� In Western Europe: Great Britain (Welwyn Garden City, chemical, biological, medical research), Belgium (Gent,

Immunology), France (Strassburg, clinical research; Village-Neuf, animal food), Italy (Milan, Immunology);� In Japan: Kamakura (biological and medical research)� Small R&D centres with special tasks and divisional development in all important markets

Sandoz � Central research in Switzerland (Basel: 580 p. and Berne: 130 p.)� Research in the USA (New Jersey: 250 p.), Austria (Wien: 200 p.) and Great Britain (London: 70 p.)� 46 R&D sites all over the world (mostly development)

Royal Dutch/Shell � Westhollow, Texas (1,350 p.), Bellaire (550 p.), Amsterdam (about 1,390 p.), Oakville (150 p.), Calgary (130p.), Hamburg (140 p.), Arnhem (250 p.), Louvain-La-Neuve (210 p.), Schwabenheim (190 p.), Grand-Couronne(270 p.), Berre (120 p.), Atsugi (190 p.), Sittingbourne (600 p.), Rijswijk (650 p.), Thornton (590 p.)

� Total of 6,900 researchersSchering � Berlin (chemicals)

� R&D in the USA: Biotechnology in Richmond since 1991 (oncology, multiple sclerosis, thrombosis); acquisitionof two biotechnology companies and a RD/Shell site in Richmond

� Diagnostics research and applications development in Japan (Osaka)

Source: Annual reports, company documents, interviews (1994–1997).

sity (R&D expenditures per sales) and R&D interna-tionalization (R&D expenditures and personneldeployment abroad). Based on a classification schemeused by the Institute for System Research and Inno-vation in Karlsruhe, companies with an R&D ratiobetween 3.5 and 8.5% are considered as high-techcompanies, while an R&D ratio above 8.5% indicates ascience-based company. Our chart represents data on acorporate level. It neglects differences in R&D ratioand internationalization on individual business unitsbut captures the shared use of R&D resources availableto several business units. Consequently, some businessunits may have ratios different from their parentcompany as exhibited in Figure 1. As our intent was todemonstrate the overall significance of R&D interna-tionalization in correlation with R&D intensity, weanticipated a certain variance per company, acceptingthe company’s average figures as satisfactoryindicators.

Technology-based companies are attracted to locatetheir R&D activities at centres of technological excel-lence, i.e. regions characterized by a high rate of newtechnology output. The knowledge creation abroad isalso intensified by the scarcity of resources at home.The location patterns we have found (see Table 1)indicate that knowledge-creating processes are asym-metrically distributed worldwide.

Based on this data we can confirm the continuingtrend of R&D internationalization as anticipated inRoberts’ (1995) MIT study. Most of the surveyedcompanies have significantly increased their ratio offoreign R&D to total R&D expenses. Many companiesplan to further increase their number of internationalR&D sites in the next five to ten years (see also Gery-badze et al., 1997, Gerybadze and Reger, 1997).

Although small and medium enterprises (SMEs)increasingly engage in international R&D activities,R&D internationalization of the 80s and 90s haslargely been restricted to multinational companies(MNCs). Our research was thus focused on MNCswith a certain amount of international R&D experi-ence. Technology-based companies from smallcountries have been leading this development. Swiss,Dutch and Belgian companies had more than 50% ofits R&D carried out outside their home country in asearly as 1992 (see Gassmann, 1997a; 53, Granstrandet al., 1993; p. 414). Until recently, most R&D waslocated in the triad regions of USA, Japan and WesternEurope, but we observe a strong trend towards estab-lishing development and design centres in South-EastAsia. Market customization and access to local know-how are the main drivers for this trend, but there is afine balance between a variety of determinants for andagainst R&D internationalization.

3.2. Factors driving and impeding R&Dinternationalization

Shorter product development cycles, global competi-tion, increased customer expectations and technologi-cal risks have been major reasons to internationalizeR&D. While the traditional motivation for internation-alizing corporate activities could be found inimproving cash-flow and turn-over quickly (inter-national exporting and marketing) or reducingoperating costs (e.g. setting up production in low-wagecountries), short-term ROI objectives should not befollowed when internationalizing R&D.

Local markets, resources and synergy potentials areamong the most cited drivers for R&D internationaliz-

150�R&D Management 28, 3, 1998 © Blackwell Publishers Ltd 1998

Oliver Gassmann and Maximilian von Zedtwitz

50

20

30

40

10

100

70

80

90

60

12 181614 202 864 10

R&D Intensity in %

Internationalizationof R&D in % ABB

Ciba

Roche

Eisai

Schering

Sandoz

Philips IBM

Hoechst

Bayer

Schindler SulzerSiemens

BASF

Bosch

Sharp

SonyKao

Matsushita NEC

HitachiN-Steel

UTCMelco

DuPont

Kubota Daimler

R&D Expenses

1-2 billion US-$

0-1 billion US-$

> 2 billion US-$

RD/Shell

Fujitsu

Figure 1. R&D internationalization of technology-based companies (1994–1997).

ation. Synergies will become even more important overthe next years because of the growth in mergers,acquisitions and cooperations.

The development of local products requires the earlyinvolvement of market and customer applicationknow-how, which is more likely to be found inregional business units. Companies with local R&Dexhibit an inclination towards over-emphasizingdifferent local market specification in order to supportlocal autonomy and independence from the parentcompany (image). Host country restrictions, such aslocal content requirements, tolls, import quota, andfulfilment of standards, can attract R&D into keymarket countries ( pull regulations). Reversely, con-straining home country restrictions may inducecompanies to move R&D abroad ( push regulations).

The 90s have seen an effort of many large com-panies to consolidate their activities in order to realizesynergy and coordination potential in internationalR&D. If the corporate R&D network consists ofcompetence centres (e.g. Roche, Schindler), transna-tional R&D projects are managed more easily, giventhat complementary competencies are provided locally.With increasing complementarity of resources, com-petencies, and knowledge bases, as well as the divisionof labour and specialization of work, synergy potentialin R&D projects carried out across several locationscan be exploited.

Spill-over effects in know-how transfer occur whenpeople from different locations become part of thesame team. Informal know-how transfer by sharingexperience between individuals is often superior toformal distribution of knowledge by means of hand-books, blueprints or patent rights. The establishment ofa corporate-wide network rests heavily on those side-

effects of informal contacts between members ofdifferent R&D sites. It is a major management chal-lenge not only to exploit informal networks forinternational R&D activities, but also to foster thecreation of new personal linkages between geographi-cally dispersed R&D units.

Table 2 lists the most important factors for conduct-ing R&D not only in the home country. The reasonslisted here both for and against R&D internationaliz-ation were collected by literature analysis and duringour research interviews. In general, we follow aclassification scheme used by Beckmann and Fischer(1994). Input-oriented factors encompass driversrelated to personnel, know-how and infrastructure.Proximity to markets and customers, collaborationsand image are output- or product-oriented factors.Political and socio-cultural factors such as localcontent rules, technology acceptance and publicapproval times, play an important role for locatingR&D abroad. Direct cost advantages (such as the oftenpublicly discussed labour costs) rarely influence theinternationalization of R&D, but other efficiency-oriented factors such as costs of coordination andtransfer, and critical laboratory size do have an impacton international R&D organization.

In addition, external factors have a great impact onthe dispersion of R&D sites. Many companies interna-tionalize their R&D in the context of mergers andacquisitions (M&As). The scope of all worldwideM&As has doubled from US$369 billion in 1992 to anestimated US$740 billion in 1995. In 1997, the volumeof M&As amounted to 1,630 billion US-$ (SecuritiesData Company, 1998). About 70% of all acquisitionsare based on a market-driven rationale (see Kutschker,1989, p. 12, Granstrand et al., 1993; 416, HØakanson

© Blackwell Publishers Ltd 1998 R&D Management 28, 3, 1998�151

Organization of industrial R&D

Table 2. Drivers of R&D internationalization.

Input-oriented Output-oriented External

� Information and communication networks � National and legal conditions � Acquisition of parent company, merger� Insufficient home personnel � Country-specific cost advantages � Historic reasons� Local pocket-of-innovation � Local economic and natural advantages � Peer pressure� Local infrastructure � Improving local image � Tax optimization� Qualified personnel abroad � Adaptation to local production processes� Local scientific community � Customer-specific development� Tapping informal networks � Closeness to lead users

� Local values� Market and customer proximity

Efficiency-oriented Political / socio-cultural

� Improving flexibility through new organization � Improvement of international patenting laws� Local critical mass / mass degression in home country � Overcoming protectionist barriers� Reduction of development failure risks � Local social and peaceful labour relations� Making use of many time zones � Local content� Closeness to production, marketing, distribution � Legal restrictions in home country� Reduction of development cycle time � Low home acceptance� Overcoming logistic barriers � Subsidies� Lower R&D personnel costs

Sources: Beckmann and Fischer (1994), De Meyer and Mizushima (1989), Gassmann and von Zedtwitz (1996), H×akanson and Zander(1988), Carnegie Bosch Institute (1994), Krubasik and Schrader (1990), von Boehmer et al. (1992).

and Nobel, 1993, p. 402). M&A activities also lead toa reduction of R&D locations. A redirection of stra-tegy following a merger has led Novartis to actuallyreduce the number of central research laboratories byabandoning its former Ciba R&D site in Takarazuka,Japan.

The case for R&D internationalization is not unchal-lenged. Patel and Pavitt (1992) show by means of1969–1986 patent data that research and developmentof core technologies were still highly domesticized in1986. Reasons against internationalization are plenty(see Table 3). We divide our determinants againstinternational R&D into a set of reasons explicitlypreventing R&D internationalization versus a set ofreasons favouring centralization of R&D. The individ-ual items in this list are not necessarily mutuallyexclusive: a reason against internationalization may atthe same time be a strong cause for centralization.Again, as in Table 2, some drivers may be moreprevalent in some industries as they are in others.

Yet, more R&D sites are established each year thanexisting R&D sites are dissolved (Howels, 1990, p.134). Divesting of acquired R&D units is usuallyassociated with high costs (e. g. pension funds pay-ments, social plans) and has a negative impact on thework force morale. Furthermore, valuable technologi-cal competence with possibly high strategic potentialmay be lost (Pearson et al., 1993, p. 259).

International R&D has thus risen from a by-productof business internationalization to a quite importantand far-reaching phenomenon. While this trend offers agreat opportunity for global corporate learning, theunderlying organization must be adapted for the newdemands of international R&D. Our study indicatesthat the organizational evolution towards a worldwidenetwork of interconnected R&D units is characterizedby the establishment of overlaying structures of boththe formal and informal kind. We shall briefly outlinethe trends towards more coordination and integrationof formerly independent R&D before introducing anorganizational framework which we will use todescribe how a company may support formal andinformal means of transnational R&D.

152�R&D Management 28, 3, 1998 © Blackwell Publishers Ltd 1998

Oliver Gassmann and Maximilian von Zedtwitz

Table 3. Determinants of R&D centralization.

Reasons for central R&D Reasons against international R&D

� Economies of scale � Immobility of top class personnel� Synergy effects � Critical mass� Optimal career planning � Parallel development� Minimal R&D costs and development time � Language and culture problems� No cases of ‘not-invented-here’ syndrome � Scientific and technical information available worldwide� Allocation of technological competence to SBU is simple � Specific know-how easily lost when support not present� Better control over basic research results � Local workshops are easier� Common R&D culture � Political risks

� No wage advantages in triad nations� Coordination and information costs

Sources: Behrmann and Fischer (1980), De Meyer and Mizushima (1989), Coombs and Richards (1993), Carnegie BoschInstitute (1994), von Boehmer et al. (1992), Beckmann and Fischer (1994), Gassmann and von Zedtwitz (1996).

3.3. Concentration on leading researchcentres and tight coordination

In accordance with the results of Casson et al. (1992),we notice a trend towards the integrated R&D networkstructure. This trend is characterized by two keydevelopments: an increase of authority of foreignR&D sites by assigning strategic roles and missions,and an integration of formerly loosely coordinatedR&D sites into a competence-based network. Once anintegrated network structure is achieved a second trendmay be observed: a concentration on leading high-techregions, and a re-centralization of decision processesin few competence centres. General Motors and ABBare two examples for these trends.

General Motors (GM) tried to reduce R&D costs bytightening R&D coordination. Aiming at a global R&Dmanagement without inefficient duplicate develop-ment, GM has started to restructure its worldwideR&D network. It established four regional competencecentres in order to centralize the development of newplatforms. Based on these platforms, the developmentof variants is carried out locally. Besides the reductionof the number of variants, GM tries to implementglobal standards for ‘invisible parts’ and regionaldifferentiation of ‘visible parts’.

ABB successfully oversees 1,300 companies, 5,000profit centres, 50 business units, 5 divisions and a greatnumber of globally distributed and independent R&Dcentres. The current dilemma with ABB’s integratedcorporate network has been formulated by CEOBarnevik: ‘ABB has three internal contradictions. Wewant to be global and local, big and small, decentral-ized with central reporting’. In 1994, ABB tried tobring down the costs of running this network bytightening and centralizing the coordination of R&Dactivities. Corporate management was reduced to eightmembers drawn from all four product divisions. Thereorganization strengthened the four product divisionsnoticeably and led to a recentralization towards fewerR&D locations. Their autonomy is controlled bycentralizing the important financial and informationflows within top management.

Many companies have realized that the benefits ofsuch a network structure exceed the high coordinationcosts if the network is properly managed. The econ-omies of scale realized by advanced coordination ofworldwide R&D, a reduction of duplicate develop-ment, the realization of synergies, and a concentrationon internal technology transfer across boundaries arethe advantages of such a network. Summarizing thetrends in the entire research sample, we observe sixmajor trends that affect multinational technologycompanies:

1. Increasing R&D internationalization. The relativeshare of foreign R&D has dramatically risen in thelast 15 years. Technology-based companies head-quartered in small Western European countrieshave assumed a pioneering role, since their homecountry could not offer a sufficiently large resourcepool for R&D nor a market of significant size. Allof the surveyed companies have increased theirR&D expenditure to sales ratios between 1980 and1995. They also set up new foreign R&D labora-tories and reinforced existing ones.

2. External orientation. Ethnocentric orientedcompanies such as Daimler, Nissan, Volvo, andToyota increasingly participate in internationalcentres of technology and know-how. Typically,this happens via strategic technology alliances,foreign manufacturing, and integration of localtechnical support in product development.

3. Listening post. A number of companies withcentralized R&D have begun to set up tightlycontrolled technology listening posts in centres ofinnovation. Japanese basic research laboratories inEurope and the USA are characteristic for thisdevelopment.

4. Empowerment. Foreign R&D sites are assigned astrategic and active role in the R&D network. Thebalance between autonomy and control shiftstowards more competence for the local unit.

5. Integration. R&D units which have experiencedlittle coordination will find themselves moreintegrated in the future. Corporate restructuringprogrammes promote some units to centres-of-excellence, which are responsible for a specificproduct group or technology field worldwide.

6. Recentralization. In heavily decentralized com-panies such as ABB, IBM and Hoechst we couldobserve a trend towards recentralization of R&Dinto a handful of leading centres. Inefficientlocations are divested and existing competencecentres more tightly controlled.

Companies that have grown by M&A typically facethe need of R&D integration (Gerpott, 1994). Com-panies whose R&D organizations have more steadilyevolved follow a more sequential development ofexternal orientation, listening post establishment andempowerment. In Gassmann (1997a, b) and Gassmann

and von Zedtwitz (1998), we describe the drivers andtrends of this organizational change in more detail. Allof the surveyed companies could be observed to strivefor a presence in a small number of leading innovationareas. Simultaneously, companies increasingly try toexploit possible synergies between decentral locations:global efficiency becomes more important than localeffectiveness.

4. Overlaying structures for global R&Defficiency

A discussion of the issues for global R&D manage-ment includes the following arguments:

i. development of R&D and technical standardsii. managing cultural diversityiii. coordination of decentralized processesiv. educating and training the global R&D project

managerv. retention, integration and utilization of dispersed

know-how.

In order to manage global processes it is helpful tounderstand the four levels of structure in internationalR&D organizations. The regional and hierarchicalstructures are the foundation for transnational organiz-ations, being simultaneously responsible for manybarriers in global R&D management. While these twostructures have been inappropriate to deal with the herelisted issues of global R&D management effectively,overlaying and informal structures can help to over-come these barriers. We will see that solutions oftenevolve naturally if implemented in a superimposed,overlaying structure (Figure 2).

We define the overlaying structure as the sum of allcorporate mechanisms and procedures (such as cross-functional project teams) that do not fit into the generalhierarchical line structure. Due to the rising demandson flexibility and dynamics in the company, thisoverlaying structure becomes increasingly important.The overlaying structure affects the creation andsustainability of an informal network encompassingrelations and linkages both within and outside thecompany. Since this informal network structurepowerfully transcends the entire organization, weexpect great potentials to be exploited by managingthis network structure more deliberately.

The idea of viewing an organization as a set ofinterlinked structures is not new. On the one hand, ourmodel is related to the hypertext organization ofNonaka and Takeuchi (1995, pp. 166–171), which isbased on a Japanese publication by Nonaka and Konnoin 1993. Nonaka and Takeuchi’s hypertext organiz-ation consisted of a business-system layer, a project-team layer, and a knowledge-base layer, and is consid-ered a dynamic synthesis of both the bureaucraticstructure and the hierarchy-spanning task force. On the

© Blackwell Publishers Ltd 1998 R&D Management 28, 3, 1998�153

Organization of industrial R&D

other hand, Schoonhoven and Jelinek (1990) introducethe concept of a quasi-formal structure. Thisintermediary structure is between a formal structuredefined as subunits, positions, and reporting relation-ships and an informal structure consisting ofunsanctioned patterns of interaction devised aroundsocial and task requirements the formal structurefailed to take into account.

Our model puts more emphasis on the informalstructure and introduces the regional structure as amajor determinant for the overall organization. In thefollowing we will concentrate on some importantelements of the linkages from the overlayingstructure  to the hierarchical and to the informalstructure. These elements are technology offices,technology management competence centres, virtualproject management pools, international job rotationschemes, project offices, and project manager careerplanning.

4.1. Schering’s technology officeThe purpose of the technology office is to seek newand relevant information and to distribute it within thecompany. Information is researched upon request orprovided internally when deemed relevant. The tech-nology head office of Schering is located in Berlin,Germany (ten people), with an additional six people inRichmond, USA and four more in Osaka, Japan. Theoffice functions as the company’s ‘eye and ear’. In thisrole, it stays in close contact with the scientific com-munity, organizes seminars and conferences, and looksafter the diffusion of critical information to Schering’sresearchers. The personal and often informal contactsare a very important element to the success of the

technology office, which is often forgotten whenoperating in network-like structures.

If R&D must be outsourced due to time constraintsor insufficient internal know-how, the office initiatescontracts with external R&D partners. In this case, therespective units or departments will contact the officeand state the internal deficit and its request for out-sourcing. The office then initiates a search process inorder to find the required knowledge. Forecasting,evaluation and distribution of global available tech-nology is one of the main tasks of the office. Insummary, the technology office relieves researchersfrom many administrative and procedural chores,while supporting them in their core research activities.

4.2. Schindler’s technology managementcompetence centre

Technology management should not be seen as a one-time project but rather as an ongoing process. Schin-dler therefore established technology management as acompetence centre. This centre combines two tra-ditional departments of Advanced Development andCompetitor Technical Observation, along with whatwas known as strategic technology management. Themission of Schindler’s TM is ‘the integrated manage-ment of new technologies for the maximum benefit ofthe Schindler Corporation’. More precisely, it isresponsible for the tasks given in Figure 3 andexplained subsequently.

In advanced development, Schindler’s TM aims atquantum leap developments in elevators throughradical new approaches. New product concepts aredefined and studies show the feasibility of suchconcepts. Advanced development is undertaken in very

154�R&D Management 28, 3, 1998 © Blackwell Publishers Ltd 1998

Oliver Gassmann and Maximilian von Zedtwitz

I

IV

III

II

RegionalStructure

HierarchicalStructure

OverlayingProject/ProcessStructure

Informal/RelationshipStructure

Personal andinformal relations

Cross-boundaryorganizational units

Hierarchical links

R&D Centres

Figure 2. Four levels of structure in international organizations.

close cooperation with universities and external tech-nology leaders in clearly defined areas. According tothe core competence philosophy, Schindler is increas-ingly engaged in strategic technology alliances withsuppliers who are technology leaders in their fields.

Technology and industry monitoring promotessystematic scanning and evaluation of new tech-nologies which arise in other industries. These areevaluated in terms of their applicability to elevatorsand escalators.

By means of competitor technical intelligence,Schindler’s TM identifies technological activities ofcompetitors. So-called ‘weak signals’ are identifiedwith a variety of instruments based on e. g. patentanalysis and publications.

Knowledge management optimizes information flowand the knowledge base of people. The goal is thedevelopment of the company’s R&D towards organiz-ational learning. Several state-of-the-art methods andinstruments such as knowledge maps, who-is-who listsand gatekeeper meetings are used in order to makeknowledge management operational. This means thatgatekeepers in R&D teams must be identified andbrought together in interdisciplinary technology fora.

Schindler’s TM is responsible for updating anddeveloping technical core competencies and tech-nology portfolios in close cooperation withdevelopment centres. Schindler’s TM also promotes theimplementation of core competencies in the R&Dorganization and the development of skills in accor-dance with the human resource department.

By engaging in benchmarking, Schindler’s R&Dcompares itself with worldclass R&D in leadingtechnology-based companies. Learning from ‘bestpractice’ across industries is promoted, in order toimprove its R&D organization, processes and tools.This also includes the implementation of the ‘lessonslearned’.

Internally, Schindler’s TM acts as a change agent toopen the mind of every R&D employee: according to

3M’s model, every employee should think about newtechnologies and radical innovations for some part ofhis time. Externally, Schindler’s TM takes care ofR&D public relations for attracting the best graduatesfrom universities. Through contacts with universitiesand scientific publications, Schindler wants to beclearly identified as a technology leader in its field.

The team at the technology management competencecentre is highly skilled and very interdisciplinary.Building on their multicultural background (Swiss,French, German, Egyptian, Chinese and Malaysian),these people reinforce Schindler’s R&D with newtechnologies, radically new concepts, state-of-the-artmethods and R&D tools. Schindler’s competencecentre combines aspects of Schering’s technologyoffice with the integration of advanced development,listening post and analysis activities.

4.3. Central project managementLearning processes that span several locations are morecomplex than sporadic exchange of information andexperience during a single project. In order to institu-tionalize and support these processes, a coordinationoffice is required. This office ensures the continuedinformation flow between all nodes of the R&Dnetwork.

A corporate office called Central Project Manage-ment maintains this role at Bayer. This office providesR&D units with standardized project managementinstruments such as project manuals, developmentguides, IS support, and project management systems.If development is to be carried out routinely, ‘standardoperating procedures’ become very important.Coordinating decentralized R&D procedures andinformation flows is the main purpose of this projectmanagement unit.

IBM’s solution is similar: for the coordination ofdecentralized VSE (Virtual Storage Extended) systemdevelopment, IBM’s project office plays a central role.

© Blackwell Publishers Ltd 1998 R&D Management 28, 3, 1998�155

Organization of industrial R&D

IM

TT

HRBC

M1 M2 M3SY

SE

R&D

Technology& IndustryMonitoring

CompetitorTechnical

Intelligence

AdvancedDevelopment

ChangeAgent& PRCore

Competencies

Benchmarking,R&D Tools

KnowledgeManagement

TechnologyManagement

Figure 3. Technology Management as a new competence centre at Schindler Lifts.

A total of 20 components of the VSE system is neededto be coordinated, controlled and managed. In teams oftwo, each member of the office is responsible for anew release of the system. The number of membersfluctuates in accordance with the number of projectsassigned to the office. Among the tasks they deal withare the collection of requirements and specificationsfor a VSE release, the technical evaluation of projectideas, technical systems design, project controlling andcoordination, project documentation, and VSE productplanning.

Experienced senior advisory programmers screen,evaluate and refurbish new project ideas and pro-posals, eventually putting together a formal projectproposal. All requirements are listed and prioritized. Ifa completely new project idea comes up, the officefinds out which parts may be developed by the originalproponent and which parts may be contracted to otherR&D sites (Figure 4).

Project supervision and coordination requires aworldwide integration of all VSE development projectswith respect to deadlines, milestones and project targets.After long years of experience, this process has beenwidely routinized by now. Nevertheless, the complexityof the task is enormous: the number of technical inter-faces between different R&D sites and the incorporationof previous project results require close attention.

The management of programme interfaces is decen-tralized and self-regulating. A ‘workbook’ includes allspecifications of the new VSE system and designchange requests. The project office freezes thespecifications once a certain milestone (or checkpoint)has been reached, as incongruent interfaces lead tohigh adaptation costs in later stages of development.Apart from the interfaces, the office also specifieswhen a certain programme module has to be delivered.

Every project member has access to the projectstatus document. This document is regularly updatedby the office and serves as an information tool foridentifying if a milestone has been reached and whenthe project may move into another phase. The projectoffice also coordinates the development of nationallanguage versions of the VSE operating system. TheEnglish original is transferred to local R&D sites in therespective countries and translated into Japanese,German, and Spanish. In short, operative managementof project schedules and technical interfaces in highlystandardized development processes is the main task ofIBM’s project office.

4.4. Cross-cultural job rotationSome companies are particularly good at internalizingexternal know-how through hosting visitingresearchers. This is especially true for those companieswithout overseas R&D sites. Hosting foreignresearchers and engineers in the parent company can bea major source of innovation. If the company hasalready established overseas R&D, then visitingresearchers can also be R&D members of another unitof the same company who have been sent abroad inorder to learn about local research. Such visitingresearchers are often assigned to collaboration projectsthat are funded by a business unit.

The Advanced Telecommunication Research Lab-oratories (ATR) – a Japanese basic research company– has recognized the importance of foreign input andestablished intensive relations with overseas researchcommunities. Research projects are defined based onthe competence and interest of new researchers, aslong as the projects fit in the overall research pro-gramme. A large number of foreign researchers work

156�R&D Management 28, 3, 1998 © Blackwell Publishers Ltd 1998

Oliver Gassmann and Maximilian von Zedtwitz

Projectmanager

Project office

• Information• Coordination • no directive authority

• Gathering VSE requirements

• Evaluating project ideas• Suggesting location of

development

• Product planningsupport

• Coordination ofcountry-specific releases

• Project documentation• Project controlling• Technical systems design• Work-flow design

Decentralized teams

• Objectives• Directives

Figure 4. The tasks of the VSE-Project-office at IBM.

in mixed teams with Japanese scientists, but all ofR&D management remains in Japanese hands. ATRemploys about 50 invited foreign researchers forperiods from one month up to several years, whilesome of the foreign researchers hold long-term jobcontracts. For instance, the ATR Human InformationProcessing Laboratory employed a total of 48 Japaneseresearchers and 35 foreign researchers in 1994. Themajority of the Japanese researchers come from over40 supporting companies and generally stay for threeyears before they return to their parent companies. Thisresulted in about 800 Japanese researchers and about400 international researchers for the past ten years.ATR researchers produce a high amount of internation-ally coauthored articles. Furthermore, collaborationsspin off after the foreign researcher has been trans-ferred back to his home country. Thus, althoughentirely domestically organized, ATR is a highlyinternationalized company.

Hicks et al. (1994) found that 91 of the Japanesecompanies in the JPC (1990) sample of 149 firmsemployed no foreign researchers at all. 88% of theremaining companies employed less than ten foreignresearchers. Because the international rotation ofscientists and engineers is a quick and economical wayof learning about external and foreign R&D (de Meyer,1992), it is often a first step towards establishing aglobal R&D network. Several Japanese companieswhich we interviewed currently pursue this approach toglobalization. For instance, Nippon Steel set up a‘Visiting Researcher Program’ with the aim to activateR&D by introducing expert researchers worldwide andto promote a stimulative R&D environment. NipponSteel has accepted over 70 researchers in ten years,sending eleven researchers abroad for study annually.Similarly, the HIVIPS program introduced by Hitachiin 1984 was devised to realize synergy for innovation

through heterogeneous cross-cultural interaction. Thisprogram is now seen as a major device for internalR&D globalization at Hitachi (Kuwahara, 1998).

Although similar programmes are undertaken inother companies, the extent of international personnelexchange remains limited. Thus the Japanese corporateresearch workforce is mainly staffed by Japanese(Hicks et al., 1994), and cannot be compared indiversity with its American or European counterparts.

Because earlier and later stages of R&D are oftenperformed in different locations, the transfer of tech-nology and know-how crosses not only organizationalbut also geographical distances. The Japanese chemicalcompany Kao conducted an 18-month developmentproject of hair cosmetics treatment. While such pro-jects had been predominantly carried out in Japan,local adaptation for culture, different treatment styles,and physiological differences required the integrationof R&D close to the eventual customer. While conceptdevelopment and advanced research took place inTokyo, much of the actual product development wascarried out in regional R&D centres. Figure 5 depictsthis concept in an example of hair treatment develop-ment in Tokyo and Darmstadt.

Know-how between Japanese advanced developmentand German development was transferred by mutualpersonnel exchange. During the concept phase, aGerman scientist was sent to Tokyo to be part of theconcept creation in the local project team. During thedevelopment phase in Darmstadt, the Japanese projectmanager supported the German team for two months.In this case, personnel continuity ensured technologyand know-how transfer.

The main objective to exchange or invite researchersis manifold. The cross-cultural rotation supportsthe  construction of a professional network, fosterscreativity, raises an awareness for cultural differences,

© Blackwell Publishers Ltd 1998 R&D Management 28, 3, 1998�157

Organization of industrial R&D

Key project memberdispatched forthree months

Project managerdispatched for

two months

P11

P22

P21

P22

P23

PL

P12

PL

Team 1Tokyo

Team 2Darmstadt

tConcept Creation(Advanced Development)

Formulation(Development)

Figure 5. Personnel exchange at Kao.

and ensures transnational project continuity andefficiency.

4.5. Virtual project management poolsInternational R&D processes result in new challengesfor human resource management. International recruit-ing, administration and development of R&Dmanagers and scientists is not restricted to one locationanymore but must be coordinated on a global scale.Especially the management of a project manager’scareer is a challenge that is rarely satisfactorilyaddressed.

Although personnel development in functional andline management is highly developed in many com-panies, it appears problematic to identify andsystematically develop a pool of R&D project mana-gers. Typically, qualified potential project managersare not recognized as such and miss chances for theirown career development (see also Domsch, 1993, p.163). On a corporate level this leads to inefficienthuman resource allocation and suboptimal individualcareer planning. One of the most important aspects ofpersonnel development is therefore the increasingsupport for ‘special design’ project manager careersoutside the conventional line promotion. Humanresource departments thus concentrate on financialincentives, formal recognition of project managers,and providing explicitly individual career springboards.

Bosch has set particular emphasis on the develop-ment and support of dual career ladders in its R&Dproject manager career planning. Based on an incentivescheme, successful project managers are assigned to

larger projects after termination of their former pro-jects, thus enabling them to rise in the ranks. Usually,they start with projects in component development orsubprojects. Subsequently, they may assume tasks withmore responsibility, more strategic orientation, andincreased degree of complexity and international-ization.

The systematic promotion of human resourcedevelopment strategies may be succeeded by a special-ized project management department (Figure 6).Hoffmann-La Roche’s Pharma Division established adepartment called ‘International Project Management’which coordinates a resource pool of about 50 projectmanagers for all R&D projects worldwide. Thisdepartment consumes about 30% of the pharmaceuticalR&D expenditures, having highly strategic importancefor the innovative potential in this business.

Every project manager is fully assigned to thisgeographically decentralized department. The directorof this ‘virtual’ resource pool assigns his people asmanagers to projects also as part of a global pro-gramme to ensure standards in quality and projectprocedures. Upon completion of his project, a projectmanager is returned to the resource pool. As there aremore projects in the pipeline as potential projectsmanagers available, he is immediately reassigned to anew project.

Since this relative new department reports directly tothe Corporate Board, the internal political position ofR&D project managers has been improved recently.Furthermore, the first director of the ‘InternationalProject Management Department’ was the wife of thethen R&D director, which increased the significance of

158�R&D Management 28, 3, 1998 © Blackwell Publishers Ltd 1998

Oliver Gassmann and Maximilian von Zedtwitz

Pharma R&D

Staff Functions• International Controlling• Quality Assurance• Human Resources

Pre-clinical R&DBasel, Nutley,

Palo Alto, Welwyn, Kamakura

International Development

Worldwide

International Clinical Research

Worldwide

International Project Management

Basel

(1) 6 months beforeproject approval:request for projectmanager

(2) Assignment of project manager to a project,independent of location

(3) After project termination:Project manager returnsto pool and is readyfor new assignment

Virtual Project Manager Pool• 50 project managers,

worldwide distributed• reporting to Project

Management Dept. in Basel

Figure 6. Project management department as a virtual project management pool at Roche.

the department. The establishment of a project mana-ger pool is a clear signal for empowering anddeveloping one of the scarcest competitive resources,hands-on operative management with long-termperspectives. Roche thus manages to retain much ofthe valuable procedural know-how to conduct and leadinternational projects not only inside the company, butalso in a position where it can be reapplied whenneeded.

The director of the ‘International Project Manage-ment Department’ is also in the International ProjectCommittee, which decides over the roughly 60 globalR&D projects at Roche Pharma. The director assumesa role as interpreter or liaison between project mana-gers and top management, thus representing theinterests of international project management at Roche.This virtual pool promotes the project managementidea. Experienced project managers are dispersedaround the world and wander from project to project –no matter where the project will be conducted. Thisvirtual pool is more than just a international matrixorganization as it expands and shrinks responding tothe requirements of the international developmentcentres.

5. Conclusions

The internationalization of R&D remains a key man-agement challenge for the next years. Technology-based companies will continue to disperse their R&Dactivities around the world. Our research indicates atrend towards a concentration on leading high-techareas and a tendency to tighten the coordination ofinternational R&D activities in competence centres.

The conventional organization of firms is inadequatefor the requirements of modern global R&D because itis constrained by hierarchical and regional barriers.New organizational forms must come to the rescue.Among those requirements are the development ofmanagerial and technical standards in R&D, managingcultural diversity, coordination of decentralizedprocesses, educating and training the global R&Dproject manager, and the use and integration of dis-persed know-how.

An overlaying structure presents a platform onwhich these requirements can be directly addressed.We have found and described a number of organiz-ation forms that successfully respond to abovementioned issues in global R&D.

The institutionalization of technology managementas a competence centre at Schindler is an excellentexample for an overlaying structure implemented onregional and functional structures. The management ofnew technologies is facilitated by integrating advanceddevelopment, competitor technical observation, andstrategic technology management.

Project offices are suitable for highly routinized,

incremental innovation projects, characterized byclearly defined responsibilities. In the IBM example,this is the case in the development of a new softwarepackage release. A central project management asfound at Bayer ensures common standards which helpreduce the complexity associated with decentralizedR&D processes. This is done by providing coordina-tion and a common IT infrastructure.

The practice of exchanging researchers or sendingscientists abroad is an excellent source of new ideasand fosters creativity for all involved. Personal link-ages are formed, and mutual trust and understandingnot only increases the cultural awareness of theindividual and his parent company, but also expeditesinterlocal project execution (e.g. around-the-clockresearch) and reduces the risk of NIH-syndromes.Companies use personnel transfer as a very efficientknowledge-management tool especially in the pre-competitive phases of the R&D process. If the com-pany has no foreign R&D sites, it can resort toinviting external scientists and offer them excellentresearch and collaboration conditions in their homebase.

Virtual project management pools like the one atRoche ensure professionalism and quality of projectexecution. Highly trained and motivated projectmanagers are an incredible asset for a companydepending on the success of each of their R&D pro-jects. It is the project manager and not the R&Dengineers who should control the innovation process.Access to external know-how as well as the dissemina-tion of internal knowledge is supported by atechnology office serving as ‘the eye and ear’ of thecompany. No company in a knowledge- or science-intensive industry can afford to neglect the search forexternal information especially in the early and explor-atory stages of technology development.

Organizing industrial R&D for global efficiency isan often neglected task in corporate management. Ourresearch has shown that integrated R&D networkswhich are based on the core competencies of its R&Dunits are an appropriate solution for many multina-tional technology firms. The described overlayingstructures are part of the glue that holds the networkunits together, improving efficiency and effectivenessof global R&D.

Acknowledgments

We would like to thank numerous interview partnerswho have spend their time to share valuable informa-tion with us. Many of them have helped to correcterrors in the company data. Nonetheless, the viewspresented in this paper are the sole responsibility of theauthors. We are also indebted to Ursula Koners,Gregory Huber, and two anonymous referees for theirinsightful comments on earlier drafts of this paper.

© Blackwell Publishers Ltd 1998 R&D Management 28, 3, 1998�159

Organization of industrial R&D

References

Amstad, M., Arvanitis, S. and Hollenstein, H. (1996) Wieinnovativ ist die Schweizer Industrie im internationalenVergleich? In Gassmann, O. and von Zedtwitz, M. (eds),Internationales Innovationsmanagement, M×unchen:Vahlen, 231–258.

Beckmann, C. and Fischer, J. (1994) Einflufaktoren auf dieInternationalisierung von Forschung und Entwicklung inder Chemischen und Pharmazeutischen Industrie. Zeit-schrift f×ur Betriebswirtschaftliche Forschung, 46, 7/8,630–657.

Behrmann, J.N. and Fischer, W.A. (1980) Transnationalcorporations: market orientations and R&D abroad.Columbia Journal of World Business, 15, 3, 55–59.

Beise, M. and Belitz, H. (1996) Internationalisierung vonF&E multinationaler Unternehmen in Deutschland. InGassmann, O. and von Zedtwitz, M. (eds), Inter-nationales Innovationsmanagement, M×unchen: Vahlen,215–230.

Brockhoff, K. (1997) Industrial Research for FutureCompetitiveness, Heidelberg, Berlin, New York: Springer.

Buderi, R, Weber, J., Hoots, C. and Neff, R. (1991) Atighter focus for R&D. Business Week, 2 December,80–84.

Cantwell, J. (1992) The internationalisation of technolog-ical activity and its implications for competitivness.Granstrand, O., HØakanson, L. and Sj×olander, S. (eds),Technology Management and International Business:Internationalization of R&D and Technology, Chichester:Wiley, 75–95.

Carnegie Bosch Institute (1994) Managing InternationalResearch and Development, Working Paper 94–8,International Executive Forum of the Carnegie BoschInstitute: Pittsburgh.

Casson, M., Pearce, R.D. and Singh, S. (1992) Businessculture and international technology: research managers’perceptions of recent changes in corporate R&D. InGranstrand, O., HØakanson, L. and Sj×olander, S. (eds),Technology Management and International Business:Internationalization of R&D and Technology, Chichester:Wiley, 117–135.

Cheng, J.L.C. and Bolon, D.S. (1993) The management ofmultinational R&D: a neclected topic. Journal of Inter-national Busines Research, 1, 1–18.

Coombs, R. and Richards, A. (1993) Strategic control oftechnology in diversified companies with decentralizedR&D. Technology Analysis & Strategic Management, 5,4, 385–396.

Dalton, D. H. and Serapio, M.G. (1995) GlobalizingIndustrial Research and Development. Washington, DC:US Department of Commerce.

De Meyer, A. (1992) Management of international R&Doperations. Granstrand, O., HØakanson, L. and Sj×olander,S. (eds), Technology Management and InternationalBusiness: Internationalization of R&D and Technology,Chichester: Wiley, 163–179.

De Meyer, A. and Mizushima, A. (1989) Global R&Dmanagement. R&D-Management, 19, 2, 135–146.

De Pay, D. (1989) Die Organsation von Innovationen. DieAnwendung des Dekompositionsprinzips von William-son.  In Albach, H. (ed.), Organisation, Wiesbaden,289–313.

Domsch, M. (1993) Laufbahnentwicklung f×urIndustrieforscher.

Domsch, M., Sabisch, H. and Siemers, S.H.A. (eds), F&EManagement, Stuttgart, 153–178.

Gassmann, O. (1997a) Internationales F&E-Management,M×unchen, Wien: Oldenbourg.

Gassmann, O. (1997b) Organisationsformen internationalerF&E. Zeitschrift f×ur Organisation, 6, 332–339.

Gassmann, O. and von Zedtwitz, M. (1998) Towards theIntegrated R&D Network − New Aspects of OrganizingInternational R&D. In Lefebvre, L.A., Mason, R.M. andKhalil, T. (eds) Management of Technology, SustainableDevelopment and Eco-Efficiency, New York, Oxford,Tokyo 1998, pp. 85−99.

Gassmann; O. and von Zedtwitz, M. (1996) Ein Referen-zrahmen f×ur das Internationale Innovationsmanagement.In Gassmann, O. and von Zedtwitz, M. (eds), Inter-nationales Innovationsmanagement, M×unchen: Vahlen,3–15.

Gerpott, T.J. (1991) Globales F&E-Management: Bausteineeines Gesamtkonzepts zur Gestaltung eines weltweitenF&E-Standortsystems. In Booz, Allen und Hamilton(eds), Integriertes Technologie- und Innovationsmanage-ment, Berlin, 49–73.

Gerpott, T. (1994) Successful integration of R&D functionsafter acquisitions: An explorative empirical study. R&DManagement, 25, 2, 161–178.

Gerybadze, A., Meyer-Krahmer, F. and Reger, G. (eds,1997) Globales Management von Forschung und Inno-vation, Stuttgart: Sch×affer-Poeschel.

Gerybadze, A. and Reger, G. (1997) Globalisation of R&D:Recent Changes in the Management of Innovation inTransnational Corporations, Working paper 97–01 atUniversity of Hohenheim. Stuttgart.

Granstrand, O., HØakanson, L. and Sj×olander, S. (1993)Internationalization of R&D – a survey of some recentresearch. Research Policy, 22, 413–430.

HØakanson, L. (1981) Organization and evolution of foreignR&D in Swedish multinationals. Geografiska Annaler,63B, 1, 47–56.

HØakanson, L. (1992) Locational determinants of foreignR&D in Swedish multinationals. In Granstrand, O.,HØakanson, L. and Sj×olander, S. (eds) TechnologyManagement and International Business: Internationaliz-ation of R&D and Technology, Chichester: Wiley, 97–115.

HØakanson, L. and Nobel, R. (1993) Determinants of foreignR&D in Swedish multinationals. Research Policy, 22,397–411.

HØakanson, L. and Zander, U. (1988) International manage-ment of R&D: the Swedish experience. R&DManagement, 18, 3, 217–226.

Hicks, D., Ishizuka, T., Keen, P. and Sweet, S. (1994)Japanese corporations, scientific research and globaliz-ation. Research Policy, 23, 375–384.

Howells, J. (1990) The location and organisation of researchand development: new horizons. Research Policy, 19,133–146.

JPC (Japan Productivity Centre/St. Paul Rikkyo University)(1990) Utilization of Foreign Universities and OtherResearch Institutes by Japanese Corporations.

Krubasik, E.G. and Schrader, J. (1990) Globale Forschungs-und Entwicklungsaktivit×aten. In Welge, M.K. (ed.),Globales Management, Stuttgart, 17–27.

160�R&D Management 28, 3, 1998 © Blackwell Publishers Ltd 1998

Oliver Gassmann and Maximilian von Zedtwitz

Kutschker, M. (1989) Akquisition, internationale. In:Macharzina, K. and Welge, M.K. (eds), Handw×orterbuch:Export und internationale Unternehmung, Col. 2–22.

Kuwahara, Y. (1998) Management practices for innovationin global industrial research. Boutellier, R., Gassmann, O.and von Zedtwitz, M. (eds), International Organization ofIndustrial R&D, Heidelberg, Berlin, New York: Springer,forthcoming.

National Science Board (1996) Science & EngineeringIndicators – 1996. Washington, DC: US GovernmentPrinting Office, NSB 96–21.

NIW, DIW, ISI, ZEW (1995) Zur TechnologischenLeistungsf×ahigkeit Deutschlands. Erweiterte Berichterstat-tung 1995. Endbericht an das Bundesministerium f×urBildung, Wissenschaft, Forschung und Technologie(BMBF). Hannover, Berlin: BMBF.

Nonaka, I. and Takeuchi, H. (1995) The Knowledge-Creating Company, New York: Oxford University Press.

Papanastassiou, M. and Pearce, R. (1994) The internation-alisation of research and development by Japaneseenterprises. R&D Management, 24, 2, 155–165.

Patel, P. (1996) Are large firms internationalizing thegeneration of technology? Some new evidence.IEEE  Transactions on Engineering Management, 43, 1,41–47.

Patel, P. and Pavitt; K. (1992) Large firms in the productionof the world’s technology: an important case of non-globalisation. Granstrand, O., HØakanson, L. and Sj×olander,S. (eds) Technology Management and International

Business: Internationalization of R&D and Technology,Chichester: Wiley, 53–74.

Pearce, R.D. and Singh, S. (1991) The overseas laboratory.Casson, M. (ed.), Global Research Strategy and Inter-national Competitiveness. Oxford: Basil Blackwell,183–212.

Pearson, A., Brockhoff, K. and von Boehmer, A. (1993)Decision parameters in global R&D management. R&DManagement 23, 3, 249–262.

Roberts, E.B. (1995) Benchmarking the strategic manage-ment of technology I. Research Technology Management,38, 1, 44–56.

Schoonhoven, C. and Jelinek, M. (1990) Dynamic tension ininnovative, high technology firms: managing rapid tech-nological change through organizational structure. In vonGlinow, M. and Mohram, S. (eds), Managing Complexityin High Technology Organization, New York: OxfordUniversity Press, 90–118.

Securites Data Company (1988) Worldwide Mergers andAquisitions. In Neue Z×urcher Zeitung (eds), Ein SteinigerWeg f×ur Junge Biotech-Firmen, 9 December 1977, No.286, p. 24.

von Boehmer, A. Brockhoff, K. and Pearson, A.W. (1992)The management of international research and develop-ment. in: Buckley, P.J. and Brooke, M.Z. (eds)International Business Studies, Oxford, 495–509.

Wortmann, M. (1990) Multinationals and the internationaliz-ation of R&D: new developments in German companies.Research Policy, 19, 2, 175–183.

© Blackwell Publishers Ltd 1998 R&D Management 28, 3, 1998�161

Organization of industrial R&D